Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
listlengths 1
91
| keywords
listlengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
Motion tracking is a challenge the visual system has to solve by reading out the retinal population . It is still unclear how the information from different neurons can be combined together to estimate the position of an object . Here we recorded a large population of ganglion cells in a dense patch of salamander and guinea pig retinas while displaying a bar moving diffusively . We show that the bar’s position can be reconstructed from retinal activity with a precision in the hyperacuity regime using a linear decoder acting on 100+ cells . We then took advantage of this unprecedented precision to explore the spatial structure of the retina’s population code . The classical view would have suggested that the firing rates of the cells form a moving hill of activity tracking the bar’s position . Instead , we found that most ganglion cells in the salamander fired sparsely and idiosyncratically , so that their neural image did not track the bar . Furthermore , ganglion cell activity spanned an area much larger than predicted by their receptive fields , with cells coding for motion far in their surround . As a result , population redundancy was high , and we could find multiple , disjoint subsets of neurons that encoded the trajectory with high precision . This organization allows for diverse collections of ganglion cells to represent high-accuracy motion information in a form easily read out by downstream neural circuits .
Our current understanding of how sensory neurons collectively encode information about the environment is limited . Being able to “read out” this information from neural ensemble activity is a major challenge in neuroscience . Discriminating among a discrete set of stimuli based on their sensory responses has been attempted in brain areas including the retina [1–3] , sensory cortex [4 , 5] , and motor cortex [6] . Some studies have attempted the more difficult task of reconstructing a time varying signal from neural activity [7–9] . Being able to reconstruct a dynamical stimulus from the neural activity would greatly improve our understanding of the neural code , but there have been only a few attempts in the visual system [10–12] . The retina is an ideal circuit in which to attempt to decode a dynamical stimulus because recordings from a large , diverse , complete population of ganglion cells—the retinal output—have recently become possible [13 , 14] . In contrast to cortical recordings , it is possible to drastically reduce the proportion of “hidden variables” in the network , i . e . unrecorded neurons that could carry relevant information . Furthermore , since the retina encodes all visual information available to the brain , the decoding performance of a complete retinal population can be rigorously compared to behavioral performance for an equivalent task [15 , 16] . Motion tracking is of major ecological relevance . Amphibians can capture small moving prey , like flies , at a variety of speeds and distances from their body [17–19] , implying that the retina must track such motion accurately . Furthermore , humans can use fixational eye movements to discriminate stimuli separated by roughly two cone photoreceptors , which is highly challenging without an accurate retinal representation of image movement [20] . Yet a reconstruction of the position of an object moving randomly from the activity of sensory neurons has not been achieved in the vertebrate visual system . The classical view suggests that ganglion cells will signal the position of the bar when it is at the peak of their receptive field , making this reconstruction easy by tracking the peak firing rate in the retinal map [21 , 22] . However , the non-linear computations performed by the retinal network seem to make this picture more complex than intially thought: for example , a synchronous peak in the activity could signal sudden changes of speed [23] or a sudden reversal of motion [24] . So it is unclear how a complex trajectory can be decoded from the retinal output , and which neurons are most useful for that purpose . In order to study how neural populations encode dynamic motion , we recorded from a large fraction of the ganglion cells in a patch of the retina while stimulating with a randomly moving dark bar . We show that the entire trajectory of the bar can be estimated by a linear decoder with a precision better than the average spacing between cones using groups of 100+ ganglion cells . Our analysis also revealed an unexpected structure to the population code . Instead of representing the object’s location with a “hill” of neural activity , the population activity was sparse and broadly distributed in space . The retina thus used a distributed and redundant code , in which several disjoint groups could be read-out to reconstruct the stimulus with high precision .
We used a large multi-electrode array with 252 electrodes to record the responses of ganglion cells in the salamander and guinea pig retinas ( see Methods ) , while presenting a randomly moving bar ( Fig 1A , 1B ) . The density of the electrodes allowed us to record a large fraction ( at least 50% ) of the ganglion cells in the retina patch covered by the array [14] . The neurons were recorded over a compact region and had highly overlapping receptive fields . Up to 189 cells were recorded simultaneously ( see example spike rasters in Fig 1C ) over several hours . The stimulus was a dark bar ( width = 100 μm ) on a grey background moving diffusively over the photoreceptor layer ( with a diffusion constant D = 62 . 5mm2 s−1 ) . The trajectory was a random walk with a restoring force to keep the bar close to the array ( Fig 1A , and Methods ) , spanning a region corresponding to roughly 10 degrees of visual angle . One of our primary motivations in choosing this stimulus ensemble was to select patterns of motion that are more complex than constant velocity punctuated by sudden discontinuities . Many previous studies of motion processing in the retina have used objects moving at a constant velocity , sometimes punctuated by sudden changes of velocity [22–25] . While such studies have contributed greatly to our understanding of retinal motion processing , they are not necessarily representative of the patterns of motion that an animal encounters in the natural environment . Obviously , not all objects in the world move at nearly constant velocity for long periods . In addition , we suspected that ganglion cell responses to complex motion might not be simply related to the manner in which they respond to constant velocity motion . Thus , we wanted a more general class of motion patterns . Building on the analogy to white noise stimuli , we selected a class of diffusive motion , which constitutes a broad ensemble of motion patterns without making highly specific choices about the set of trajectories . Under these stimulus conditions , the activity of ganglion cells was sparse and strongly modulated ( Fig 1C and 1D ) . Our goal was to decode the position of the bar at all times from this retinal activity . We used a linear decoder that took as an input the spike trains of all the ganglion cells and predicted the position of the bar as a function of time [26] . A temporal filter is associated with each cell and added to the prediction each time the cell fires an action potential ( Fig 2A ) . The filter shapes were fitted on a training fraction of the data ( 2/3 ) to minimize the squared error between the predicted and true trajectories ( see Methods for details ) . To test the decoder , the prediction was evaluated on the previously withheld test fraction of the data ( 1/3 ) . With 140 cells recorded in a single experiment in the guinea pig retina , the prediction was very precise and followed closely the position of the bar ( Fig 2B ) . The normalized cross-correlation ( see Methods ) between the two traces was CC = 0 . 95 ( corresponding to an error of ∼ 5μm , or 0 . 15 degree of visual angle ) on the testing and the training dataset , which also indicated that the effects of over-fitting were minimal . Similar results were found in the salamander retina ( with CC = 0 . 9 in the testing set , CC = 0 . 93 on the training dataset , Fig 2C ) . They were confirmed in n = 4 retinas for the salamander , and n = 2 for the guinea pig . Similar results were obtained when forcing the filters to be causal , by restricting their parameters to the range of -500 ms to 0 ms before a spike ( Fig 2E; CC = 0 . 9 ) . Thus , the activity of a large population of the ganglion cells contained enough information to reconstruct the position of an object moving randomly across the visual field with high precision . Furthermore , a simple readout mechanism—the linear decoder—was able to perform this reconstruction . How large a population of cells is necessary to decode the position of the bar so precisely ? To address this question , we selected random subsets of cells , performed linear decoding on them , and quantified their decoding performance by measuring the cross-correlation between the real and the predicted trajectories . The decoding performance as a function of the number of selected ganglion cells grew rapidly for few cells and then slowed down for ≳70 cells in both the guinea pig and the salamander retina ( Fig 2D , 2E ) . Therefore , recording a large number of ganglion cells in the same area of the retina was essential to be able to track the trajectory of a moving object with high accuracy . While the high cross-correlation between the real and predicted bar trajectories demonstrates high performance , it is instructive to make this comparison along other dimensions . The decoding performance is expected to depend significantly on the frequency of bar motion for many reasons . Ganglion cells are unlikely to be able to follow rapid motion ( high frequencies ) , because of their temporal integration time; similarly , ganglion cells are unlikely to be able to follow extremely slow motion ( low frequencies ) , because of adaptation . Alternatively , many ganglion cells have highly transient responses that might emphasize moderately high frequencies . Finally , a large neural population might be able to extract information at higher and lower frequencies than individual ganglion cells . The capacity of the retinal circuit to track the position of a moving object is also related to the geometry and signal-to-noise ratio of its sensors . We thus wanted to analyze more precisely the prediction errors and relate them , at least qualitatively , to the properties of the photoreceptor array . In some cases , an upper bound has been derived for the decoding performance that is related to the properties of the photoreceptor array ( spacing between photoreceptors , transfer function of the photoreceptors , etc ) [27] . To quantify the dependence of decoding performance on frequency , we generated a new bar trajectory: a white noise trajectory low-pass filtered at 5 Hz ( see Methods ) . This stimulus ensemble is desirable because a large band of frequencies are represented equally ( Fig 3A; blue ) . This allowed us to explore how the retina represents both slow and fast fluctuations in the bar’s position . We recorded the responses of 158 neurons in the salamander retina to this moving bar , performed the same linear decoding , and estimated the power spectral density of the error ( Fig 3A; see methods ) . The error spectrum was clearly frequency-dependent with a dip between 1 and 4 Hz . Note that the error estimation over this large band illustrates the benefit of the white noise stimulus . If we had used the random walk described in the previous section , we could not have measured the error spectrum for the highest frequencies ( above 3 Hz ) , because this trajectory does not have enough power in this band . Nevertheless , for the lower frequencies ( below 2Hz ) , we found error spectra that were consistent with what was measured with this white noise stimulus . In the guinea pig , where only the random walk stimulus was employed , we also found an error spectrum that was consistent with what we found in the salamander for the same stimulus ensemble . The bar’s motion also spanned a wider field than the array , so we hypothesized that errors would be larger for eccentric positions . We estimated the mean absolute error as a function of the position of the bar itself ( Fig 3B ) . As expected , the error was position-dependent with the minimal error found for the position that was at the middle of the bar’s range of motion . Note that this dependence on position can have more than one explanation . It can come from an incomplete sampling of the side positions by the recorded cells , but also from the fact that , when minimizing the error , more weight is given to the center positions , since they are explored more often . The error spectrum estimated above thus reflected an average between heterogeneous error levels taken at different positions of the bar . We thus aimed at separating simultaneously the effects of position and signal frequency on the prediction error . For that purpose , we designed an error measure that took these two aspects into account . For each frequency , we filtered the error signal ( i . e . the difference between the traejctory and the prediction ) with a band-pass filter centered on that frequency , and then computed the variance separately for each position ( see Methods ) . For proper normalization , we required that a weighted average of this position-frequency spectrum over the positions gives back the power spectrum shown in ( Fig 3A ) . At the middle of the bar’s range of motion , the error spectrum had a soft dip between 1 and 4 Hz , a slightly broader range of frequencies than when averaging over all bar positions ( Fig 3C ) . Over that range of frequencies , the error was significantly below the average spacing between cones in the salamander retina [28 , 29] , illustrated by the dashed line . To our knowledge , this is the first demonstration of high precision in decoding the time-varying trajectory of an object’s motion in the vertebrate visual system . We split our population of ganglion cells into different functional types to understand how they contributed to estimating the moving bar’s trajectory . Here we focus on analyzing single cells . When using the linear decoder , there was no difference between the decoding performance of OFF and ON cells ( p ≥ 0 . 75 , Mann Whitney U-test ) . A similar result was obtained for the guinea pig retina , with no significant difference in decoding performance between OFF and ON cells ( p ≥ 0 . 7 , Mann Whitney U-test ) . Furthermore , when we compared the overall firing rate during random bar motion , we found that ON and OFF cells were similarly responsive ( in the salamander , 1 . 12 ± 0 . 09 Hz ( ON cells ) and 1 . 96 ± 0 . 49 Hz ( OFF cells ) , with no significant difference , ( p ≥ 0 . 37 , Mann Whitney U-test ) ; in the guinea pig , 1 . 15 ± 0 . 14 Hz ( ON cells ) and 1 . 18 ± 0 . 2 Hz ( OFF cells ) , and p ≥ 0 . 9 ) . We went on to sub-divide the salamander ganglion cell population into six functional types according to the temporal dynamics of the receptive field center mechanism , as in previous studies [14 , 30] . Biphasic and monophasic OFF cells were brisk , firing at high rates with spikes clustered in bursts , while the other types ( medium and slow OFF cells ) were sluggish , having slower temporal dynamics , lower peak firing rates and longer refractory periods . However , we did not find any clear difference between the decoding performance for the different cell types ( p ≥ 0 . 05 , Mann Whitney U-test , for all the pairwise comparisons ) , similar to the result for ON and OFF cells . These results illustrate that the functional properties of ganglion cells cannot be solely described by their classical receptive fields . A striking aspect of this sub type analysis was that ON cell contained as much information about the stimulus than OFF cells . If an ON cell is monophasic , and if it integrates the visual input linearly like a LN model does , in principle , it should not respond to the bar leaving the receptive field , and these ON cells should carry less information about the stimulus than OFF cells . This is not what we observed . Our results can be understood as arising from the biphasic temporal response of some ON cells , and/or from the pooling of excitatory and inhibitory non-linear subunits , as found in Y-type cells [31–35] . In this latter case , individual ON-type subunits would have small enough receptive fields that they can be excited by the trailing edge of the dark bar , which causes an increase in the local light intensity when it moves . Thus , ON cells with nonlinear receptive field subunits can also encode local motion with high accuracy . Our naive expectation was that ganglion cells would fire when the bar moved in their receptive field center , such that the population had a peak of activity that travelled with the moving object [25] . In this case , the moving object’s position is represented well by the peak in the neural image . Following a sudden reversal of motion , there is an error in the location of the neural image [24] . However , the spatial location represented by the synchronous burst of firing following motion reversal was shifted in the new direction of motion , so that it begins to “correct” for the retina’s mistaken representation of the object position . As a result , a decoder based on the population vector could achieve good results for an object moving at constant velocity punctuated by sudden reversals [22] . A related example comes from place cells in the hippocampus , where a simple population vector decoder can reconstruct the position of the animal with good resolution [7 , 8] . Thus , we wanted to test if this previously used population vector decoding could be successful at predicting the bar’s trajectory , and see if the retinal activity showed a moving hill that followed the bar’s position . Compared to our previous method of linear decoding , it is interesting to note that population vector decoding corresponds to linear decoding with a temporal window of just a single time bin , and the weight of the filter being related to the value of the receptive field center . We first constructed the “neural image” as in previous studies , where its peak peak corresponded well with the position of the moving bar [24 , 25] . The neural image is the spatial pattern of firing in the ganglion cell population as a function of time . We calculated it by plotting the firing rate of the cells as a function of their receptive field position ( see Methods ) . Despite the success of the neural image in tracking simpler object motion [22 , 25] , we found that it did a poor job with our complex trajectory ( CC = 0 . 06 , Fig 4A ) . Similar results were obtained on the 3 salamander retinas ( average performance CC = 0 . 09 ± 0 . 04 ) . The performance of the neural image decoder was also poor for the guinea pig retina ( CC = 0 . 17 , n = 1 ) . One of the reasons for this poor performance might be that the bar can move to regions where not all of the ganglion cells are recorded by our array . To take this into account , we divided the neural image at each position by the total number of spikes recorded at this position ( see Methods ) . Another reason for the poor performance could be that each cell can respond with a different latency and have different temporal dynamics . So we tried to improve the neural image by taking into account the temporal extent of the receptive field . To this end , we added the temporal profile of the receptive field each time a neuron spiked . Together these two corrections significantly improved the correlation between the neural image and the real trajectory , but the final performance was CC = 0 . 25 for the salamander and CC = 0 . 28 for the guinea pig , far below the performance of the linear decoder . One of the primary reasons for this poor performance was that neural activity was too sparse to continuously represent the moving object’s location ( Fig 4B ) . We quantified the sparseness of a cell as in [36] , by measuring the amount of time where the firing rate was above 5% of its maximum value . On average , we obtained that it was above this threshold 11% of the time ( n = 117 cells ) , which means that neurons remained inactive 89% of the time . The population activity was also sparse: it was above 10% of its maximum only 36% of the time . The neural image performance slightly improved for epochs where the global firing rate was higher , as illustrated by a small negative correlation between the decoding error and the global firing rate ( CC = −0 . 13 ) . However , the performance was still poor during periods of relatively high firing rate: CC = 0 . 27 for the salamander if we restrict ourselves to epochs where the total firing rate was at least half of the max firing rate ) . Thus , there was no clear moving hill of neural activity from which the position of the bar could be inferred . Instead , we observed sparse neural activity with no obvious spatial structure . From this lack of spatial structure we hypothesized that even the ganglion cells whose receptive fields were far from the bar carried information about the trajectory . To test this , we displayed the randomly moving bar stimulus in three different average locations , each separated by 430 microns and lasting 20 minutes ( see Methods and Fig 5A ) . We then estimated the bar’s trajectory for each stimulus ensemble one-at-a-time . The trajectory could be decoded for the three locations at nearly the same level of high performance ( Fig 5B , 5C , 5D ) . We then tried to decode the trajectory with each individual cell , thus re-learning the decoding filter for each of them . We plotted the individual decoding performance of different cells as a function of the distance between the cell’s receptive field center coordinate and the mean position of the bar ( Fig 5E ) . The decoding performance displayed a mild decrease as a function of distance ( on average , a CC loss of 0 . 11 per mm ) . This decrease was partly due to an increasing number of cells falling to CC = 0 . Many cells still carried information about the bar at distances of more than 800 μm away from the bar , while the average receptive field radius is 115μm [30] . We then checked if these results could be explained by some ganglion cells having large receptive fields that would still overlap with the bar trajectory . We defined a normalized distance between the bar and each receptive field as the difference between their average position divided by the receptive field width ( see methods ) . When estimating this distance , the bar width is taken into account by convolving the receptive field with the bar . In these units , the point at which the strength of the surround exceeds that of the center ( the zero crossing point ) is roughly 1 . 5 [37] . However , if we want to know where motion is entirely in the surround , we need to take into account the range of bar positions . If we assume the bar rarely goes further than 3 σ ( where σ = 73μm ) , then this normalized distance is roughly 219 μ m 115 μ m ∼ 1 . 9 . Therefore , normalized positions beyond ∼ 3 . 5 show examples of ganglion cells where motion is entirely in the surround , many of which still show high fidelity motion tracking . These results show that the representation of the moving bar was distributed across a large population of cells spread widely in spatial location , far beyond what the extent of the receptive field center would predict . This also explains why the neural image failed to code for the bar position . We can gain additional insight into why ganglion cells can precisely encode motion on their surround by inspecting the corresponding neural responses . When the bar moved over the receptive field center , ganglion cells fired in discrete firing events on a background of silence ( Fig 6A , 6B ) . Similar spike train structure has been seen under many other stimulus conditions [14 , 36 , 38] . Motion in the surround also triggered responses with similar event-like structure ( Fig 6C , 6D , 6E , 6F ) . Another way to explore the nature of the precise coding of an object’s trajectory is to compute the spike-triggered speed average , which is analogous to the standard spike-triggered stimulus average [39] . This reveals that the speed was higher than average for a period ∼ 200 ms before a spike ( depending on the cell ) and slightly below average for a period of almost a second before a spike ( Fig 6G , 6H ) . Thus , ganglion cells tended to spike after a brief acceleration . Importantly , this spike-triggered speed average was nearly invariant to the overall location of the moving bar ( Fig 6G , 6H; red , green , blue ) . Together , these results support the interpretation that the structure of the code for a moving object’s location was substantially similar for motion in both the center and surround , with sudden accelerations triggering firing events in ganglion cells . Yet at the same time , ganglion cells could encode the position of the moving bar quite well . Even though the spike-triggered speed average was similar for the 3 positions , the decoding filters were very different ( Fig 6I , 6J ) . The positive ranges on the y-axis correspond to the bar getting closer to the receptive field . So for the green and red curves , the cell tends to signal when the bar gets closer to its receptive field . In the far surround case , this might correspond to excursions of the bar into the closer surround . Thus , we found that ganglion cells could encode different features of the trajectory depending whether the bar is close or far from its receptive field . Because of the distributed nature of the code , we wondered how much redundancy was present in the ganglion cell population . As seen above ( Fig 2 ) , after a rapid initial increase , the decoding performance essentially saturated as a function of the number of cells . This saturation suggests that the information encoded by a new cell was highly redundant with the rest of the population . To quantify the redundancy in a principled way , we estimated the information rate between the real and predicted trajectories for single cells and for groups of cells of different sizes . Note that this measure is a lower bound on the true mutual information between the spike trains of a group of cells and the stimulus [26 , 40] . Information about the moving bar’s trajectory encoded by subsets of ganglion cells increased as more cells were added ( Fig 7A ) . But for the same number of cells , the performance varied substantially depending on the particular cells chosen in the subset . A natural hypothesis for this variability is that some cells carry more information about the stimulus than others . To explore this relationship , we compared the “total” information encoded by a group of ganglion cells ( y-axis in Fig 7A and 7B ) versus the sum of the individual informations encoded by each cell in the group ( x-axis in Fig 7B ) . Plotted in these units , there was much less variability in the information encoded by different groups of ganglion cells . The same result was obtained in the guinea pig retina ( Fig 7C , 7D ) . In making this comparison , we picked subsets randomly . To better test the predictive power of the sum of individual informations , we looked for the subsets with the best and the worst decoding performances for the same number of cells . There were too many combinations of cells having a given group size to allow for a comprehensive search over all possible subsets . Instead , we ordered the cells by their individual informations , and for each number of cells N , we estimated the total information for the N “best” cells and the N “worst” cells . While there was a large difference in the performance for the best and the worst subsets when plotted against the number of cells ( Fig 7E ) , this difference was almost entirely compensated when plotting them against the sum of individual information ( Fig 7F ) . For a group of ganglion cells having a similar sum of individual informations ( i . e . a difference less than 0 . 5 bits/s ) , the average difference between worst and best subsets was 8% of the best . Part of the remaining difference is likely due to the error in the estimation of the individual informations ( horizontal error bars ) due to the finite size of the data . In contrast , the “worst” cell groups had on average 55% less information than “best” groups when matched for number of cells . These results together show that we could predict the decoding performance of a group of ganglion cells from the properties of individual cells . In both the guinea pig and the salamander , small groups ( less than 10 cells ) had a total information that was close to the sum of individual informations , so the different cells carried nearly independent information . But as the sum of individual informations increased , it became much larger than the total information encoded by the group . Thus , the redundancy among ganglion cells ( defined as one minus total information divided by sum of individual informations ) increased with the number of cells and became quite large as the performance saturated , with a 6 . 4-fold redundancy for the 123 cells in the salamander , and 6 . 6-fold for 140 cells in the guinea pig . This suggests that multiple subsets of cells might be able to encode the bar trajectory . To find these disjoint subsets , we used L1 regularized decoding , a method that simultaneously minimizes the error in the prediction while setting to zero as many filter coefficients as possible ( see Methods ) . Compared to the previous optimization where there were no constraints on the filter amplitude , here most filters were driven to zero ( Fig 8A ) , but the prediction performance remained unchanged ( CC = 0 . 93 ) . The filters with the highest amplitude corresponded to the neurons that were the most useful at decoding the trajectory . We ordered them from the highest to the lowest filter amplitude , and split them into groups of 10 that we used to decode the bar position . The decoder was then re-trained without L1-regularization for each group of 10 . As expected , the cells with higher filter amplitude were better at decoding the trajectory than those with lower amplitude ( Fig 8B ) , which confirmed that the L1 method picked the most informative cells . Of particular interest , we also found that the performance remained above CC = 0 . 75 for the first seven groups , indicating many different groups of 10 neurons could decode the trajectory with high accuracy . We then looked for disjoint subsets that would give a prediction performance above CC = 0 . 85 , which was 90% of the performance for the entire population . We first tried to decode using only the cell with the highest filter amplitude , and then kept adding cells to the subset , ordered from the highest to the lowest amplitude , until the decoding performance using the subset reached 0 . 85 . We then restarted the same process after discarding the cells already used in the subset . By iterating this selection , we could find 6 disjoint subsets of increasing size that were all able to decode the trajectory of the bar ( Fig 8C ) . All these subsets had a decoding performance between 0 . 85 and 0 . 87 . These analyses together demonstrate that we could define 6 or 7 subsets with high decoding performance , similar to the estimated redundancy of information in the population . So far we have estimated redundancy for random subsets , including ganglion cells from all the different types . Perhaps , the redundancy is significantly different for cells of the same functional type . We could define one cell type in the salamander retina , the biphasic OFF type cell ( Fig 9B ) , whose receptive fields clearly tiled visual space ( Fig 9A ) , as expected from previous studies [14 , 30] . Using the same method as before , we took many subsets of cells belonging to the same cell type and plotted the sum of the individual information against the information carried by the subset . The values obtained in this redundancy plot were similar to the ones obtained with random subsets ( Fig 9C ) . When averaging over subsets having the same sum of individual informations , we found that the information encoded by small subsets was essentially identical for ganglion cells of the same functional type as for random functional types . However , for large enough subsets , the information encoded by cells of the same functional type was significantly smaller than for random subsets ( Fig 9D ) .
Linear decoding was first developed and applied to the H1 interneuron in the fly to estimate wide-field motion [10] . Linear decoding has also been applied to small populations of retinal ganglion cells to estimate the light intensity of a spatially uniform , white noise stimulus [11 , 41] , and to reconstruct a checkerboard stimulus for large populations of primate parasol cells [12] . However , in these previous studies , the LN model worked well for the stimuli employed . Some of the most notable quantitative successes of the LN model have been demonstrated with spatially uniform stimulation [42–44] . Pillow et al ( 2008 ) showed that the LN model performed very well for parasol cells under their stimulus ensemble , perhaps due to the fact that the checkers were large enough to cover a cell’s entire receptive field center . This is not the case for the kind of diffusive motion we used here , which includes pauses , starts , and reversals that trigger transient bursts of firing that cannot be explained by the LN model [24 , 35] . One important factor in the failure of the LN model is the presence of nonlinear spatial subunits inside ganglion cell receptive field [34 , 35 , 39 , 45 , 46] . During spatially uniform stimulation , all nonlinear spatial subunits are activated together , allowing models that do not include this structure to approximate well the ganglion cell’s firing rate . But during irregular motion , the subunits can be independently activated , such that a model without this structure performs poorly [35 , 47] . Other studies have tried to estimate a visual stimulus from the activity of populations of ganglion cells . The speed of a moving bar could be estimated from the activity of parasol ganglion cells [1 , 2] . However , this study was restricted to the case of motion at constant speed , so only one number was estimated: the speed of the moving bar . For the case of a moving texture that switched speeds every 500 ms , discrete classification of the speed was possible [23] ( see also [48] ) . Similar classification studies have been performed in the visual cortex [4 , 49] . Our study differs from these by reconstructing the full time-varying position of the bar , a task that is much more difficult than the classification of several stimuli or the estimation of a single parameter . One of the most significant surprises in this study was the lack of precise spatial structure of the retina’s population code for this diffusive motion stimulus . One key factor was the complexity of our motion stimulus . Most early studies of how ganglion cells respond to motion involved an object moving at constant velocity [22] , a case in which the picture of the neural image seems to hold . In our stimulus ensemble , the object moved at varied speed , including discontinuities of motion and moments of high acceleration . Motion discontinuities like the sudden onset or reversal of motion trigger transient bursts of firing [24 , 35] , as does acceleration in general [23] . Finally , while our diffusive motion is more complex than the stimuli aforementioned , it is still far from the complexity of a natural movie . It would be interesting for future studies to investigate if the responses to natural movies show a precise spatial structure or not . Another striking result was how well ganglion cells encoded diffusive motion of a dark bar in their surrounds , which partly explained the diffuse spatial structure of the code . Motion in the surround has long been known to be able to generate excitatory responses in ganglion cells [50] . Our analysis demonstrates that this activation is not just a global alert signal , but can be used to encode the precise trajectory of the bar . Detailed investigation has revealed that the response of ganglion cells to gratings drifting in their surround can be modeled as the pooling of excitatory and inhibitory non-linear subunits [51 , 52] . In the salamander , the surround response to annulus of modulated luminance can also be modeled by pooling non-linear subunits [53] . Many other studies have demonstrated the presence of disinhibitory circuits within the receptive field center , involving transient amacrine cells turning off sustained inhibition from sustained amacrine cells [54–57] . These circuits may allow precise information about motion in the surround to be conveyed by the firing of ganglion cells . Finally , we found that the decoding performance of ON and OFF-type ganglion cells was comparable . This is surprising if one thinks about how the moving bar interacts with the classical receptive field . In this view , the dark bar should cause excitation of OFF cells and inhibition of ON cells . Thus , ON cells should fire at lower rates and convey less visual information . Neither of these properties was observed in our data . However , we can understand these results if we instead think about ganglion cells as having nonlinear subunits in their receptive fields . Such subunits have smaller receptive fields than the entire ganglion cell , and therefore can be excited by the trailing edge of the moving dark bar , which is effectively an ON-type stimulus . Many ganglion cell types have been shown to possess spatial subunits , including brisk cells like the alpha cell in the guinea pig [32] and in the mouse [34] and the fast OFF cell in the salamander [33] as well as sluggish cells like the local edge detector [56 , 58] . One might also think that our results apply primarily to retinal ganglion cells with transient response types and that sustained ganglion cells might be able to continuously follow a moving object . However , our recorded populations included sustained cells . More specifically , the fast OFF cells in the salamander , which were the focus of the original work on motion anticipation [25] as well as a recent study using a neural image decoder [22] have a sustained response to smooth motion . Yet these same cells exhibit sporadic , event-like firing for diffusive motion . Thus , the stimulus ensemble changes the qualitative nature of ganglion cell spike trains , and studies of how sustained cells respond to constantly drifting gratings do not predict how those cells would respond to a diffusive motion . And in fact , our recordings in the guinea pig also showed sparse , event-like firing across the ganglion cell population ( Fig 1C ) in contrast to the intuition derived from many studies of mammalian retina using drifting gratings . But important caveats to our findings are needed: our stimulus ensemble—a diffusively moving bar—is artificial and only explores a subset of all possible motion patterns . It is conceivable that more naturalistic stimuli could strongly activate dedicated subpopulations of ganglion cells . Furthermore , we could only define broad functional groups rather than “true” cell types with distinct anatomy . Perhaps true anatomical cell types would have more clearly differentiated function . Clearly , the distributed and redundant neural code that we have observed is not how a human engineer would design a system to track motion . In most theoretical studies where a population of neurons has to code for the value of a parameter , the single cell selectivity is modeled by a tuning curve . As a result , the activity in response to the stimulus is local , and neurons code best for stimuli at the peak of their sensitivity , or at the side of their tuning curves , depending on the noise level [59] . Experimentally , this corresponds to the case of the neural image . While the peak of neural activity does track objects moving at constant velocity [22 , 25] , we have shown that , in our case of diffusive motion , the retinal code was too sparse and spatially diffuse for the peak of neural activity to track the position of the bar . This organization challenges the classical theory that the position of an object could be simply tracked by following the peak of firing rate in the retinal output . Even though the retinal code did not have a straightforward spatial organization , it could still be read out with one of the simplest of all decoding mechanisms: the linear decoder . One of the implications of the success of the linear decoder is that it makes information easy to extract by subsequent neural circuits [60] . More generally , a fundamental function of sensory circuits might be to compute and actively maintain an “explicit” or linearly decodable representation of the most relevant features of the environment [61] . This appears to be the case for IT cortex: the identity of an object can be decoded linearly from IT neurons , but not from V1 neurons [62] . Our results thus show that in this sense , the retina has an “explicit” representation of the position of a moving object . The high redundancy of the retinal code was both notable and somewhat surprising . This observation is consistent with a previous study finding extensive redundancy during stimulation with natural movie clips [38] , so the high redundancy we observed is not only due to the structure of the stimulus we employed here . Redundancy results from the high spatial overlap of ganglion cell’s receptive fields: in the salamander , the number of receptive field centers covering a point in visual space is roughly 60 [63] . It also results from the distributed nature of the code , namely the fact that cells of different functional type do not convey categorically different visual information . Why might such a distributed and redundant code be a beneficial organization ? One advantage is that neural circuits in the central brain would have access to high precision motion tracking information from several groups of ganglion cells . This is useful because there are multiple features in the visual scene beyond just the location of a moving object that the brain might want to follow . For instance , if such circuits sampled all ∼ 100 ganglion cells with receptive fields overlapping one point in visual space , they could extract high resolution spatial and chromatic information about object identity in addition to the object’s trajectory . Alternatively , if ∼ 100 ganglion cells with distributed spatial locations were sampled , then information about the characteristics of a larger object could be extracted in addition to its motion trajectory . Thus , the distributed , redundant population code maintains maximum flexibility with respect to the purposes of downstream neural circuits . Of course , our analysis does not establish whether the brain uses a linear decoder nor how it might learn effective decoding kernels . In particular , we have shown that the decoders that reached the best performance had to be learned on the moving bar data directly . This means that the detailed parameters of the decoder that achieves high performance might depend on the properties of the stimulus ensemble . By demonstrating the broad spatial selectivity of ganglion cells , our study shows that neurons with complex tuning curves or mixed selectivity [64] do not appear only when merging information from different modalities in associative areas of the brain , but already at the earliest stage of sensory processing . Interestingly , these features of the retinal code were only apparent when the stimulus dynamics were sufficiently rich . The redundancy of trajectory representation that we uncovered could be a trade-off between discrimination and generalisation [65] . Future studies will have to understand how a representation of motion that is invariant to the context can be extracted from the retinal activity .
This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee ( IACUC ) of Princeton University ( Protocol Number: 1828 ) . Retinal tissue was obtained from larval tiger salamanders ( Ambystoma tigrinum ) of either sex and continuously perfused with oxygenated Ringer’s medium at room temperature . For guinea pig experiments , the tissue was perfused with AMES solution and maintained at 37 °C . Ganglion cell spikes were recorded extracellularly from a multi-electrode array with 252 electrodes spaced 30 μm apart ( custom fabrication by Innovative Micro Technologies , Santa Barbara , CA ) . Details of the recording and spike sorting methods are described elsewhere ( [14 , 63] . Experiments were performed in accordance with institutional animal care standards . The main dataset presented in this paper is available for download on the Dryad repository ( doi: 10 . 5061/dryad . 1dp55 ) . The stimuli were displayed on a CRT screen with a 60 Hz refresh rate [38] and a background light of 12 mW/m2 . We focused the stimulation plane precisely on the photoreceptor plane during each experiment . The stimulus presented was a dark bar of 100% contrast moving randomly over a gray background . The length of the bar was 2300 μm and the width 100 μm . The trajectory was given by Brownian motion with a spring-like force to bring the bar back to the center: d v d t = - v τ + σ Γ ( t ) - ω 0 2 x ( 1 ) where x is the position of the bar , v = d x d t is the velocity of the bar , σ2 = 0 . 05m2 s−3 , Γ ( t ) is a Gaussian white noise , ω0 = 9 . 42 Hz and τ = 50 ms . For the experiments where we tested the decoding performance as a function of frequency , the trajectory was sampled every stimulus frame ( 16 . 7 ms ) from a Gaussian distribution of standard deviation 60 μm and low-pass filtered at a frequency of 5 Hz using a Butterworth filter . The cut-off was chosen to be at 5 Hz because salamander cones do not seem to convey much information beyond this frequency [11] . Spatio-temporal receptive fields were measured by reverse correlation to a flickering checkerboard composed of squares of 69 μm size that were randomly selected to be black or white at a rate of 30 Hz [30] . Since salamander eyes are ∼ 4 mm in diameter , one degree of visual angle should correspond to approximately 35 microns in the retina plane , assuming a focal length of 2 mm . So the size of the array corresponds to 12–13 degree of visual angle . We used a linear model that takes the spike trains as an input and gives a prediction about the position of the bar as an output: p ( t ) = ∑ i , j K i ( t - t i j ) + C ( 2 ) where tij is the j-th spike of the neuron i , and p ( t ) is the predicted position of the bar over time . The filters Ki and the constant C are found by minimization of ⟨ ( p ( t ) − x ( t ) ) 2⟩ ( least square minimization ) [11] , where x ( t ) is the real position of the bar , and angular brackets denote averages over time . The filters extended 500 ms before and after the spike . All the results shown here were cross-validated: we trained our model on 2/3 of the data , and tested it on the other 1/3 . We used one hour of recordings , which corresponds to 40 minutes for the training set , and 20 minutes for the test set . If the performance on the testing was below 0 ( e . g . for decoding using a single cell ) , we set it to 0 , since a negative performance is an effect of overfitting . To characterize decoding performance , we estimated the normalized cross-correlation ( CC ) between the prediction and the real trajectory . For two signals x ( t ) and y ( t ) , it is classically defined as C C = ⟨ ( x ( t ) − m x ) . ( y ( t ) − m y ) ⟩ σ x . σ y , where mx and σx are the mean and standard deviation of x ( t ) , respectively , and ⟨⟩ denotes an average over time . To get an estimation of the error signal as a function of both the real position and the frequency , we filtered the error e ( t ) = p ( t ) − x ( t ) by a bandpass filter that did not introduce a phase delay , i . e . a Morlet Wavelet . The wavelet was normalized so that the averaged squared value of the filtered signal matched the corresponding value in the total power spectrum . For each possible position , we picked the times where the bar was at this position , and took the average of the corresponding squared values in the filtered signal . Formally , if ef ( t ) is the filtered version of e ( t ) for frequency f , then the error e ( f , p ) for position p and frequency f was computed as: 1 N p ∑ i = 1 N p e f ( t i ) 2 ( 3 ) where the ti’s are all the times such that x ( ti ) = p , and Np is the number of such points . Note that the error spectra measured here are in microns2 Hz−1 . This means that , if the trajectory was a sinusoid at a given frequency , the variance of the decoding error should be the value estimated in this error spectrum . The dashed line representing cone spacing in Fig 3 corresponds to the square of the average distance between cones in the salamander ( 20μm ) . If the error spectrum value at frequency f is below this dashed line , it means that a sinusoidal trajectory at this frequency should be decoded with a root-mean-squared error lower than 20 μm . For more complex stimuli , the error should be integrated over the entire frequency band of the stimulus before being compared to this line . To construct the neural image , we assigned a spatial position to each cell as the peak of its receptive field . The activity of the cells was binned in 16 . 6 ms bins ( corresponding to the refresh rate of the stimulus ) . At each time bin , we counted the number of spikes emitted by the cells at each spatial location . The resulting matrix was then smoothed across spatial positions with a Gaussian smoothing kernel of width 21 μm . The “most likely” position was obtained by taking the peak location of the neural image at each time where there was at least one cell firing . We also tried to decode the position by estimating an average of the position weighted by the firing rate , but the performance was even worse ( CC = 0 . 027 ) . We then improved this neural image analysis in two ways . First , to take into account the fact that there might be “holes” in the coverage , we normalized the neural image with the following method . We assumed that , if we were to record all the retinal ganglion cells , the total firing rate should be the same for each spatial position . If this is not the case , the neural image will be biased to favor some positions just because cells spike more often there . To remove this bias , we divided the neural image at each position by the total number of spikes recorded at this position . Mathematically , it means that if we call the neural image A ( x , t ) , where x is the spatial position and t the time , we divided A ( x , t ) by ∑t A ( x , t ) Second , to take into account the fact that each cell will respond with a specific latency and time course , we tried to improve the neural image by taking into account the temporal extent of the receptive field . To this end , each time a neuron at position x spiked at time t , instead of simply incrementing the neural image at ( x , t ) , we added the temporal profile of the receptive field to the window ( x , [t − τ;t] ) . The information rate between the true and decoded bar trajectory was estimated from their mutual coherence , γ ( f ) , as I = − ∫ 0 f max d f log 2 ( 1 − γ 2 ( f ) ) [40] . Debiased coherence was computed using Chronux [66] using trajectory windows of duration 256/60s sampled at 60 Hz . The frequency range of integration [0 , fmax] was determined as the contiguous range where the estimated coherence is above zero at 10−2 significance level . Alternatively we considered a larger frequency range by lowering the significance threshold to 10−1 , and performing explicit debiasing by estimating the information on multiple subsets of the data of various sizes , and extrapolating to infinite sample size [67] . The results agreed within error bars . To validate our method , we generated synthetic “real” and “decoded” traces with Gaussian statistics and power spectra that were similar to the real traces , and of similar length . In this case the true information is analytically computable . We compared this exact result to our estimators to assess their accuracy and verify that the chosen parameters ( window length , determination of frequency range ) were suitable . For information rates above 0 . 5 bit/s the estimated error in the information rate is between 5–10% , while for rates of ∼0 . 1 bit/s the error increases to about 20% . The relation between the normalized cross-correlation and the mutual information estimated this way was examined . If there were no temporal correlation , the mutual information would be I = −0 . 5 log ( 1 − CC2 ) . The measured information was significantly different from the value predicted by this formula . This is unsurprising since the formula does not take into account the different frequencies . However , we could fit the information values with a linear interpolation of this formula , i . e . I = −0 . 5α log ( 1 − CC2 ) + β . With this formula , we could explain 98% of the variability in the information values . So an empirical relationship could be inferred between the two quantities . Note that α was larger than one , indicating a synergy between frequencies . Redundancy in a subset of cells A was defined as R = 1 − M I A ∑ i ∈ A M i , where MIA is the mutual information for the subset , and Mi the mutual information for each cell . In L1-regularized decoding the least squares minimization problem ⟨ ( p ( t ) − x ( t ) ) 2⟩ is substituted by ⟨ ( p ( t ) − x ( t ) ) 2⟩ + λ∑i‖Ki‖1 where λ ≥ 0 is the regularization parameter . Unlike the simple least squares case , this regularized minimization cannot be reduced to a simple straightforward linear algebra problem and a numerical solution is required . For this purpose we have used the implementation of the interior-point method by Boyd et al [68] . The value of the regularization parameter λ is usually chosen to minimize the mean-squared error ( MSE ) . However , our goal is to assign zero weight filters to as many cells as possible while keeping a good decoding performance . Therefore , we choose a larger-than-optimal regularization parameter by allowing a 5% decrease in performance ( CC ) and a 20% increase of the MSE . Also , we run a 5-fold cross-validation procedure on the training set to validate the choice of regularization parameter . For each cell i , we estimated its receptive field using a classical checkerboard stimulus ( see above ) and convolved it with the bar . From the resulting spatial distribution , we estimated the average position mi and the receptive field width si as the standard deviation of this distribution . The normalized distance was then defined as d i = ∣ m b − m i ∣ s i . Note that the receptive fields were measured with the same stimulus display . If there was a blur in the optical system that we used to display the stimulus , it would equally affect the receptive field estimation . So our distance measure should not be affected by a blur in the optical system .
|
It remains unclear how the brain is able to track the location of moving objects by reading the spike trains received from the retina . To address this question , we recorded a large population of ganglion cells , the retinal output , in a dense patch of salamander and guinea pig retinas while displaying a bar moving in complex motion . From previous studies , the naive expectation was that individual ganglion cells would spike when an object was moving on their receptive field center and that the entire population’s activity would resemble a “hill” that continuously tracked the object’s location . However , our analysis revealed that this picture did not hold . Instead , ganglion cells fired sparsely and coded for the bar trajectory even when it was far from their receptive field center . Nevertheless , we showed that the bar’s position could be reconstructed from retinal activity with an accuracy better than the spacing between photoreceptors , when using more than 100 cells . We also showed that the retinal code was highly redundant , over-representing the same information more than 6-fold . Yet , this unexpected representation allowed for precise object tracking using a simple decoder , as long as the temporal structure of the spike trains was accounted for .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
High Accuracy Decoding of Dynamical Motion from a Large Retinal Population
|
Cytotoxic T-lymphocytes play an important role in the protection against viral infections , which they detect through the recognition of virus-derived peptides , presented in the context of MHC class I molecules at the surface of the infected cell . The transporter associated with antigen processing ( TAP ) plays an essential role in MHC class I–restricted antigen presentation , as TAP imports peptides into the ER , where peptide loading of MHC class I molecules takes place . In this study , the UL49 . 5 proteins of the varicelloviruses bovine herpesvirus 1 ( BHV-1 ) , pseudorabies virus ( PRV ) , and equine herpesvirus 1 and 4 ( EHV-1 and EHV-4 ) are characterized as members of a novel class of viral immune evasion proteins . These UL49 . 5 proteins interfere with MHC class I antigen presentation by blocking the supply of antigenic peptides through inhibition of TAP . BHV-1 , PRV , and EHV-1 recombinant viruses lacking UL49 . 5 no longer interfere with peptide transport . Combined with the observation that the individually expressed UL49 . 5 proteins block TAP as well , these data indicate that UL49 . 5 is the viral factor that is both necessary and sufficient to abolish TAP function during productive infection by these viruses . The mechanisms through which the UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 block TAP exhibit surprising diversity . BHV-1 UL49 . 5 targets TAP for proteasomal degradation , whereas EHV-1 and EHV-4 UL49 . 5 interfere with the binding of ATP to TAP . In contrast , TAP stability and ATP recruitment are not affected by PRV UL49 . 5 , although it has the capacity to arrest the peptide transporter in a translocation-incompetent state , a property shared with the BHV-1 and EHV-1 UL49 . 5 . Taken together , these results classify the UL49 . 5 gene products of BHV-1 , PRV , EHV-1 , and EHV-4 as members of a novel family of viral immune evasion proteins , inhibiting TAP through a variety of mechanisms .
Evolving under the selective pressure of the host immune system , herpesviruses have developed countermeasures to prevent recognition of infected cells by cytotoxic CD8+ T lymphocytes ( CTLs ) . CTLs recognize viral antigens presented as peptides bound to major histocompatibility complex ( MHC ) class I molecules at the surface of infected cells . Herpesviruses in particular have acquired diverse mechanisms to inhibit antigen presentation in the context of MHC class I molecules , thereby escaping from elimination by CTLs [1]–[4] . Most peptides presented by MHC class I molecules are transported into the endoplasmic reticulum ( ER ) lumen by the transporter associated with antigen processing , TAP . TAP is a heterodimer composed of TAP1 and TAP2 subunits and belongs to the ATP-binding cassette family of transporters [5] , [6] . TAP translocates peptides across the ER membrane via a conformational transition that is energized by the hydrolysis of ATP . TAP is part of the MHC class I peptide-loading complex that also contains tapasin , MHC class I heavy and light chains , and several auxiliary proteins including calreticulin and ERp57 [5] , [7]–[10] . Several herpesviruses have acquired mechanisms to interfere with TAP function . Interestingly , inhibition of TAP transport is achieved through different strategies , exerted by unique gene products . Although the varicellovirus bovine herpesvirus 1 ( BHV-1 ) and the simplexviruses herpes simplex virus type 1 and 2 ( HSV-1 and -2 ) all belong to the subfamily of alphaherpesviruses , they block TAP through proteins that have an entirely different structure and mode of action . The inhibition of TAP by BHV-1 relies on the UL49 . 5 ( Unique Long 49 . 5 ) gene product , a type I transmembrane protein of 75 amino acids [11] . Inactivation of TAP by UL49 . 5 involves two events: the arrest of the peptide transporter in a translocation-incompetent state and the proteasomal degradation of both subunits of TAP [11] . In contrast , the ICP47 proteins of HSV-1 and -2 are soluble cytosolic proteins acting as high-affinity competitors for peptide binding to TAP [12]–[18] . Within the subfamily of betaherpesviruses , human cytomegalovirus ( HCMV ) was found to encode a protein , US6 , that inhibits TAP function by reducing the interaction of ATP with TAP [19]–[24] . The murine gammaherpesvirus-68 ( MHV-68 ) encodes the mK3 protein that acts as a ubiquitin ligase linking MHC class I molecules and TAP to the ubiquitin/proteasome degradation pathway [25]–[35] . Recently , the BNLF2a protein of Epstein-Barr virus ( EBV ) and of related primate gamma-1 herpesviruses has been characterized as a potent TAP inhibitor , preventing the binding of both peptides and ATP to TAP [36] . Homologs of UL49 . 5 ( commonly known as glycoprotein N; gN ) are encoded by every alpha- , beta- and gammaherpesvirus sequenced to date [37]–[39] . The UL49 . 5 genes are all predicted to encode a type I membrane protein with a putative cleavable signal sequence . The UL49 . 5 proteins interact with another herpesvirus protein , glycoprotein M ( gM ) , with which they form a disulfide-linked heterodimer through a conserved cysteine residue within their ER-luminal/extracellular domain [37] , [40]–[44] . Nevertheless , the amino acid sequences of UL49 . 5 proteins demonstrate considerable heterogeneity , even among varicellovirus UL49 . 5 proteins ( Fig . 1 ) . The only exceptions are EHV-1 and EHV-4 UL49 . 5 , which differ by only seven amino acid residues . Thus , at this moment , it is impossible to predict on the basis of amino acid sequence whether any of these proteins have the same capacity to inhibit TAP that was found for BHV-1 UL49 . 5 . The UL49 . 5 gene products of HSV-1 , HSV-2 , HCMV , and EBV fail to block TAP , indicating that not all UL49 . 5 molecules act as inhibitors of TAP [11] , [37] . In this study , the effects on TAP function were assessed in more detail for UL49 . 5 encoded by various members of the genus Varicellovirus . The UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 were found to down-regulate MHC class I cell surface expression through TAP inhibition . Their ability to block TAP was observed in cells of the relevant host species , as well as in human cells . Using UL49 . 5 deletion mutants of BHV-1 , PRV and EHV-1 , it was shown that the UL49 . 5 proteins of these viruses are responsible for the inhibition of TAP-dependent peptide transport . The UL49 . 5 homologs of canine herpesvirus ( CHV ) and VZV did not affect MHC class I surface expression . BHV-1 UL49 . 5 strongly reduces the steady state protein levels of TAP in both bovine and human cells , whereas the UL49 . 5 proteins of EHV-1 , EHV-4 or PRV were not observed to have this capacity . Interestingly , the EHV-1 and EHV-4 UL49 . 5 homologs interfere with the binding of ATP to TAP , a function that is not influenced by BHV-1 or PRV UL49 . 5 . The UL49 . 5 proteins of PRV and EHV-1 arrest TAP in a translocation-incompetent state , a property that is shared with BHV-1 UL49 . 5 . Thus , the BHV-1 , PRV , EHV-1 and EHV-4-encoded UL49 . 5 proteins all induce a similar phenotype , i . e . inhibition of peptide transport , but their modes of action demonstrate a surprising diversity .
To evaluate the TAP-inhibiting capacity of the UL49 . 5 proteins encoded by the varicelloviruses PRV , EHV-1 , EHV-4 , CHV and VZV , cell lines of the relevant host species were transduced using a retrovirus-based gene delivery system to express the corresponding UL49 . 5 proteins . Down-regulation of MHC class I expression by the UL49 . 5 gene products was evaluated using flow cytometry . In cells expressing UL49 . 5 of BHV-1 , PRV , EHV-1 and EHV-4 , MHC class I surface expression was reduced ( Fig . 2A ) . The UL49 . 5 proteins of CHV and VZV failed to down-regulate MHC class I surface expression . These results indicate that UL49 . 5 of BHV-1 , PRV , EHV-1 and EHV-4 interfere with MHC class I-restricted antigen presentation . To investigate whether the observed down-regulation of MHC class I cell surface expression relies on the inhibition of TAP , species-specific cell lines stably expressing these UL49 . 5 homologs were evaluated for TAP-dependent peptide transport . The UL49 . 5 proteins of BHV-1 , PRV , EHV-1 and EHV-4 strongly inhibited TAP activity in the corresponding natural host cell lines ( Fig . 2B ) . Despite the absence of a detectable reduction in cell surface MHC class I levels ( Fig . 2A ) , some inhibition of TAP-dependent peptide transport was observed in canine cells expressing the CHV UL49 . 5 protein ( Fig . 2B ) . Apparently , the inhibition of TAP by CHV UL49 . 5 was insufficient to observe MHC class I downregulation at the cell surface . VZV UL49 . 5 had no significant effect on TAP activity . Thus , although the amino acid sequences of the UL49 . 5 proteins of BHV-1 , PRV , and EHV-1/EHV-4 demonstrate considerable variation ( Fig . 1 ) , their ability to inhibit TAP was found to be a common property of these varicellovirus gene products . VZV infection has been shown to cause down-regulation of MHC class I expression at the cell surface [45]–[47] . This phenotype could not be reproduced by the VZV-encoded UL49 . 5 protein when expressed individually ( Fig . 2A ) . To examine whether the absence of MHC class I down-regulation by VZV UL49 . 5 is due to a loss of the interaction of the viral protein with the TAP complex , TAP was immunoprecipitated from VZV UL49 . 5-expressing MJS cells that were solubilized in the presence of the mild detergent digitonin . The resulting protein complexes were separated by SDS PAGE and analyzed for the presence of UL49 . 5 by immunoblotting . Surprisingly , VZV UL49 . 5 was found to interact with the TAP complex ( Fig . 3A ) . Although VZV UL49 . 5 associates with TAP , this appears to be insufficient to inhibit peptide transport effectively ( Fig . 2A and B ) . VZV UL49 . 5 could , however , interfere with peptide-loading and MHC class I-restricted antigen presentation in a different way . For instance , the US3 protein encoded by human cytomegalovirus binds both tapasin and TAP , without having an effect on TAP function . Instead , US3 impairs tapasin-dependent peptide loading and optimization of the MHC class I peptide cargo [48] , [49] . To investigate whether VZV UL49 . 5 inhibits MHC class I-mediated antigen presentation via a mechanism similar to that of US3 , functional T cell assays were performed using a panel of human leukocyte antigen ( HLA ) -A1 and HLA-A2-restricted CTL clones . It is known that especially HLA-A1-restricted peptide presentation strongly depends on the function of tapasin . Antigen-presenting phytohemagglutinin ( PHA ) -treated T cell blasts and the melanoma cell-line Mel518 were transduced to express the UL49 . 5 proteins of VZV and BHV-1 . While the presence of BHV-1 UL49 . 5 greatly reduced specific lysis of the PHA-blasts by CTLs , the expression of VZV UL49 . 5 had no detectable effect ( Fig . 3B ) . This was observed for HLA-A1 and HLA-A2-restricted CTL clones . VZV UL49 . 5-expressing and control target cells induced IFNγ production by the CTLs , while reduced IFNγ production was observed when BHV-1 UL49 . 5 was expressed by the target cells ( Fig . 3C ) . This reflects effective inhibition of CTL recognition by BHV-1 but not VZV UL49 . 5 . Thus , despite the interaction of the VZV UL49 . 5 protein with the peptide-loading complex , no interference with MHC class I-restricted antigen presentation could be detected . Having observed that the UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 interfere with MHC class I-restricted antigen presentation when expressed individually , we next investigated whether the various UL49 . 5 proteins are responsible for TAP inhibition during infection with BHV-1 , PRV , or EHV-1 . Peptide transport activity was examined in natural host cells infected with wild type viruses or with the corresponding recombinant viruses lacking a functional UL49 . 5 gene [41] , [43] . Whereas the wild-type viruses effectively blocked peptide transport , this inhibition was not observed in cells infected with the mutant viruses lacking UL49 . 5 ( Fig . 4 ) . These findings indicate that during infection with BHV-1 , PRV and EHV-1 , the UL49 . 5 gene products of these viruses are responsible for the inhibition of peptide translocation by TAP observed in virus-infected cells previously [50]–[52] . Next , the mechanism of TAP inhibition by the various UL49 . 5 proteins was investigated . Expression of BHV-1 UL49 . 5 strongly reduced TAP1 and TAP2 protein levels in human MJS cells [11] . It was shown that the cytoplasmic domain of UL49 . 5 is required for mediating proteasome-dependent degradation of TAP . To investigate whether BHV-1 UL49 . 5 has a similar mode of action in natural host cells , bovine MDBK cells were infected with wild type BHV-1 or a recombinant virus expressing a UL49 . 5 protein that lacks its cytoplasmic domain ( UL49 . 5Δtail ) . Steady state protein levels of bovine TAP were evaluated by immunoblotting . Whereas bovine TAP was readily detectable in uninfected MDBK cells , it was no longer observed in cell lysates from wild-type BHV-1 infected cells ( Fig . 5; upper panel , compare lanes 1 and 2 ) . Interestingly , in cells infected with the recombinant virus expressing the UL49 . 5Δtail mutant , TAP1 steady state levels were not affected ( compare lanes 2 and 3 ) . As a control , α-tubulin was consistently detected in all samples ( Fig . 5 , middle panel ) . Immunoprecipitation of UL49 . 5 from the infected cells confirmed the expression of the wild-type and recombinant proteins ( Fig . 5; lower panel ) . These findings indicate that the degradation of TAP by UL49 . 5 previously observed in human cells also occurs in bovine cells . In addition , like in human cells , the cytoplasmic domain of UL49 . 5 is critical to TAP degradation in the natural host cells . To further address the molecular basis of TAP inhibition mediated by PRV , EHV-1 , EHV-4 , and CHV UL49 . 5 , these proteins were stably expressed in human melanoma ( MJS ) cells . Like BHV-1 UL49 . 5 , the PRV and EHV-1 UL49 . 5 proteins were capable of blocking human TAP ( Fig . 6A ) . CHV UL49 . 5 did not inhibit peptide transport in human cells , while in canine cells some reduction in TAP activity was observed without a reduction of MHC class I surface expression ( Fig . 2A and B ) . Expression of BHV-1 UL49 . 5 in MJS cells resulted in reduced TAP1 and TAP2 protein levels , which is in accordance with previous observations ( Fig . 6B; compare lanes 1 and 2 ) [11] . In contrast , expression of the UL49 . 5 homologs of PRV and EHV-1 did not affect TAP1 and TAP2 steady state levels in MJS cells ( Fig . 6B; lanes 3 and 4 ) . These findings indicate that the UL49 . 5 homologs of PRV and EHV-1 inhibit peptide transport by TAP through a different mechanism than by mediating degradation of TAP . The translocation of peptides into the ER lumen is initiated by the association of peptides with the peptide-binding site of TAP [53] . To investigate whether the inhibition of peptide transport by PRV and EHV-1 UL49 . 5 involves blocking of peptide binding to TAP , microsomes were isolated from MJS cells expressing PRV and EHV-1 UL49 . 5 . Microsomes were incubated with a 125I-labeled reporter peptide ( Fig . 7 ) . At all concentrations tested , the peptide-binding capacity ( Bmax ) was similar for microsomes prepared from control cells and from cells expressing the PRV or EHV-1 UL49 . 5 proteins . Most importantly , the binding affinity ( Kd ) for the peptides was not changed by the viral inhibitors , demonstrating preservation of the peptide binding site of the TAP complex . This has also been observed for BHV-1 UL49 . 5 [11] and indicates that inhibition of TAP-mediated peptide transport by the BHV-1 , PRV , and EHV-1 UL49 . 5 proteins does not rely on interference with peptide binding . Since ATP-binding and hydrolysis are required to energize peptide translocation by TAP [54]–[56] , it was investigated whether the expression of the PRV , EHV-1 , and EHV-4 UL49 . 5 proteins affected binding of ATP to TAP . Previous experiments indicated that the BHV-1 UL49 . 5 protein did not influence the interaction of ATP with TAP [11] . The ATP-binding capacity of TAP in lysates from MJS cells ( control ) was compared to the binding in lysates from MJS cells stably expressing UL49 . 5 of PRV , EHV-1 or EHV-4 , or the HCMV-encoded US6 protein . US6 is known to strongly inhibit ATP-binding to TAP [22] , [24] . Cell lysates prepared in the presence of the mild detergent digitonin were incubated with ATP-agarose beads . Proteins bound to the ATP-agarose ( Fig . 8; pellet “P” ) were eluted from the beads with EDTA and displayed next to the unbound supernatant fractions ( Fig . 8; “S” ) . TAP1 and TAP2 were detected by immunoblotting . PRV UL49 . 5 did not alter the binding of ATP to TAP1 or TAP2 ( Fig . 8A; compare lanes 2 and 4 ) . As expected , the expression of US6 completely abolished the interaction of ATP with TAP ( Fig . 8A; lane 6 ) . These data show that TAP retains the capacity to bind ATP in the presence of PRV UL49 . 5 . In EHV-1 and EHV-4 UL49 . 5-expressing cells , neither TAP1 nor TAP2 could be detected in the ATP-agarose fraction ( Fig . 8B; compare lane 2 with lanes 4 and 8 ) . Since the C-terminus of UL49 . 5 is exposed in the cytosol , this domain might be responsible for the inhibition of ATP-binding to the nucleotide-binding domains of TAP . To evaluate whether the C-terminus of UL49 . 5 blocks ATP-binding to TAP , a truncated form of EHV-1 UL49 . 5 lacking the cytoplasmic domain was constructed and expressed in MJS cells . The EHV-1 UL49 . 5Δtail recombinant still interfered with ATP-binding to TAP ( Fig . 8B; lane 6 ) . When the association of wild type or mutant EHV-1 UL49 . 5 with TAP was disrupted by lysis of the cells in NP-40 , the ability of TAP1 and TAP2 to bind to the ATP-agarose was restored ( Fig . 8C , lanes 4 and 6; also compare Fig . 8B lanes 4 and 6 with Fig . 8C lanes 4 and 6 , respectively ) . These results indicate that the EHV-1 UL49 . 5 protein is capable of interfering with the recruitment of ATP by human TAP independent of the cytoplasmic domain of UL49 . 5 . The ability of EHV-1 and EHV-4 UL49 . 5 to interfere with the binding of ATP to equine TAP was assessed in E . derm cells ( data shown for EHV-4 ) . Like human TAP ( Fig . 8B lane 8 ) , equine TAP2 was not able to bind ATP-agarose in the presence of EHV-4 UL49 . 5 ( Fig . 8D; compare lanes 2 and 4 ) . When the experiment was performed in the presence of NP-40 , the ability of equine TAP2 to bind ATP was restored ( Fig . 8D; compare lanes 4 and 6 ) . These results indicate that the UL49 . 5 proteins of EHV-1 and EHV-4 inhibit human and equine TAP through similar mechanisms , rendering both human and equine TAP molecules incapable of recruiting ATP . To obtain further insight into the strategies used by EHV-1 and PRV UL49 . 5 to block TAP transport , Fluorescence Recovery After Photobleaching ( FRAP ) assays were performed . With this technique , conformational changes of TAP that occur during peptide translocation can be indirectly visualized by measuring the lateral mobility of green fluorescence protein ( GFP ) -tagged TAP within the ER membrane . It has been shown that the lateral mobility of TAP is inversely proportional to its activity , as peptide-transporting TAP molecules diffuse at a slower rate than inactive , closed TAP complexes [57] . In the absence of ATP , the translocation cycle cannot be initiated and consequently TAP will have a closed , more compact conformation . In agreement with this , depletion of ATP results in increased mobility of TAP in the ER membrane ( Fig . 9; control samples , compare black and grey bars ) . The complex can be trapped in the active conformation by adding long side chain peptides ( l . s . c . p . ) . These peptides bind to TAP , but cannot be translocated over the ER membrane , which results in a retained open conformation and therefore a slow diffusion rate of TAP in the ER membrane [57] ( Fig . 9; control samples , white bar ) . Expression of EHV-1 and PRV UL49 . 5 results in a decreased mobility of TAP ( Fig . 9; compare untreated samples/black bars ) . Whereas the diffusion rate of TAP increased considerably in ATP-depleted control cells , only a slight increase in TAP mobility was detected upon ATP depletion in the UL49 . 5-expressing cells ( grey bars ) . The failure of TAP to respond to ATP depletion in the EHV-1 UL49 . 5 cells is in agreement with the observation that this protein interferes with ATP binding to TAP ( Fig . 8B ) . Although ATP can still bind to TAP in the presence of PRV UL49 . 5 ( Fig . 8A ) , ATP depletion induces only a minor change in TAP mobility in the PRV UL49 . 5 cells ( Fig . 9 ) . Apparently , the presence of PRV UL49 . 5 prohibits conformational transitions that normally follow ATP-binding . In the presence of the UL49 . 5 proteins , l . s . c . p . were also unable to induce conformational changes within the TAP complex ( Fig . 9 ) . Since peptides can still bind to TAP in the presence of EHV-1 and PRV UL49 . 5 ( Fig . 7 ) , the failure of l . s . c . p . to induce conformational changes again suggests that the UL49 . 5 proteins arrest TAP in a translocation-incompetent state .
This study identifies the UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 as members of a novel class of viral immune evasion proteins . The UL49 . 5 gene products interfere with MHC class I antigen presentation by blocking the supply of antigenic peptides in the ER lumen through inhibition of TAP . Within the UL49 . 5 family of TAP inhibitors , heterogeneity is observed with respect to the mechanisms that underlie TAP inhibition . Whereas BHV-1 UL49 . 5 targets TAP for proteasomal degradation [11] , PRV and EHV-1 UL49 . 5 do not diminish the steady state levels of TAP1 or TAP2 . Interestingly , EHV-1 and EHV-4 UL49 . 5 interfere with the binding of ATP to TAP , a function that is not influenced by BHV-1 or PRV UL49 . 5 . All TAP-inhibiting UL49 . 5 proteins arrest the transporter complex in a translocation-incompetent state . UL49 . 5 homologs are encoded by all Herpesviridae analyzed to date [38] . However , the TAP-inhibiting capacities of these proteins appear to be restricted to certain members of the genus Varicellovirus . Members of this virus genus have co-evolved with their respective host species [39] . Viruses of even-toed ungulates or Artiodactyla like BHV-1 and PRV co-evolved with cattle and pigs; viruses of odd-toed ungulates or Perissodactyla ( EHV-1 and EHV-4 ) with horses; the carnivore viruses FHV-1 and CHV with cats and dogs , and the Old World primate virus VZV with humans [39] ( Fig . 10 ) . The identification of the UL49 . 5 proteins encoded by BHV-1 , PRV , EHV-1 , and EHV-4 as members of the UL49 . 5 family of TAP inhibitors suggests that more UL49 . 5 proteins with this property may be found in varicelloviruses of even- and odd-toed ungulate hosts . Considering the shared evolution of ( herpesviruses from ) carnivores and ( herpesviruses from ) odd-toed ungulates [39] , CHV UL49 . 5 was expected to inhibit TAP as effectively as EHV-1 and EHV-4 UL49 . 5 . However , the reduction of TAP-dependent peptide transport caused by CHV UL49 . 5 was very moderate compared to the inhibition by the other TAP-inhibiting UL49 . 5 proteins . The identification of the UL49 . 5 domains contributing to TAP inhibition will provide more insights into these differences . VZV infection of human cells results in reduced expression of MHC class I at the cells surface [45]–[47] . The VZV ORF66-encoded serine-threonine protein kinase has been shown to be one of the VZV proteins contributing to MHC class I down-regulation in VZV-infected cells [47] . However , a VZV recombinant lacking a functional ORF66 product still causes down-regulation of MHC class I surface expression , indicating that additional modulators of MHC class I-restricted antigen presentation are encoded by VZV . The observed down-regulation of MHC class I surface expression on VZV-infected cells [45]–[47] is not induced by UL49 . 5 when expressed individually . Despite the observed interaction between VZV UL49 . 5 and the peptide-loading complex , this protein alone did not block peptide transport by TAP and it had no effect on antigen recognition by HLA-A1 and HLA-A2-restricted CTL clones . As VZV occupies a somewhat isolated position in the phylogenetic tree of varicelloviruses ( Fig . 10 ) , it seems likely that evolutionary divergence has influenced VZV to acquire a separate mechanism to interfere with MHC class I-restricted antigen presentation . Alternatively , UL49 . 5 might co-operate with another unidentified VZV-encoded protein in order to reduce antigen presentation by MHC class I molecules . During virus infection , UL49 . 5 can be found in a complex with glycoprotein M ( gM ) . However , the co-expression of VZV UL49 . 5 and glycoprotein M has no effect on the expression of MHC class I molecules at the cell surface [47] , indicating that gM does not act as a modulator of UL49 . 5 with respect to TAP inhibition . Interaction with the conserved viral membrane glycoprotein M appears to be a common property of all UL49 . 5 homologs , as is the presence of a single cysteine residue in their ER-luminal/extracellular domain [41]–[44] . This cysteine residue is involved in the interaction of UL49 . 5 with gM , with which it forms a disulfide-linked heterodimers [37] , [40] . The complex of UL49 . 5 and gM is implicated in virion maturation and membrane fusion processes [58] , [59] . Interestingly , the interaction of BHV-1 UL49 . 5 with gM interferes with its capacity to block TAP [42] . Nevertheless , UL49 . 5 blocks peptide transport by TAP in BHV-1-infected cells . This may be explained by the fact that UL49 . 5 is expressed prior to and in excess of the early-late gM [42] . Interference with TAP-mediated peptide transport is an effective way of reducing CTL recognition and is used by several other herpesviruses , including HSV-1 and -2 , HCMV , MHV-68 , and EBV [12]–[36] . Compared to the other herpesvirus-encoded TAP inhibitors , the cross-species activity of UL49 . 5 proteins is remarkable . Except for CHV UL49 . 5 , the UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 all exhibit the ability to target human TAP . In addition , BHV-1 UL49 . 5 inhibits peptide transport by murine [60] , rat , equine , and porcine TAP ( D . K . L . and M . V . , unpublished observations ) . Human , porcine , bovine , and rodent TAP1 and TAP2 demonstrate a substantial degree of amino acid identity ( 70–80% ) [61] . Thus , the ability of UL49 . 5 proteins to act across species barriers most likely relies on structural homology within the TAP domains critically involved in UL49 . 5-TAP interaction . Apparently , this is less so for the domains within TAP that are targeted by US6 , mK3 and BNLF2a , whose actions seem to be restricted largely to the natural host species . BHV-1 UL49 . 5 reduces TAP protein levels in bovine , human , and murine cells , and also mediates degradation of human TAP in insect cells when co-expressed with UL49 . 5 [11] , [60] , [62] , indicating conservation of the pathway involved in this degradation process . The UL49 . 5 proteins exhibit unexpected differences in their mechanisms of TAP inhibition , despite their close evolutionary relatedness . The cytoplasmic domain of BHV-1 UL49 . 5 is essential for mediating degradation of both human and bovine TAP . EHV-1 and PRV UL49 . 5 have no influence on the stability of TAP . Apparently , degradation of TAP is facilitated by a yet unknown signal within the C-terminal domain of BHV-1 UL49 . 5 , which is not present in the other homologs . Studies to identify the nature of this sequence motif are in progress . The interaction of EHV-1 and EHV-4 UL49 . 5 with TAP blocks ATP binding to TAP . This feature distinguishes EHV UL49 . 5 from the other homologs studied . Interestingly , removal of the cytoplasmic domain of the EHV-1 UL49 . 5 protein did not restore the ability of TAP to bind ATP . Therefore , a direct interaction of EHV-1 UL49 . 5 with the cytosolic nucleotide binding domains of TAP is unlikely . Instead , the viral protein appears to arrest TAP in a translocation-incompetent state , incompatible with ATP-binding . This may resemble the type of structural change caused by HCMV US6 [23] , [63] . US6 , a type I transmembrane protein , interacts with the luminal side of the TAP transporter and blocks ATP-binding by prohibiting essential conformational rearrangements within TAP . The inability of the BHV-1 and PRV UL49 . 5 homologs to interfere with ATP-binding could be due to a slightly different conformational change induced by these proteins . Based on the results presented in this study , the UL49 . 5 proteins encoded by BHV-1 , PRV , EHV-1 , and EHV-4 can be classified as a new family of TAP-inhibiting proteins . These proteins share the ability of inducing a conformational arrest of TAP , which results in impaired peptide transport and inhibition of MHC class I-restricted antigen presentation . In view of these joint features it is likely that the TAP inhibiting UL49 . 5 proteins originate from a common ancestral protein , which acquired this capacity earlier during evolution . The VZV UL49 . 5 protein may be a rudimentary form with respect to TAP inhibition , or it may have lost its TAP inhibitory capacity later on . Alternatively , it may require additional VZV proteins for the inhibition of TAP . This study has revealed unexpected variation among UL49 . 5 proteins of varicelloviruses with respect to their mechanisms of TAP inhibition . These differences can be related to distinct evolutionary pathways of these varicelloviruses . The UL49 . 5 family of TAP-inhibiting proteins does not demonstrate any structural or functional similarity to TAP-inhibiting proteins encoded by other herpesviruses , for instance ICP47 , US6 , mK3 , or BNLF2a . This diversity of TAP-inhibiting proteins acquired by distantly related members of the subgroups of alpha- , beta- , and gammaherpesviruses is remarkable and presents a striking example of functional convergent evolution . At the same time , this identifies TAP as an Achilles' heel of the MHC class I antigen presentation pathway . Inhibition of TAP has apparently provided a strong advantage to these herpesviruses during co-evolution with their hosts .
Purified viral DNA from BHV-1 strain Lam and CHV strain Eva ( Animal Sciences Group , Lelystad , The Netherlands ) , PRV strain Kaplan [41] , EHV-1 strain Ab-4 ( kindly provided by J . Rola; National Veterinary Research Institute , Pulawy , Poland ) and EHV-4 ( kindly provided by R . de Groot; Dept . of Infectious Diseases and Immunology , Utrecht University , The Netherlands ) , and VZV ( viral DNA extracted from patient material; kindly provided by E . Klaas , Leiden University Medical Center , Leiden , The Netherlands ) were used as a template for polymerase chain reaction ( PCR ) amplification . PCR-reactions were performed under standard conditions using Pfu DNA polymerase ( Invitrogen ) and specific primers ( Table 1 ) for amplification of the full length coding sequence of the UL49 . 5 genes of BHV-1 [11] , PRV , EHV-1 , EHV-4 , CHV and VZV UL49 . 5 . The sequences of the primers are based on published sequences found in the NCBI database , except for the sequence of the CHV primers ( Haanes , E . and Rexann , F . ‘Recombinant canine herpesviruses’ , patent number EPO910406 , publication date 1997-08-21 ) . To generate the EHV-1 UL49 . 5Δtail construct , primers ( Table 1 ) were used to obtain a PCR product lacking 3′-terminal 45 nucleotides , thereby deleting the 15 carboxy-terminal amino acids . PCR-generated products were sequenced and inserted into the retroviral expression vectors pLZRS-IRES-GFP or pLZRS-IRES-ΔNGFR , upstream of the internal ribosome entry site ( IRES ) element . pLZRS vector information can be obtained at www . stanford . edu/group/nolan/retroviral_systems/retsys . html ) . The human melanoma cell line Mel JuSo ( MJS ) , MJS TAP1-GFP [57] and Madin-Darby bovine kidney ( MDBK ) cells ( American Type Culture Collection , ATCC ) were maintained in RPMI-1640 medium; GP2-293 cells , porcine kidney ( PK15 ) cells , the embryonic bovine trachea ( EBTr ) cell line , Madin-Darby canine kidney I ( MDCK I ) cells , and the equine epithelial cell line E . derm were maintained in DMEM medium . Media were supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) ( with the exception of E . derm cells that required 20% ) , 2 mM L-glutamine ( Invitrogen ) , 140 IU/ml penicillin and 140 µg/ml streptomycin . PHA-treated T-cell blasts positive for HLA-A1 and HLA-A2 were generated from PBMCs by stimulation with 0 . 8 µg/ml PHA and were subsequently cultured in IMDM supplemented with 100 IU/ml IL-2 and 10% FBS . The HLA-A2-expressing melanoma cell line 518 , Mel518 ( a kind gift from E . Verdegaal , department of Clinical Oncology , Leiden University Medical Center , Leiden , The Netherlands ) was maintained in DMEM containing 4 . 5 mM glucose , supplemented with 8% FBS , 2 mM L-glutamine ( Invitrogen ) , 140 IU/ml penicillin and 140 µg/ml streptomycin . Recombinant retroviruses were prepared using the Phoenix amphotropic packaging system as described previously ( www . stanford . edu/group/nolan/retroviral_systems/retsys . html ) . MJS , MDCK I , PK15 , and E . derm cells were transduced with recombinant retroviruses to generate the following stable cell lines: MJS , MDCK I , PK15 , and E . derm controls ( containing BHV-1 UL49 . 5 in the anti-sense orientation , GFP+ ) ; MJS UL49 . 5BHV-1 , PHA T-cell blast UL49 . 5BHV-1 and Mel518 UL49 . 5BHV-1 ( containing BHV-1 UL49 . 5 in the sense orientation ( SO ) , GFP+ ) ; MJS UL49 . 5VZV , PHA T-cell blast UL49 . 5VZV and Mel518 UL49 . 5VZV ( containing VZV UL49 . 5 SO , GFP+ ) ; MJS UL49 . 5CHV and MDCK I UL49 . 5CHV ( containing CHV UL49 . 5 SO , GFP+ ) ; MJS UL49 . 5PRV and PK15 UL49 . 5PRV ( containing PRV UL49 . 5 SO , GFP+ ) ; MJS UL49 . 5EHV-1 and E . derm UL49 . 5EHV-1 ( containing EHV-1 UL49 . 5 SO , GFP+ ) ; MJS UL49 . 5EHV-1Δtail ( containing tail-less EHV-1 UL49 . 5 SO , GFP+ ) ; MJS UL49 . 5EHV-4 and E . derm UL49 . 5EHV-4 ( containing EHV-4 UL49 . 5 SO , GFP+ ) . In addition , MJS TAP1-GFP cells were transduced with recombinant retrovirus to generate MJS TAP1-GFP control ( containing the empty pLZRS construct , ΔNGFR+ ) ; MJS TAP1-GFP UL49 . 5BHV-1 ( containing BHV-1 UL49 . 5 SO , ΔNGFR+ ) ; MJS TAP1-GFP UL49 . 5PRV ( containing PRV UL49 . 5 SO , ΔNGFR+ ) and MJS TAP1-GFP UL49 . 5EHV-1 ( containing EHV-1 UL49 . 5 SO , ΔNGFR+ ) . To generate recombinant retroviruses for MDBK cell line transductions , the GP2-293 pantropic packaging cell line was used according to the protocol obtained from BD Bioscience Clontech ( www . bdbiosciences . com ) . In brief , 1×106 of GP2-293 cells were co-transfected with retroviral expression vector ( pZLRS-IRES-GFP containing the BHV-1 UL49 . 5 gene in anti-sense or in the sense orientation ) and pVSV-G construct ( envelope vector ) for retrovirus production . Retrovirus-containing medium was collected 48 hours post-transfection . MDBK cells were transduced four times with VSV-G containing recombinant retroviruses to generate the following stable cell lines: MDBK control ( containing BHV-1 UL49 . 5 in anti-sense orientation , GFP+ ) and MDBK UL49 . 5BHV-1 ( containing BHV-1 UL49 . 5 SO , GFP+ ) . All cell lines generated in this study were selected for GFP or ΔNGFR expression using a FACSVantage cell sorter ( Becton Dickinson ) . To obtain MJS cells stably expressing the HCMV-encoded US6 ( MJS US6 ) , MJS cells were transfected with pcDNA3-US6-IRES-NLS-GFP and selected for neomycin resistance [64] . The following antibodies were used in this study: anti-transferrin receptor ( TfR ) monoclonal antibody ( mAb ) 66Ig10 , anti-TfR mAb H68 . 4 ( Roche ) , anti-human MHC class I complexes mAb W6/32 , anti-human MHC class I heavy chain mAb HC-10 ( kindly provided by H . Ploegh , Whitehead Institute , Cambridge , Massachusetts , USA ) , anti-human class II HLA-DR mAb Tü36 ( kindly provided by A . Ziegler , Institute for Immunogenetics , Universitätsklinikum Charité , Berlin , Germany ) , anti-TAP1 mAb 148 . 3 [54] and anti-TAP2 mAb 435 . 3 ( kind gift from P . van Endert , Institut National de la Santé et de la Recherche Médicale , Paris , France ) . For the detection of equine TAP2 , the polyclonal antibody anti-rat TAP2 Mac394 was used ( kindly provided by M . Knittler , Institute of Immunology , Friedrich-Loeffler-Institute , Tübingen , Germany ) . For preparation of bovine TAP1 specific antibody , the bovine TAP1 ORF sequence encoding amino acid residues 117 to 167 were amplified from bovine genomic DNA and cloned into pGEX-4T-2 ( GE Healthcare ) . The TAP1 polypeptide encompassing residues aa117-167 was purified as described previously [65] . The monoclonal antibody IL-A19 directed against bovine MHC class I molecules ( a kind gift from Dr . J . Naessens , ILRAD , Nairobi , Kenya ) . The anti-equine and anti-canine MHC class I complexes mAb H58A and anti-porcine MHC class I mAb PT85A were purchased from VMRD Inc . , Pullman , WA , U . S . A . Mouse anti-BHV-1 UL49 . 5 serum was kindly provided by G . J . Letchworth ( University of Wisconsin , Madison , Wisconsin , USA ) . Polyclonal rabbit anti-BHV-1 UL49 . 5 serum H11 was raised against a synthetic peptide representing the N-terminal sequence of BHV-1 UL49 . 5 and has been described [42] . In fig . 5 , a different polyclonal rabbit anti-BHV-1 UL49 . 5 was used , obtained using a synthetic peptide corresponding to amino acid residues 27–41 of UL49 . 5 ( [H] DAMRREGAMDFWSAGC*-[OH] ) . To facilitate conjugation to keyhole limpet hemocyanin , an additional irrelevant cysteine was added at the C terminus of the peptide ( indicated by * ) . Rabbits were immunized by as described earlier [66] . The rabbit antiserum raised against PRV UL49 . 5 ( gN ) has been described [67] , as was the anti-EHV-1 UL49 . 5 rabbit serum [43] . The VZV UL49 . 5-specific antibody was raised against two synthetic peptides: the N-terminal peptide EPNFAERNFWHASCSARGVYIDGSMITTLFKK and the C-terminal peptide RLFTRSVLRSTW . Both peptides were conjugated to glutathione S-transferase ( GST ) according to the methods described in [42] . The peptide-GST conjugates were mixed at a 1:1 ratio and emulsified in Freund's complete adjuvant for the first immunization and Freund's incomplete for the following immunizations . At 3-weeks intervals , the rabbit received four additional subcutaneous immunizations with the conjugates . Cells were trypsinized and resuspended in phosphate-buffered saline ( PBS ) containing 1% bovine serum albumin ( BSA ) and 0 . 05% sodium azide . Cells were incubated with specific antibodies on ice for one hour . After washing , the cells were incubated with phycoerytrin ( PE ) -conjugated anti-mouse antibody for 45 min . Stained cells were analyzed by flow cytometry on a FACSCalibur flow cytometer ( Becton Dickinson ) . To exclude dead cells , 7-aminoactinomycin D ( 7-AAD , Sigma-Aldrich ) was added at a concentration of 0 . 5 µg/ml to all samples before analysis . Cells were analyzed using CellQuest software ( Becton Dickinson ) . The fluorescence-based peptide transport assay was performed as previously described [11] , [68] . In brief , MJS cells were permeabilized with Streptolysin O ( Murex Diagnostics Ltd . ) at 37°C , followed by incubation with the fluorescein-conjugated synthetic peptide CVNKTERAY ( N-core glycosylation site underlined ) in the presence or absence of ATP . Peptide translocation was terminated by adding ice-cold lysis buffer containing 1% Triton X-100 . After lysis , cell debris was removed by centrifugation , and supernatants were collected and incubated with Concanavalin A ( ConA ) -Sepharose ( Amersham ) . After extensive washing of the beads , the peptides were eluted with elution buffer ( 500 mM mannopyranoside , 10 mM EDTA , 50 mM Tris-HCl pH 8 . 0 ) by vigorous shaking and further separated from ConA by centrifugation at 12 , 000×g for 2 minutes . The fluorescence intensity was measured using a fluorescence plate reader ( CytoFluor , PerSeptive Biosystems; excitation 485 nm/emission 530 nm ) . The data were analyzed using the unpaired t-test . Statistical significance was set at p <0 . 05 . Cells were lysed in a buffer containing 1% ( wt/vol ) digitonin , 50 mM Tris·HCl ( pH 7 . 5 ) , 5 mM MgCl2 , 150 mM NaCl , 1mM leupeptin , and 1 mM AEBSF ( 4- ( 2-Aminoethyl ) -benzenesulfonyl fluoride ) , and subjected to immune precipitations using anti-TAP1 mAb 148 . 3 o/n . To determine steady state protein levels , cells were lysed in Nonidet P-40 ( NP-40 ) lysis mix containing 50 mM Tris-HCl , pH 7 . 4 , 5 mM MgCl2 and 0 . 5% NP-40 , supplemented with 1 mM AEBSF ( 4- ( 2-Aminoethyl ) -benzenesulfonyl fluoride ) , 1 mM leupeptin and 20 µM Cbz-L3 ( Carbobenzoxy-1-Leucyl-1-Leucyl-1-Leucinal-H; Peptides International , Inc ) . The samples were kept on ice throughout the experiment . Protein complexes were denatured in reducing sample buffer ( 2% SDS , 50 mM Tris pH 8 . 0 , 10% glycerol , 5% β-ME , 0 . 05% bromophenol blue ) for 5 min at 96°C . Immunoblotting ( IB ) analysis was performed on denatured cell lysates separated by SDS-PAGE and blotted onto polyvinylidene fluoride ( PVDF ) membranes . Blots were incubated with the antibodies as indicated , followed by horseradish peroxidase-conjugated goat-anti-mouse or swine-anti-rabbit Igs ( DAKO and Jackson Laboratories ) , and visualized by ECLplus ( Amersham ) . Steady-state labeling of MDBK cells with [35S]-methionine/cysteine and subsequent immunoprecipitations with rabbit BHV-1 UL49 . 5-specific antibody were performed as described [69] . Immunoblotting procedures with rabbit anti-bovine TAP1 and rabbit anti-α-tubulin have been described [65] . A total of 1 , 000 51Cr-labeled target cells were incubated with different CD8+ CTL clones at various effector to target ratios . The HY-A1 clone recognizes an HY epitope in the context of HLA-A1 , and the HA2 . 27 clone recognizes the histocompatibility antigen HA-2 in the context of HLA-A2 . After 4 hours of incubation at 37°C , 51Cr release into the supernatant was measured using standard methods . The mean percentage of triplicate wells was calculated as follows: % specific lysis = ( experimental release–spontaneous release ) / ( maximal release–spontaneous release ) ×100 . For analysis of IFN-γ production , 20 , 000 T-cells were co-cultured with 10 , 000 target cells . After 24 hours , the supernatant was harvested and the concentration of IFN-γ was measured by standard ELISA ( Sanquin , Amsterdam , The Netherlands ) . The wild type viruses used in this study were: BHV-1 strain Lam , BHV-1 strain Cooper ( Fig . 5 ) , PRV strain Kaplan and EHV-1 strain RacL1 . The UL49 . 5 deletion mutant of PRV used in this study has been described before [41] , [43] . The UL49 . 5 deletion mutant of EHV-1 ( strain RacL11 ) was a gift from J . von Einem ( College of Veterinary Medicine , Cornell University , Ithaca , NY , USA ) . Infections with wild type and mutant herpesviruses were carried out on the following cell lines: MDBK cells for BHV-1; PK15 cells for PRV and E . derm cells for EHV-1 . The cells were washed once with PBS and infected with BHV-1 or PRV at an m . o . i . of 10 , and with EHV-1 at an m . o . i . of 5 at 37°C in serum-free medium . After 2 hours , medium containing 10% FBS was added . Mock-infected cells were treated under the same conditions as infected cells . After 5 hours of infection , cells were collected and prepared for the peptide translocation assay . For immunoblotting and metabolic labeling experiments , MDBK cells were infected with BHV-1 wt and UL49 . 5Δtail viruses for 12 hours . The BHV-1 UL49 . 5 mutant was generated by homologous recombination , using BHV-1 strain Lam as parent strain . The recombination region upstream of the UL49 . 5 gene was a 1 . 4 kb fragment running from nucleotide residue 7670 to 9061 ( residue numbers based on the complete BHV-1 genome with NCBI accession number NC_001847 , updated 30 March 2006 ) . This fragment starts at a BstXI site 1 . 3 kb upstream of the start codon of the UL49 . 5 open reading frame and ends at its amino acid residue 31 . The recombination region downstream of the UL49 . 5 gene was provided by a 1 . 9 kb fragment from nucleotide residue 9075 to 10972 . This fragment starts at amino acid residue 36 of UL49 . 5 and ends at an FspI site 1 . 7 kb downstream of the UL49 . 5 open reading frame . A 2 . 2 kb NruI–PvuII fragment was cloned between the two UL49 . 5 recombination fragments that carries the hGFP gene in the expression cassette of pcDNA3 ( Invitrogen ) . The complete recombination fragment ( 5 . 5 kb ) was co-transfected with purified BHV-1 Lam DNA into EBTr cells using a calcium phosphate-based transfection method . After plating the supernatant of freeze/thawed transfected cells , a green plaque was found that , following three rounds of plaque purification , failed to react with anti-BHV-1 UL49 . 5 serum . The BHV-1-UL49 . 5 mutant could be grown to a titer of 107 . 0 TCID50/ml and was capable of penetrating bovine cells with the same kinetics as the wild type Lam strain . The BHV-1 recombinant virus gN Am80 , expressing a form of UL49 . 5 lacking its cytoplasmic domain , was constructed by introducing an amber mutation at gN residue 80R ( AGG to TAG ) by using a BHV-1 BAC clone ( Liu and Chowdhury , manuscript in preparation ) . Cellular microsomes were prepared as described [70] . Microsomes isolated from 7×106 homogenized cells were pre-incubated in 50 µl of AP buffer ( 5 mM MgCl2 in phosphate-buffered saline , pH 7 . 0 ) on ice for 45 min in the absence or presence of a 200-fold molar excess of the non-labeled TAP-specific viral inhibitor ICP47 [18] . Different concentrations of radiolabeled peptide ( RR[125I]YQKSTEL ) were added equally to the samples with or without ICP47 and incubated on ice [71] . Non-bound peptides were removed by washing the membranes with 400 µl of AP buffer and subsequent centrifugation at 20 , 000 g for 8 min . The amount of radioactivity bound to the membranes was quantified by γ-counting and corrected for the signal obtained in the presence of ICP47 . All experiments were performed in triplicate . TAP binding to ATP-agarose was assayed as described [11] . In brief , cells were solubilized in 1% ( w/v ) digitonin , 50 mM Tris-HCl ( pH 7 . 5 ) , 5 mM MgCl2 , 150 mM NaCl , 5 mM iodoacatamide , and 1 mM AEBSF . Hydrated C-8 ATP-agarose ( Fluka/Sigma ) was added to the post-nuclear supernatant and incubated by rotation at 4°C . After 2 hours , the supernatant was separated from the ATP-agarose pellet by 5 minutes centrifugation . The resulting pellet was washed three times with 0 . 1% ( w/v ) digitonin , 50 mM Tris-HCl ( pH 7 . 5 ) , 5 mM MgCl2 and 150 mM NaCl . Proteins bound to the ATP-agarose were eluted with 500 mM EDTA and SDS sample buffer was added to both the supernatant and the pellet . The samples were separated using SDS-PAGE and analyzed by immunoblotting . Confocal microscopy and Fluorescence Recovery After Photobleaching ( FRAP ) assays were performed as described [11] , [57] . In short , a circular spot in the ER was bleached at full intensity , and an attenuated laser beam was used to monitor recovery of fluorescence . The half-time for recovery was calculated from each recovery curve after correction for loss of fluorescence caused by imaging ( usually <4% ) . The diffusion coefficient D was determined from at least seven cells measured in different experiments . UL49 . 5 sequence data used to generate the alignment shown in Fig . 1 and the phylogenetic tree shown in Fig . 7 , have been obtained from the NCBI ( www . ncbi . nlm . nih . gov ) database with the accession numbers: bovine herpesvirus 1 ( BoHV-1 ) [NP_045309] , bovine herpesvirus 5 ( BoHV-5 ) [NP_954898] , cercopithecine herpesvirus 1 ( CeHV-1 ) [AAP41468] , cercopithecine herpesvirus 9 ( CeHV-9 ) [NP_077423] , cercopithecine herpesvirus 16 ( CeHV-16 ) [YP_443897] , equid herpesvirus 1 ( EHV-1 ) [AAT67267] , equid herpesvirus 4 ( EHV-4 ) [CAA35670] , gallid herpesvirus 1 ( GaHV-1 ) [YP_182341] , gallid herpesvirus 2 ( GaHV-2 ) [NP_057812] , gallid herpesvirus 3 ( GaHV-3 ) [NP_066882] , human herpesvirus 1 ( HSV-1 ) [NP_044652] , human herpesvirus 2 ( HSV-2 ) [NP_044520] , human herpesvirus 3 ( VZV ) [YP_068406] , meleagrid herpesvirus 1 ( MeHV-1 ) [AAG30090] , psittacid herpesvirus 1 ( PsHV-1 ) [AAQ73691] , suid herpesvirus 1 ( PRV ) [YP_068325] , transporter 1 ATP-binding cassette sub-family B [Bos taurus] [AAY34698] . Not obtained from the NCBI database are: bubaline herpesvirus 1 ( BuHV-1 ) [MSRSLLVALATAALLAMVRGLDPLLDAMRREE AMDFWSAGCYARGVPLSEPPQAMVVFYAALTVVMLAVALYAYGLCFRLMSAGGPNKKEVRGRG; FAMR , unpublished] , canid herpesvirus ( CHV ) [patent EPO910406 http://ep . espacenet . com ] , cervid herpesvirus 1 ( CvHV-1 ) [MARMPRSLLSALAVAALLAIAGARDPLLDAMRHEGAMDFWSASCYARGVPL SEPPQALVVFYVALAVVMFSVAVYAYGLCLRLVGADSPNKKDSRGRG; FAMR , unpublished] .
|
Herpesviruses have the conspicuous property that they persist for life in the infected host . This is also the case for varicelloviruses , a large subfamily of herpesviruses with representatives in humans ( varicella zoster virus or VZV ) , cattle ( bovine herpesvirus 1 or BHV-1 ) , pigs ( pseudorabies virus or PRV ) , and horses ( equine herpesvirus or EHV type 1 and 4 ) , among many others . Cytotoxic T-lymphocytes play an important role in the protection against viral infections , which they detect through the recognition of virus-derived peptides , presented in the context of MHC class I molecules at the surface of the infected cell . The transporter associated with antigen processing ( TAP ) plays an essential role in this process , as TAP imports peptides into the compartment where peptide loading of the MHC class I molecules takes place . In this study , we show that the UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 all block the supply of peptides through the inhibition of TAP , but that the mechanisms employed by these proteins to inhibit TAP function exhibit surprising diversity . VZV UL49 . 5 , on the other hand , binds to TAP , but does not interfere with peptide transport . Our study classifies the UL49 . 5 proteins of BHV-1 , PRV , EHV-1 , and EHV-4 as members of a novel family of viral immune evasion proteins , inhibiting TAP through a variety of mechanisms .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology/immunomodulation",
"immunology/antigen",
"processing",
"and",
"recognition",
"virology/immune",
"evasion"
] |
2008
|
Varicellovirus UL49.5 Proteins Differentially Affect the Function of the Transporter Associated with Antigen Processing, TAP
|
The effect of biodiversity on the ability of parasites to infect their host and cause disease ( i . e . disease risk ) is a major question in pathology , which is central to understand the emergence of infectious diseases , and to develop strategies for their management . Two hypotheses , which can be considered as extremes of a continuum , relate biodiversity to disease risk: One states that biodiversity is positively correlated with disease risk ( Amplification Effect ) , and the second predicts a negative correlation between biodiversity and disease risk ( Dilution Effect ) . Which of them applies better to different host-parasite systems is still a source of debate , due to limited experimental or empirical data . This is especially the case for viral diseases of plants . To address this subject , we have monitored for three years the prevalence of several viruses , and virus-associated symptoms , in populations of wild pepper ( chiltepin ) under different levels of human management . For each population , we also measured the habitat species diversity , host plant genetic diversity and host plant density . Results indicate that disease and infection risk increased with the level of human management , which was associated with decreased species diversity and host genetic diversity , and with increased host plant density . Importantly , species diversity of the habitat was the primary predictor of disease risk for wild chiltepin populations . This changed in managed populations where host genetic diversity was the primary predictor . Host density was generally a poorer predictor of disease and infection risk . These results support the dilution effect hypothesis , and underline the relevance of different ecological factors in determining disease/infection risk in host plant populations under different levels of anthropic influence . These results are relevant for managing plant diseases and for establishing conservation policies for endangered plant species .
Understanding the relationship between the risk of infectious diseases and host ecology is a long-standing goal of biological research , central for the management of current infectious diseases and for preventing the emergence of new ones . Indeed , changes in host ecology are among the most frequently identified causes of disease emergence ( i . e . the increase of disease incidence following its appearance in a new , or previously existing , host population ) [1]–[3] . Because infectious diseases involve interactions between at least two species , it has been proposed for a long time that ecosystem biodiversity will play a key role in disease risk . Current declines in biodiversity have been proposed to be linked with the emergence of infectious diseases , which have fueled a renewed interest on this subject [4] . Two major hypotheses with different predictions relate biodiversity to disease risk . The “Amplification Effect” hypothesis predicts that diversity will be positively correlated with disease risk , as it will result in increased abundance of inoculum sources for a focal host . The “Dilution Effect” hypothesis predicts a negative correlation between biodiversity and disease risk , as a reduction in diversity could result in an increased abundance of the focal host species facilitating disease transmission [5] . These two hypotheses can be considered to represent extremes of a continuum , as the effects of diversity on disease risk would be related to the host range of the pathogen: an Amplification Effect would require a generalist pathogen , while the more restricted the host range of the pathogen , or the higher the differences between shared hosts in their ability to amplify or transmit the pathogen , the higher the Dilution Effect . Increasing evidence derived from pathogens with broadly different life-styles indicates that biodiversity reductions most often result in increased disease risk [4] . The idea linking biodiversity with disease risk is not new in animal or plant pathology . Two classical hypotheses in plant pathology state that the high impact of plant diseases in crops is associated with: i ) the reduced species diversity , and higher host density , of agroecosystems as compared to wild ecosystems [6]; ii ) the reduced genetic diversity of crops as compared to their wild ancestors or relatives [7] . However , despite that a number of recent studies on the ecology of plant diseases have been added to those dating from the 1980s , support for these hypotheses is still often circumstantial [8] . Attention has focused on analyses of foliar diseases caused by fungi , which mostly indicate that increased biodiversity reduces disease risk [9]–[14] . Remarkably , there are fewer reports referring to viral diseases , which represent a large fraction of emergent plant pathogens [15] , and may differ from fungal ones in their relationship to biodiversity . While most plant pathogenic fungi are directly transmitted specialists [16] , most plant-infecting viruses are vector transmitted , and are host generalists but often vector specialists [17] . Most studies with plant viral diseases have focused on generalist viruses infecting grasses , generally finding an amplification effect [18]–[21] . Interestingly , work on plant diseases largely failed to assess the role of various possible mechanisms by which reduced biodiversity may affect disease risk ( but see [10] , [11] , [22] ) . Particularly , it is often difficult to differentiate the effects of increased host density and of reduced species diversity [4] . Hence , there is a need of research aimed at analyzing the effects of biodiversity on plant disease risk and , specifically , at disentangling the role of the various factors associated to ecosystem diversity . This is the goal of the present work . The focal host in this study is the wild pepper Capsicum annuum var . glabriusculum ( Dunal ) Heiser and Pickersgill [23] , also known as “chiltepin” . Chiltepin is found in Mexico in a variety of habitats from the Yucatan peninsula and the Gulf of Mexico to the Sonoran desert [24] , [25] . Chiltepin is a deciduous , perennial bush that grows for 5–8 years and vegetates and reproduces during the rainy season . Birds disperse the seeds from its red pungent fruits [24] . Human harvesting of fruits from wild chiltepin plants is a common practice in central and northern Mexico [26] , [27] . A second level of human exploitation involves tolerance or favoring the growth of spontaneously dispersed chiltepin plants in anthropic habitats , such as pastures and living fences ( i . e . , let-standing plants , sensu [28] ) . Last , chiltepin cultivation in home gardens or in small traditional plots has started in the recent past [25] . Cultivation has not yet lead to domestication , and cultivated chiltepin populations , which are managed as annual crops , do not show obvious phenotypic differences with wild ones [25] . Wild chiltepin populations show a large genetic variation and a strong spatial structure associated with the biogeographical province of origin , and human management results in a significant loss of both spatial structure and genetic diversity [25] . This habitat diversity makes chiltepin a uniquely good system to analyze the relationship between biodiversity and disease risk . We focused on two contrasting pepper-infecting virus groups . The first involves two species of the genus Begomovirus ( Geminiviridae ) : Pepper golden mosaic virus ( PepGMV ) , and Pepper huasteco yellow vein virus ( PHYVV ) , here treated collectively as “begomoviruses” . These species have a two-segmented single-stranded ( ss ) DNA genome; narrow host ranges limited in nature mostly to species of the genera Capsicum , Solanum and Datura ( Solanaceae ) , and are transmitted in a persistent manner by the whitefly Bemisia tabaci Gennadius ( Homoptera , Aleyrodidae ) [29]–[31] . The B biotype of B . tabaci , characterized by a broad plant host range , a high reproductive potential , and a high efficiency as a vector for begomoviruses , is prevalent in Central and North America [32] . The second virus is Cucumber mosaic virus ( CMV ) , genus Cucumovirus ( Bromoviridae ) , with a tripartite , ssRNA genome , and a typical generalist , infecting more than 1000 species of both mono- and dicotyledonous plant families . CMV is transmitted in a non-persistent manner by more than 80 species of aphids , thus being also a vector generalist [33] . Utilizing these host-pathogen systems we specifically addressed if: i ) modification of chiltepin habitat associated with different levels of human management resulted in changes in disease or infection risk , ii ) reduction of species diversity increases disease or infection risk , iii ) decreased host genetic diversity had an effect on disease or infection risk , iv ) increased host plant density resulted in increased disease or infection risk and v ) the above effects were different for viruses with different life-histories . To answer these questions , we visited over three years neighboring wild , and human managed ( i . e . , of let standing and cultivated plants ) chiltepin populations from different biogeographic provinces in Mexico . For each population , species diversity , host density and the prevalence of plants showing symptoms of virus infection were quantified in the field as an estimate of disease risk . Plants were collected at each population and their status ( infected/non-infected by several viruses ) was determined in the laboratory in order to estimate infection risk . Results indicate that disease and begomovirus infection risks , but not CMV infection risk , decrease with increasing biodiversity . We propose that observed differences between begomovirus and CMV infection risk can be due to different transmission modes .
Chiltepin populations were visited during the summers of 2007 to 2009 at different sites over the species distribution range in Mexico ( Figure 1 and Table 1 ) . A total of 26 populations were localized in different habitats representing three levels of human management: i ) ten wild populations ( W ) in which fruit gathering by local people may occur; ii ) six populations of let-standing plants ( here from called “let-standing populations” ) , in anthropic habitats , either pastures ( LSP ) or live fences ( LSF ) , in which chiltepin plants are tolerated or favored , and iii ) ten cultivated populations ( C ) either at home gardens ( CHG ) or at small monocultures ( CMC ) . Population sites were assigned to 6 biogeographical provinces: Yucatan ( YUC ) , Eastern side of the Sierra Madre Oriental ( SMO ) , Altiplano Zacatecano-Potosino ( AZP ) , Costa del Pacífico ( CPA ) , Costa del Pacífico Sur ( CPS ) and Sonora ( SON ) [34] . A total of 14 populations , located in YUC , SMO , AZP , CPA and CPS , were visited during 2007 and 2008 . The 2009 survey was extended to other populations of these five biogeographical provinces and SON , to a total of 26 populations ( Table 1 ) . Populations were visited between the 15 of July and the 30 of August , in an attempt to homogenize plant phenology among locations at the stage of flowering and beginning of fruit setting . Due to the highly unpredictable rain regime at some regions , or to extinction , not all populations could be surveyed for the three years . At each location , the following information was collected: 1 ) The census of the chiltepin population . 2 ) The status of each censused plant: asymptomatic or showing symptoms commonly related to virus infection ( i . e . , mosaic , leaf curl , leaf lamina reduction , and/or stunting ) . 3 ) The area ( m2 ) occupied by the chiltepin population . 4 ) The inventory of the non-herbaceous vegetation , determined as the number of individuals of each bushy or arboreal species , in the same area of the chiltepin population , to estimate species richness , and evenness according to the Shannon index [35] . Populations BER-w , PEL-w , MOC-w and MAU-w ( Table 1 ) were too large – i . e . , more than 200 plants – to census all plants , and both chiltepin censuses and biodiversity inventories were limited to a fixed transect . In this case , the area occupied by the chiltepin population was calculated by prospecting a width of 4 m along the fixed itinerary . At each population and visit , plants were systematically sampled for laboratory analyses . Plants were sampled regardless of their showing or not symptoms: One plant out of every x plants was sampled along fixed itineraries , with itinerary length and x ( 0<x≤4 ) depending on population size , 1–3 young branches with fresh leaves were collected per plant . Infection by CMV and by Potyvirus species was analyzed by DAS-ELISA , using commercial antisera against CMV or a monoclonal antibody against a highly conserved motif in the coat protein of potyviruses ( Agdia Biofords ) , according to the manufacturer's instructions . Infection by Chiltepin yellow mottle virus ( ChYMV , Tymoviridae ) was analyzed by molecular hybridization using a 32P-labeled RNA probe complementary to nucleotides 5365–5777 of ChYMV genomic RNA ( Accession No . FN563124 ) [36] . Infection by species of the genus Begomovirus was detected by PCR using degenerate primers designed on the alignment of DNA-A sequences of 43 begomovirus species from the New World: BAOPsp ( 5′-GCGCCCTGCAGGGGCCYATGTAYAGGAAGCC-3′ ) and BAONsp ( 5′-GCGCGCGGCCGCGANGCATGNGTACATGCCAT-3′ ) , which amplify a region in the coat protein gene located between nucleotide positions 392 and 884 in the genome of PepGMV ( Accession No . AY928512 ) . Molecular hybridizations and PCR were performed on total nucleic acid preparations from chiltepin leaves extracted by grinding 200 mg of fresh leaf tissues in three volumes of 200 mM Tris-HCl pH 9 , 25 mM EDTA , 1% SDS , 400 mM LiCl , followed by phenol-chloroform extraction [25] . Plants genotyped using the set of 9 nuclear microsatellites markers described in [25] were used to estimate genetic diversity of the 26 chiltepin populations . Generalized linear mixed models ( GLMM ) were used to analyze the difference in the prevalence of virus infection ( Begomovirus and CMV ) , and in the frequency of symptomatic plants , according to chiltepin population , biogeographical province and level of human management of the population , considering these factors as fixed effects . The rationale for considering population as a fixed effect is that all the chiltepin populations that we were able to find were included in the analyses , rather than using a random representation of them . The symptom and virus prevalence values determined for each population in the different years were considered as dependent measures; thus , they were treated as repeated measures in the GLMM . This seems the correct approach for wild and let-standing populations , in which at least a subset of the plants sampled over the years were the same , since chiltepin plants live for several years . This might not be so for cultivated populations , in which plants are managed as an annual crop and may change plots over the years . However , we considered that plots from different years were close enough to be spatially correlated , and therefore repeated measures are warranted . In addition , results did not differ when data from cultivated populations were analyzed as independent measures ( not shown ) . To determine whether values of analyzed traits were significantly different among classes within each factor , Bonferroni analyses were employed in all cases using the GLMM marginal means calculated for each class [37] . GLMM accommodates missing data , so that the 26 chiltepin populations sampled could be included in the analysis . Parallel analyses using only the 8 populations for which data on the 3 years of sampling were available yielded comparable results ( data not shown ) . The contributions of each ecological factor to the variation in virus and symptom prevalence were estimated using a Principal Component Analysis ( PCA ) . Host plant density ( d ) , host plant genetic diversity estimated as expected heterozygosity ( He ) , species diversity estimated as species richness ( number of species , SR ) and Shannon index ( Sh ) , of 24 chiltepin populations were scaled to zero mean and unit variance , inserted in a regression matrix and rotated to obtain the principal components ( PCs ) . Importantly , species diversity was not measured in TLA-w and HER-cmc , so that these populations were excluded from the analysis . Significance thresholds for the load of each ecological factor on a PC were determined using a broken-stick model [38] . Bivariate analyses , considering both linear and non-linear models , of begomoviruses , CMV and symptom prevalence onto the ecological factors and their corresponding PCs , yielded the proportion of the variance in each of these variables explained by each factor and each principal component ( R2 ) , and the significance of these correlations . For these bivariate analyses , we utilized the GLMM marginal means of begomoviruses , CMV and symptom prevalence calculated for each population . Statistical analyses were performed using the statistical software package SPSS 17 . 0 ( SPSS Inc . , Chicago , USA ) . Information theory was used to determine the relative importance of the ecological factors in the variation of symptom and virus prevalence [39] . This approach was chosen because it allows making inferences across a set of causal model structures for symptom and virus prevalence [39] . To do so , a set of models that included a global model , which contained all ecological factors ( species richness , Shannon index , expected heterozygosity and host plant density ) , and nested models , which contained different combinations of the predictor variables was fitted . Since species richness and Shannon index always loaded in the same PC , different selection model analyses in which the nested models considered SR , or Sh , or both variables together were performed . The three approaches gave similar results . For simplicity , only results considering species richness are shown . We ranked the models according to second order Akaike's Information Criteria ( AICc ) to account for small sample size ( R library: AICcmodavg ) [39] . The model with the lowest AICc score was selected as the best-ranked model . We calculated AICc Delta ( Δi ) , as the difference between the AICc of a given model and that of the best-ranked model . Delta quantifies how strongly models compete ( Δi = 0 for best-ranked model; Δi = 1–2 indicates substantial empirical support; Δi = 4–7 indicates considerable less support; and Δi>10 indicates no support [39] ) . Finally , the Akaike relative weight ( ωi ) of each model was calculated following the expression: ωi = exp ( Δi ) /Σexp ( Δi ) .
The status of a total of 1820 censused plants was recorded during the summers of 2007–2009 . The prevalence of plants showing symptoms of virus infection ( symptomatic plants ) ( Table 1 ) marginally varied among year ( χ2 = 5 . 86 , P = 0 . 060 ) , ranging between 16 . 2% and 21 . 6% of the census . A subset of 1081 plants , either symptomatic or asymptomatic , was analyzed for infection by ChYMV ( a chiltepin-infecting tymovirus , see [36] ) , CMV , begomoviruses or potyviruses . Low prevalence of potyvirus infection ( 2 . 87% ) precluded further analyses . ChYMV infection was limited to locations around Tula , AZP ( Table 1 ) where its prevalence was high ( 42 . 86% ) . Infection by CMV and by begomoviruses was detected during the three years of the study in all biogeographical provinces and under different levels of human management ( Figure 1 and Table 2 ) . PepGMV and PHYVV were the only begomovirus species detected infecting chiltepin , and their relative prevalence did not depend on the level of human management of the chiltepin population ( data not shown ) . Therefore , from here on these two species will be considered together and referred to as “begomoviruses” . CMV prevalence remained stable ( ≈7% ) among years ( χ2 = 0 . 06 , P = 0 . 970 ) , while begomovirus prevalence was about 3–5 times higher ( 19–36% ) and varied largely according to year ( χ2 = 58 . 25 , P<1×10−5 ) ( Table 2 ) . Begomoviruses , CMV or ChYMV infection explained the symptoms of 212/281 ( 78 . 7% , of these 81% being infected by begomoviruses ) laboratory-analyzed symptomatic plants from all populations and years . This fraction did not differ according to the level of human management of the population ( 59/76 analyzed symptomatic plants for wild populations; 50/70 for let-standing populations , and 103/135 for cultivated populations ) ( χ2 = 0 . 86 , P = 0 . 651 ) . The fraction of infected plants showing symptoms ( 212/369 , i . e . , 57 . 4% in total ) was lower in wild populations ( 59/133 , 44 . 4% ) than in cultivated ( 103/153 , 67 . 3% ) or in let-standing populations ( 50/83 , 60 . 2% ) ( χ2≥4 . 54 , P≤0 . 033 ) . This fraction did not differ between the later two levels of human management ( χ2 = 0 . 89 , P = 0 . 345 ) . The effect of geography ( biogeographical province and chiltepin population ) , and level of human management in the prevalence of symptomatic plants , begomoviruses or CMV , was analyzed . GLMM analyses using biogeographical province as a fixed effect showed that neither the prevalence of symptomatic plants , begomoviruses or CMV did depend on this factor ( F5 , 47>0 . 797 , P<0 . 557 ) . Similarly , the prevalence of CMV infection did not vary among chiltepin populations ( F25 , 47 = 1 . 512 , P = 0 . 108 ) . However , population was a factor determining the prevalence of symptomatic plants and begomovirus infection ( F25 , 47>4 . 369 , P<1×10−4 ) . Bonferroni-corrected multiple comparisons showed that this was solely due to the higher prevalence in populations HUJ-chg and LIB-cmc ( P<0 . 046 in 21/25 populations in both cases ) , and when these populations were removed from the analysis , population was no longer a factor in the prevalence of symptomatic plants and begomovirus infection ( F23 , 45 = 1 . 984 , P = 0 . 183 ) . Populations HUJ-chg and LIB-cmc were not excluded from further analyses in order to consider as much of the variability in the analyzed factors as possible . These results show that the biogeographical factors analyzed largely do not affect viral and symptom prevalence . Consequently , the populations corresponding to each level of human management could be analyzed together . The level of human management was a factor in determining the prevalence of symptomatic plants , ( F2 , 47 = 7 . 619 , P<1×10−4 ) , which was significantly lower in wild than in cultivated populations ( P<1×10−3 ) , with let-standing populations showing an intermediate value ( P≥0 . 276 ) . Similarly , human management also affected the prevalence of begomovirus infection ( F2 , 47 = 5 . 774 , P = 6×10−4 ) . Values were higher in cultivated populations than in wild ( P = 6×10−4 ) and intermediate in let-standing populations ( P≥0 . 076 ) . However , CMV prevalence did not vary depending on this factor ( F2 , 47 = 1 . 459 , P = 0 . 243 ) . Thus , increased levels of human management are associated with higher prevalence of symptomatic plants and begomoviruses , but not to prevalence of CMV . We therefore explored which ecological factors varying between populations with different levels of human management were linked to these differences in disease and virus infection risk . The relative importance of focal host plant density ( d ) , host genetic diversity ( expected heterozygosity , He ) and species diversity of the habitat expressed either as species richness ( SR ) or considering also species evenness ( Shannon index , Sh ) , on chiltepin populations was analyzed by including these variables in a PCA . Relevant statistical parameters of these ecological factors are provided in Table S1 of the Supporting Information . Parallel PCAs were performed considering all populations together and individually for each level of human management ( Table 3 ) . Importantly , SR and Sh loaded in the same PC in all cases . Since both variables represent the same ecological factor , we performed separate PCAs considering either SR or Sh , but the choice of index did not alter the results ( data not shown ) . The PCA using the data set that included all the populations ( All ) yielded three main PCs collectively explaining 95 . 7 percent of the total variance . Species diversity ( SR and Sh ) was highly associated with PC1 , d with PC2 , and He with PC3 ( Squared loadings>81 . 9 ) ( Table 3 , All column ) . The PCA restricted to wild populations , largely mirrored the results obtained with the All data set . However , the fraction of total variance explained by PC1 was higher , and that explained by PC2 and PC3 lower than in the All analysis ( Table 3 , Wild column ) . PCAs considering either let-standing or cultivated populations separately yielded PCs explaining similar percentages of the variance than in the All data set . However , variables loading in each PC differed from All and wild data sets . For let-standing populations , He was now associated with PC1 , SR and Sh with PC2 , and d with PC3 ( Squared loadings>82 . 7 ) ( Table 3 , Let-standing column ) . Similarly , in cultivated populations He was associated with PC1 , SR/Sh with PC2 , and d with PC3 ( Squared loadings>85 . 0 ) ( Table 3 , Cultivated column ) . Importantly , SR , Sh and d loaded positively into their respective PCs , but the loading of He was always negative ( not shown ) . The results above indicate that the relative importance of the ecological factors considered in this study vary depending on the level of human management of the chiltepin population . To determine how human management affects species and genetic diversity , and plant density , we performed GLMM analyses on each PC obtained with the All data set using level of human management as a factor . The three major PCs significantly differed depending on the level of human management ( F2 , 24≥4 . 995 , P≤0 . 015 ) . Values of PC1 ( species diversity ) were significantly higher in wild than in cultivated populations ( P = 0 . 045 ) , with intermediate values in let-standing populations . The opposite trend was observed for PC2 ( plant density ) . For PC3 ( host genetic diversity ) , values for cultivated populations were lower than in let-standing and wild populations ( P≤0 . 015 ) , the later two types not differing ( P = 0 . 952 ) . To further explore the association between the considered ecological factors and disease risk , the influence of each ecological factor on symptom , begomovirus and CMV prevalence was studied using model selection analyses . For the All data set , symptom , begomovirus and CMV prevalence were chiefly determined by species diversity , either measured as SR ( Table 4 ) or Sh ( not shown ) . The model including only species richness was unambiguously the best ( ω>0 . 79 ) ( Table 4 ) . In wild populations , symptom prevalence was mainly associated with species richness and host density ( ω = 0 . 45 and 0 . 33 , respectively ) , and begomovirus prevalence was chiefly associated with host density ( ω = 0 . 40 ) . However the best-ranked model included also host genetic diversity ( ω = 0 . 55 ) . Although the best-ranked model explaining CMV prevalence included all the ecological factors , the single factor that best explained the variable was host genetic diversity ( ω = 0 . 12 ) ( Table 4 ) . Model selection analyses were largely similar for let-standing and cultivated populations . In both types of populations , host genetic diversity best explained symptom and begomovirus prevalence ( ω = 0 . 51 in both cases ) , with the model including all the ecological factors showing slightly lower weight . Finally , host density chiefly determined CMV prevalence ( ω = 0 . 57 and 0 . 35 ) , but the model considering all factors showed slightly higher weight in cultivated populations ( Table 4 ) . Thus , these analyses are in agreement with the PCA . The effect of ecological factors in symptom and virus prevalence was analyzed by bivariate analyses of each factor onto the prevalence of symptomatic plants , begomoviruses and CMV . For the All data set , SR was negatively associated with symptom and begomovirus prevalence ( P<0 . 050 ) , explaining 31 . 1% and 20 . 2% of the variance in these variables , respectively ( Figure 2 and Table S2 ) . Therefore , species diversity was the primary predictor of symptom and begomovirus prevalence . In wild populations , SR explained 20 . 2% of the variance being negatively associated with symptom prevalence ( P = 0 . 026 ) ( Figure 2 and Table S2 ) , and d explained 27 . 7% ( P = 0 . 039 ) of the variance in symptom prevalence , and 44 . 4% ( P = 0 . 038 ) of the variance in begomovirus prevalence . Thus , in wild populations species diversity had also a principal role in determining symptom prevalence , with a lesser effect of plant density in symptom and begomovirus prevalence . In contrast , in let-standing populations He was negatively associated with symptom and begomovirus prevalence ( P≤0 . 025 ) , explaining 85 . 7% and 65 . 1% of the variance in these two traits , respectively . In addition , a negative correlation between SR and CMV prevalence was found ( P = 0 . 041 , 45 . 8% of the variance explained ) , and d showed a positive association with CMV prevalence ( P = 0 . 048 , 31 . 1% of the variance explained ) ( Figure 2 and Table S2 ) . Finally , He in cultivated populations was also negatively associated with symptom prevalence ( P≤0 . 015 , 55 . 2% of the variance explained ) , and d explained 28 . 5% of the variance in CMV prevalence ( P = 0 . 022 ) ( Figure 2 and Table S2 ) . Parallel bivariate analyses using the PCs associated with species richness , Shannon index , expected heterozygosity and host density of each PCA , instead of the original variables yielded similar results ( Table 3 and Figure S1 ) .
We have analyzed the prevalence of virus disease and virus infection in populations of a wild plant to test whether increased host density and decreased host genetic diversity in agroecosystems , as compared with wild ecosystems , favors disease risk . These two classical hypotheses of plant pathology [6] , [7] are particular cases of a more general one , which is receiving much attention recently , stating that habitat biodiversity is a determinant of disease risk [4] and may be at the root of disease emergence [3] , [4] . The wild pepper or chiltepin was the focal host for this study , taking advantage of some unique characteristics of this species . First , wild populations of chiltepin are found in a large variety of habitats in different biogeographical provinces of Mexico [24] , [25] , which anticipated large differences in species diversity among habitats , as was indeed the case ( Table S1 ) . Second , the genetic diversity of wild chiltepin populations differs according to their geographical origin as shown for the 10 wild populations analyzed here [25] . Last , chiltepin populations show different levels of human management , including populations of let-standing plants , which are not sown or planted , but are tolerated or protected in anthropic habitats; and cultivated populations , in which plants are sown in home gardens or in small traditional plots . The risk of virus disease was estimated as the prevalence of symptomatic plants . Although unapparent virus infection may affect plant fitness [40] , we call here diseased plants those showing macroscopic symptoms . This is grounded in our ( unpublished ) observations of a fecundity reduction in symptomatic plants as compared to both infected or non-infected asymptomatic ones . However , we are aware that prevalence of symptomatic plants may underestimate disease risk , if symptom development were correlated with increased host mortality . Considering these caveats , the risk of disease was positively correlated with the level of human management of the population , being higher in cultivated than in wild populations , and intermediate in let-standing populations . Hence , results support the concept that transition of host habitat from wild ecosystems to agricultural ones results in an increase of disease risk . A GLMM analysis of the variation of three PCs – associated with species diversity , host genetic diversity and host density – according to the level of human management , strongly suggested that the higher disease risk associated with increased human management is determined by a reduction of biodiversity , both as species diversity and host genetic diversity , and/or by an increased host density . It should be noted that habitat biodiversity and host genetic diversity , estimated as SR and He , respectively , vary along a continuum over the three levels of human management of chiltepin populations , which should avoid spurious associations . However , we cannot discard that other factors structured according to the level of human management , not specifically addressed in this work , may influence disease risk . Examples could be time of exposure to virus infection , and nutrient availability . Nevertheless , the three PCs associated to the analyzed ecological factors explained more than 95% of the variance of the analyzed variables , regardless that all populations were considered together or differentiating between wild , let-standing and cultivated populations . Therefore , although other ecological factors might have minor effects on disease risk , those here considered accounted for most of the variance within and between levels of human management , and may largely explain the emergence of viral disease associated with human management of chiltepin populations . More specifically , both PC and model selection analyses showed that , for the All and wild data sets , species diversity of the habitat was the major predictor of disease risk ( Figure 2 , Tables 3 and 4 ) . For let-standing and cultivated populations host genetic diversity was the major predictor of disease risk ( Figure 2 , Tables 3 and 4 ) . The risk of infection by begomoviruses , which mostly explained symptoms , followed a largely similar pattern , except for a noticeable role of host density in determining virus prevalence in wild populations . These results support the dilution effect hypothesis for a plant-virus system . Moreover , they stress the importance of preserving biodiversity to maintain ecosystem services , a key concept in conservation biology [4] . Results also agree with most analyses of a variety of animal [5] , [41] and plant systems [9]–[14] , contributing to extend this hypothesis to plant virus diseases . The relationship between biodiversity and disease risk has received comparatively little attention in wild plant-infecting viruses . To our knowledge , Cereal- and Barley yellow dwarf luteoviruses ( C/BYDV ) , which infect many species of grasses and are transmitted in a persistent manner by aphids in a highly species-specific way , is the best characterized system . In this case , most results are compatible with the amplification effect hypothesis , although the complex relationships between grass species , vector multiplication , and virus multiplication/transmission , make the effect of biodiversity on disease risk largely dependent on species composition [18]–[21] , [42] , [43] . Differences in life histories between luteoviruses and begomoviruses , which cause most symptoms of virus disease in chiltepin ( Tables 1 and 2 ) , could explain why effects of biodiversity vary between both systems . Begomoviruses have a narrow host range [29]–[31] , and are persistently transmitted by B . tabaci , which has a wide host range [32] . Consequently , the larger the species diversity of the habitat , the larger the number of plant species in which B . tabaci can feed , and the lower fraction of meals resulting in begomovirus transmission to the focal host , resulting in host encounter reduction ( sensu [5] ) . Interestingly , other reports of a dilution effect of biodiversity also refer to persistently transmitted viruses in which the host range of the virus is narrower than that of the vector [44] , [45] . Importantly , which of the two different components of biodiversity was the primary predictor of disease and begomovirus infection risk depended on the level of human management . The reduced weight of species diversity in anthropic habitats could be explained by species diversity being largely reduced in cultivated vs . wild populations , and not varying largely among let-standing populations ( Table S1 ) . Host genetic diversity has been shown to have a negative effect on the risk of fungal diseases in crops [46]–[48] . Results from fungal pathogens were interpreted as due to differences in resistance-susceptibility among host genotypes , resulting in decreased transmission efficiency [46]–[48] . This mechanism could be also invoked to explain our results as differences in resistance to begomovirus infection have been reported among chiltepin genotypes [49] . However , genotype diversity might also reduce pathogen transmission by other mechanisms , for instance , microenvironment changes [13] or modification of the behavior of insect vectors [50] . The reduction of disease and begomovirus-infection risk with higher biodiversity was not coupled to a lower chiltepin density , as host plant density always loaded into a different PC than species or host diversity . An accepted axiom in plant pathology is that higher host density leads to higher disease risk . However , data are scarce and mostly inconclusive [6] , [8] , [9] , and the effects of biodiversity and host density on disease risk are often difficult to differentiate [4] . The few works that attempted to differentiate these effects yielded contrasting results: density was the primary factor determining disease risk [10] , [11] or there were independent and complex interactions between the effects of both factors [22] . The methodology used here avoids artificial correlations [51] and allowed disentangling the effects of these two ecological factors on disease and begomovirus infection risk . Interestingly , infection by CMV followed a different pattern: infection risk did not depend on the level of human management , and host plant density was a relevant parameter in managed populations , but not in wild ones . The different pattern of infection risk found for begomoviruses and CMV could be due to differences in their life histories . At odds with begomoviruses , CMV is a generalist regarding the host and the vector and , perhaps more importantly , it is transmitted in a non-persistent manner . While persistent transmission is effected during feeding periods among plants that are hosts of the aphid vector , non-persistent transmission occurs during probing visits to plants that need not be hosts of the aphids , which remain viruliferous for short periods of time [52] . Thus , proximity of plants susceptible to the virus could be more important than biodiversity in determining CMV infection risk , similarly to directly transmitted fungi infecting leaves [10] , [11] . Consequently , the mechanisms of transmission , in addition to the host range of the pathogen and/or its vectors , could be a primary factor in determining the relationship between biodiversity and disease risk , an unexplored issue , to our knowledge . Finally , a larger fraction of begomovirus- or CMV-infected plants showed symptoms in managed populations than in wild ones , strongly suggesting a higher virulence of virus infection in the former , perhaps due to a higher susceptibility of plants in human-managed populations to virus infection and its effects . The relationship between host physiological condition and disease susceptibility is an underexplored subject [53] . However , we could speculate that plants of managed populations , which benefit from higher levels of water and/or nutrients than those from wild habitats , as shown by their production of about five times as much fruits ( our unpublished observations ) , would be more competent hosts for virus vectors [20] , [54] . This would encourage more frequent and longer meals , thus being under higher inoculum pressure of persistently transmitted viruses . Also , a more favorable host condition could result in higher levels of virus multiplication [21] , [54] , [55] . If this were the case , in addition to suffering more virulent virus infections , plants in cultivated and let-standing populations would be more competent hosts for virus vectors , virus multiplication and transmission . These factors would contribute to the higher disease risk , and thus to disease emergence , in human managed populations , regardless of the ecological factors here analyzed . However , we cannot exclude that the larger proportion of symptomatic plants in managed habitats would be the result of increased life span of infected plants due to the enhanced availability of resources in cultivated and let-standing populations , which could contribute to explain our observations . In summary , our results show the important role of biodiversity reduction in the emergence of viral diseases associated to human management of plant populations . Our work also suggest that other ecological and genetic factors , perhaps resulting in increased virulence in anthropic habitats , need to be considered in order to fully understand the dynamics of emergence , which should be the subject of future research .
|
Biodiversity has been proposed as a major ecological factor determining disease prevalence . However , the relationship between biodiversity and disease risk remains underexplored . Few studies focus on host-virus systems and , particularly on plant viruses . To address this subject the prevalence of virus infection and disease symptoms was monitored in wild-pepper ( chiltepin ) populations under different levels of human management . For these populations , species diversity , host genetic diversity and host plant density were determined . Higher levels of human management resulted in increased disease and virus infection risk , which was associated with decreased habitat species diversity and host genetic diversity , and with increased host plant density . More specifically , for wild chiltepin populations , species diversity of the habitat was the primary predictor of disease risk; and host genetic diversity was the primary predictor in managed populations , with host density being generally a poorer predictor of disease risk . These results support a dilution effect of biodiversity on disease risk , and underline the relevance of different ecological factors in determining disease risk in wild and in human-managed habitats .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"ecology",
"virology",
"biology",
"microbiology",
"biodiversity",
"host-pathogen",
"interaction"
] |
2012
|
Effect of Biodiversity Changes in Disease Risk: Exploring Disease Emergence in a Plant-Virus System
|
The mechanism of circadian oscillations in mammals is cell autonomous and is generated by a set of genes that form a transcriptional autoregulatory feedback loop . While these “clock genes” are well conserved among animals , their specific functions remain to be fully understood and their roles in central versus peripheral circadian oscillators remain to be defined . We utilized the in vivo inducible tetracycline-controlled transactivator ( tTA ) system to regulate Clock gene expression conditionally in a tissue-specific and temporally controlled manner . Through the use of Secretogranin II to drive tTA expression , suprachiasmatic nucleus– and brain-directed expression of a tetO::ClockΔ19 dominant-negative transgene lengthened the period of circadian locomotor rhythms in mice , whereas overexpression of a tetO::Clockwt wild-type transgene shortened the period . Low doses ( 10 μg/ml ) of doxycycline ( Dox ) in the drinking water efficiently inactivated the tTA protein to silence the tetO transgenes and caused the circadian periodicity to return to a wild-type state . Importantly , low , but not high , doses of Dox were completely reversible and led to a rapid reactivation of the tetO transgenes . The rapid time course of tTA-regulated transgene expression demonstrates that the CLOCK protein is an excellent indicator for the kinetics of Dox-dependent induction/repression in the brain . Interestingly , the daily readout of circadian period in this system provides a real-time readout of the tTA transactivation state in vivo . In summary , the tTA system can manipulate circadian clock gene expression in a tissue-specific , conditional , and reversible manner in the central nervous system . The specific methods developed here should have general applicability for the study of brain and behavior in the mouse .
Most organisms possess an endogenous circadian system that drives the daily timing of many physiological and behavioral processes . Genetic screens , spontaneous mutants , and gene-targeting approaches have been key in unraveling the essential set of genes underlying the circadian mechanism in mammals , Drosophila , and other model systems [1–4] . At the molecular and biochemical levels , a set of core clock genes govern positive and negative autoregulatory feedback loops of transcription and translation to form the core mechanism of the circadian clock in mammals [2 , 5] . The central oscillator is primarily driven by two bHLH-PAS transcription factors within the positive feedback loop , CLOCK and BMAL1 , which heterodimerize and transactivate downstream clock and clock-controlled genes by binding to E-box elements that lie within their promoters [6–9] . The core constituents of the negative feedback loop are the Cry and Per genes , which are transcriptionally driven by CLOCK and BMAL1 . PER and CRY proteins accumulate , associate with each other in the cytoplasm , translocate to the nucleus , and inhibit the CLOCK and BMAL1 activation of their own transcription [9] . As the negative elements turn over , CLOCK and BMAL1 renew their cycle of transcription of the Per and Cry genes . In mammals , nearly all cells in the body contain circadian oscillators organized in a hierarchical fashion , with a master pacemaker located in the suprachiasmatic nucleus ( SCN ) of the anterior hypothalamus [5 , 10 , 11] . The SCN is entrained to the 24-h daily light–dark cycle via retinal light input and , in turn , synchronizes and coordinates the rhythms of peripheral tissue clock cells [2 , 5] . In mammals , luciferase reporters of circadian genes [10 , 11] in conjunction with single cell imaging have been valuable in revealing self-sustained circadian oscillators in virtually every cell in the body [12–15] . These studies have shown that most peripheral organs and tissues can express circadian rhythms in isolation; however , inputs from the dominant circadian pacemaker in the SCN are essential in coordinating circadian rhythms in an intact animal [10 , 11 , 16 , 17] . For example , SCN transplant experiments have shown that SCN-lesioned arrhythmic animals and genetically arrhythmic mice take on the rhythm of the donor SCN [17–19] . Similarly , transplanted mouse embryonic fibroblasts exhibit a circadian period and phase characteristic of the host rhythm and phase [20] . These findings have led to a widely accepted hierarchical model of the mammalian circadian system in which the SCN acts as pacemaker that drives and synchronizes peripheral circadian oscillators . Thus , understanding the physiological and functional relationships among central and peripheral clocks is essential; however , we still do not fully understand how the SCN governs peripheral oscillators to regulate circadian rhythms in physiology and behavior in multicellular organisms . In nonmammalian systems , such as in Drosophila , analyses of the regulatory interactions of circadian genes led to the exploitation of novel tools to drive circadian genes , such as tissue-specific expression of transgenes ( TGs ) and reporters , which are valuable in elucidating the complexity of circadian system [21–23] . Conditional systems utilizing heat shock promoters have been developed in order to drive the temperature-dependent expression of circadian TGs [24–26] . Exogenous promoters in conjunction with the GAL4-UAS bipartite transgenic system have been valuable in expressing circadian TGs in subregions of the brain or distinct groups of circadian neurons , and even in ablation of discrete circadian neurons [27–30] . In mammals , however , other than ubiquitous inactivation of circadian genes by gene targeting techniques in embryonic stem cells or studies using the culture/explant-based system , the use of tissue-specific conditional regulation of circadian genes has not been reported [2 , 3] . Thus , to elucidate cellular and behavioral networks in the mammalian circadian system , more refined approaches are required , especially those affording temporal and spatial control of gene expression in vivo . The ability to regulate TG expression in a conditional manner has made the tetracycline-controlled transactivator ( tTA ) system an attractive tool for use in mammalian systems . The tTA system was originally developed by Bujard and colleagues for the conditional expression of reporter genes in mammalian cells [31–34] and has been successfully applied in many experiments to study a variety of developmental processes and brain function in mammals [35–45] . The first component of this system contains a tissue-specific promoter that drives the expression of tTA , a fusion of the Escherichia coli tetracycline repressor sequence to the C-terminal transactivation domain of the herpes simplex virus VP16 gene that converts the repressor into a transcriptional activator . Expression of the target TG by tTA is achieved by introducing the TG of interest downstream of a minimal cytomegalovirus promoter sequence linked to multiple copies of the tet operator ( tetO ) sequence . Conditional and inducible regulation of the target TG is contingent on the ability of tTA to bind to the tetO sequences and activate transcription in a tetracycline-dependent manner . TG expression can be turned off with the administration of doxycycline ( Dox ) , a tetracycline derivative , which prevents binding of tTA to the tetO sequence . Thus , unlike the site-specific recombinase Cre-loxP and Flp-FRT systems [46 , 47] , which only allow conditional and/or tissue-specific inactivation of genes [48] , the tTA system permits repeated cycles of conditional activation and inactivation of genes within the same animal . Hence , the tTA system provides truly conditional investigation of gene function . Despite the widespread use of the Tet system to manipulate various genes in a variety of tissues in mammals , only a few studies have used the system in the brain , and even fewer have used the system to study regulation of behavior [35 , 37–39 , 41–43 , 45 , 49] . A limitation of the tTA system in the study of brain function in vivo has been the slow induction of the TG ( i . e . , reversal ) upon removal of standard dose of Dox [49] . Here we show that regulation of ClockΔ19 or Clockwt TG expression occurs with rapid kinetics of induction and repression , causing an immediate reversion of the mutant to wild-type ( WT ) phenotype , and vice versa , in a tissue-specific manner . We demonstrate that the CLOCK protein is an excellent indicator for the kinetics of Dox-dependent induction/repression in the brain . Using activity rhythms as an output , the circadian period length provides a daily readout of the transactivation state of the Tet system in vivo . The development of the tTA system for conditional TG expression in the brain/SCN will open new avenues of research to answer fundamental questions of mammalian circadian biology .
The tTA ( also known as Tet-Off ) system requires two independent lines of transgenic mice ( Figure 1A ) : one line expresses tTA under the control of a specific promoter , and a second line carries a tTA-responsive tetO promoter linked to the target gene of interest [31–34] . When two TGs are introduced into a single mouse through mating , the tetO-linked gene is activated but only in those cells that express tTA . Expression of the target TG can be suppressed by Dox . To generate an SCN/brain-enriched transactivator line , we first searched for SCN-enriched transcripts that have expression restricted to the brain from the comprehensive mouse tissue expression databases [50 , 51] . This analysis revealed about 35 candidate genes that were either very highly expressed in the SCN or SCN/brain enriched . We then analyzed these candidate genes for SCN expression using in situ hybridization . From this screen , we found one gene , Secretogranin II ( Scg2 ) , that has very high constitutive expression in the SCN and has an expression pattern limited to brain , pituitary , and adrenal glands [52] . Scg2 is a member of the neuroendocrine/neuronal secretory proteins , which are widely distributed in endocrine , neuroendocrine , and neuronal cells [53 , 54] . Although the exact role of Scg2 is not well understood , research suggests that Scg2 plays biologic roles in neurotransmission and paracrine regulation of central and peripheral actions of the nervous and neuroendocrine systems [54 , 55] . Because of its brain-enriched pattern of expression and , in particular , its strong expression in the SCN , we explored Scg2 as a tissue-specific activator for the tTA system . We generated an SCN/brain-enriched transactivator line using a 9 . 8-kb promoter region of Scg2 . We examined the pattern of expression using an oligo probe specific to the tTA transcript and showed distinct and strong tTA expression in the SCN but also throughout the brain ( Figure 1Ba and 1Bb ) . To characterize tTA expression further , we crossed the Scg2::tTA line to mice carrying a tetO promoter-lacZ reporter construct [35] . We tested two independent tetO promoter-lacZ reporter lines , namely lines lac1 and lac2 ( kindly provided by Mark Mayford , Scripps Research Institute , San Diego , California , United States ) . Both reporter lines showed similar SCN/brain-enriched expression patterns as demonstrated by β-galactosidase staining ( Figure 1Bc and 1Bd ) . We used the Scg2::tTA line for all subsequent experiments described in this report . Next , we generated the target line , which carries a tetO promoter fused to a dominant-negative ClockΔ19 mutant allele [56 , 57] . In order to clearly assess the inducible expression of the TG , we also fused a hemagglutin ( HA ) epitope tag on the 3′ end of the cDNA , which does not interfere with CLOCK transcriptional activation [58] . We produced eight independent tetO::ClockΔ19-HA transgenic lines , and all lines were crossed to Scg2::tTA mice to evaluate whether we could control the expression of the TG in a temporal manner . We assessed whether the ClockΔ19-HA TG is expressed in Scg2::tTA/tetO::ClockΔ19-HA double transgenic mice by in situ hybridization using oligo probes targeted to exon 19 of the Clock gene and to the HA tag ( Figure 1C ) . The original ENU-induced ClockΔ19 mutation is an A-to-T transversion in a splice donor site , causing skipping of exon 19 [57]; thus , the exon 19-specific probe detects only the endogenous WT Clock transcript . Conversely , the HA tag-specific probe detects only the ClockΔ19-HA TG transcript . All mice showed endogenous WT Clock transcript expression as expected . However , while only double transgenic Scg2::tTA/tetO::ClockΔ19-HA mice showed clear hybridization with the HA tag-specific probe , single transgenic ( Scg2::tTA or tetO::ClockΔ19-HA ) and WT mice did not show any expression of the HA tag . Thus , expression of the ClockΔ19-HA TG , which is evident in the SCN and throughout the brain , is induced by tTA in the double transgenic mice only . Similarly , Western blot analyses showed that only the double transgenic mice expressed the mutant TG protein ( Figures 1D and S1A ) . Anti-CLOCK antibody detects both the mutant and WT proteins , which can be discriminated by their size difference . The lack of exon 19 in the mutant allele results in the deletion of 51 amino acids and subsequently produces a shorter protein for the ClockΔ19-HA TG compared to the WT allele . All eight independent tetO::ClockΔ19-HA transgenic lines expressed the mutant protein when crossed to Scg2::tTA mice ( Figures 1D and S1A ) . In addition , the mean density ratio measurement of the WT and mutant proteins in Western blot analyses revealed that expression of the mutant protein is 0 . 51- to 1 . 05-fold relative to the WT protein in these lines . This was also demonstrated by Southern analysis ( unpublished data ) , which showed that target lines carried approximately one or two copies of the TG . In order to test for leakiness of the tetO promoter , we inspected whether the tetO::ClockΔ19-HA single transgenic mice from each independent line expressed the TG using the anti-HA antibody against the brain lysates ( Figure S1B ) . Of the eight independent target lines , only one ( line 12 ) showed activation of the TG ( likely due to a position effect on the TG ) , and therefore this line was excluded from further analysis . Thus , we obtained seven independent tetO target lines that showed tightly regulated inducible expression . We next assessed tissue specificity and spatial expression of the ClockΔ19-HA TG in double transgenic mice . Western blot analyses indicated that expression of the TG is brain and SCN enriched ( Figure 1E ) . Because Scg2 participates in the neuroendocrine system [54 , 55] , as expected , the TG was also expressed in the pituitary gland ( Figure 1E ) ; however , there was no expression in other peripheral tissues , such as the kidney , lung , liver , and spleen . Using fluorescence microscopy and an HA tag antibody , we explored the spatial expression of the TG within the SCN and found that ClockΔ19-HA TG expression is nuclear ( Figure 1F ) . In WT mice , CLOCK has been shown to be nuclear and is expressed constitutively in the mouse SCN [59] . In contrast , rhythmically expressed negative factors of the circadian gene , such as PER1 , PER2 , and CRY1 , accumulate in the nuclei at the end of circadian night [59 , 60] . To delineate further the spatial organization of the HA-containing cells within the SCN , we compared HA expression with the distribution of neuropeptide markers ( Figure 1G–1L ) . Classically , the SCN is organized into two major subdivisions: a core that lies adjacent to the optic chiasm and is composed of neurons that predominantly produce vasoactive intestinal polypeptide ( VIP ) and a shell that surrounds the core and contain a large population of neurons that produce arginine vasopressin ( AVP ) [61–64] . Examination of the HA-containing cells showed that more than 90% of the cells in the SCN express the TG and that expression occurs throughout both the core and the shell . Serial sections through the rostral-caudal extent of the SCN showed that HA-expressing cells colocalized with AVP- and VIP-expressing cells ( Figures 1F–1L , S2 , and S3 ) . Indeed , all AVP- and VIP-positive cell bodies in the SCN were also HA positive . Taken together , the detailed evaluation using Western blot and immunocytochemical analyses demonstrates that Scg2-driven ClockΔ19-HA TG expression is tissue specific and enriched in the brain and in the majority of cells in the SCN , including the core and the shell . The original N-ethyl-N-nitrosourea ( ENU ) -induced ClockΔ19 mutant mice exhibit several behavioral alterations compared to WT mice , one of which is the lengthened free-running period of wheel-running activity . Heterozygous ClockΔ19/+ mice show a lengthening of the circadian period of approximately 1 h , while homozygous ClockΔ19/ClockΔ19 mice exhibit periods about 4 h longer compared to WT mice and fail to express persistent circadian rhythms when maintained in constant darkness [65] . Because the ClockΔ19 mutant allele behaves as an antimorph ( one type of dominant-negative mutation ) , it competes with the WT allele in the generation of circadian rhythms [56 , 57 , 66] . In this study , because the Scg2::tTA/tetO::ClockΔ19-HA double transgenic mice express the ClockΔ19-HA TG as well as both copies of the endogenous WT CLOCK , we hypothesized that these mice would exhibit a free-running period of wheel activity rhythm similar to that of the heterozygous ClockΔ19/+ mice . To test this prediction , running-wheel behavior of all four genotypes from the Scg2::tTA x tetO::ClockΔ19-HA matings was recorded ( Figure 2 ) . All mice entrained to a 12/12-h light–dark cycle , initiating their nightly bouts of activity at the beginning of the dark phase with the majority of locomotor activity restricted to the night . Upon release into constant darkness ( DD ) , however , only the double transgenic mice displayed a lengthened circadian period approximately 1 h greater than that of WT mice ( 23 . 9 to 24 . 8 h versus 23 . 4 to 23 . 7 h , respectively ) . Pairwise comparisons indicated that there were no significant differences in circadian period among the single transgenic lines ( Scg2::tTA or tetO::ClockΔ19-HA ) compared to WT littermates , while the period observed in double transgenic mice was significantly longer compared to the WT , Scg2::tTA , and tetO::ClockΔ19-HA mice ( mean DD comparison among all genotypes , F3 , 78 = 118 . 77 , p < 0 . 00005; each pairwise comparison [double transgenic versus others] , p < 0 . 0005 ) . Double transgenic mice from all seven independent tetO lines showed lengthening of the circadian period similar to the circadian period exhibited in the heterozygous Clock mutant mice . Therefore , conditional expression of the dominant-negative mutant allele TG in the SCN and brain is sufficient to drive altered running wheel activity behavior . As described above , these independent lines bear one or two copies of the TG , and each line displays slightly different expression levels ( 0 . 51- to 1 . 05-fold; Figure 1D ) . It is noteworthy that those double transgenic mice , which carry one or two copies of the ClockΔ19 TG and two endogenous WT copies of Clock , display a free-running period similar to that of the Clock heterozygote mice ( ClockΔ19/+ ) , which carry one copy of each allele ( mutant and WT Clock ) . ( There was a trend toward longer circadian period in lines that expressed higher levels; however , the differences were subtle . ) It is not surprising , therefore , that we do not observe a circadian rhythm phenotype similar to that of Clock homozygous-like behavior ( i . e . , 4-h lengthening and arrhythmicity ) in double transgenic mice , given the relative magnitude of TG expression in relation to the expression of WT CLOCK . One advantage of tetracycline controlled gene expression is that a gene of interest can be repeatedly turned on and off noninvasively . To examine Dox-dependent transactivation , we treated all mice initially with 2 mg/ml Dox in their drinking water ( Figure 2 ) . Upon treatment , the Scg2::tTA/tetO::ClockΔ19-HA double transgenic mice showed a shortening of circadian period in about a day , indicating a rapid suppression of ClockΔ19-HA TG expression . Period estimates during the Dox treatment showed a mean circadian period of 23 . 7 h ( SD = 0 . 196 , n = 16 ) in double transgenic mice , similar to their WT littermates ( 23 . 6 h , SD = 0 . 108 , n = 22 ) . Next , we examined whether the effect was reversible by removing Dox from the drinking water after 3 wk of treatment . The circadian period of double transgenic mice , however , did not immediately return to the previous long circadian period; rather , it took longer than 3 mo for reversal ( Figure 3 ) . One possibility for this long and gradual reversal of the circadian period may be due to period “after effects” , a form of behavioral plasticity in which the free-running period of circadian behavior undergoes experience-dependent changes , such as , in this case , exposure to Dox [67 , 68] . However , Dox is a potent effector substance for the tetracycline-controlled gene expression system and it is also lipophilic in nature; thus , a likely possibility is that the dosage used may have been too high to achieve rapid clearance and subsequent rapid reversal . Earlier studies have shown that 2 mg/ml Dox produces a rapid switch-off of TG expression ( within 1 d ) in many tissues and organs [69–71]; however , after 1 mo of treatment , clearance of the antibiotic can take as long as 6 wk [71] . Consequently , we tested lower Dox concentrations in the range of 10 ng/ml to 100 μg/ml and found that 10 μg/ml Dox was optimal . Similar to 2 mg/ml Dox treatment , the ClockΔ19-HA TG was rapidly turned off at 10 μg/ml ( Figure 4 ) ; double transgenic mice showed a shortening of their circadian period of about 1 h , while single transgenic or WT mice showed no effect of Dox treatment . No difference in circadian period was observed among the genotypes with 10 μg/ml Dox treatment ( F3 , 62 = 1 . 93 , p = 0 . 1339 ) . More important , rapid reversal ( switching-on ) was achieved when the double transgenic mice were returned to water; their period lengthened by approximately 1 h within one or two cycles ( paired t-test , t22 = −17 . 2177 , p < 0 . 00005 ) . In addition , in situ hybridization revealed that ClockΔ19-HA transcript in the SCN and throughout the brain is sensitive to the presence of 10 μg/ml Dox . By day 3 of treatment , the transcript is not detectable . Therefore , a dose of 10 μg/ml Dox in the drinking water can easily cross the blood-brain barrier and is sufficient to rapidly switch expression of the TG on and off in the brain . In addition , we found that 100 μg/ml tetracycline in the drinking water also worked effectively ( Figure S4 ) . The magnitude and ease by which we can alter circadian wheel running behavior using a low-dose provide flexibility of repeated activation and repression . Importantly , our data argue that the circadian period is regulated through the dynamic and daily expression of Clock and Clock-controlled genes rather than through a static process established during embryonic development . Light remains one of the most well-understood circadian entraining signals and perturbation analyses with light has been exploited to demonstrate functional properties of the circadian clock [72] . The phase-response curve ( PRC ) can be considered a fundamental pacemaker property [67] . Exposure to light early in the night phase shifts the clock so that subsequent cycles begin at a later time; however , exposure to light late in the night advances the circadian clock . These time-dependent responses to light are important for synchronization to environmental light conditions . Another altered behavioral phenotype in ClockΔ19/+ mice is high-amplitude phase-resetting responses to a 6-h light pulse ( type 0 resetting ) compared to WT mice , which exhibit low-amplitude ( type 1 ) phase resetting [73] . In the C57BL/6J inbred strain of mice , a 6-h light pulse near the “break-point” ( the transition from phase delays to phase advances at approximately circadian time [CT] 17 ) produces large phase shifts of about 4 to 6 h; however , Clock heterozygotes display phase shifts of longer than 6 h [73] . In order to assess whether the conditional induction of ClockΔ19-HA TG , restricted to brain/SCN , recapitulates behavioral properties of ClockΔ19/+ heterozygous mice , we chose to characterize the phase shifting response . Within the same animal , when ClockΔ19-HA TG is expressed ( or when the mice are on water ) , their phase shift is significantly larger than when the TG is turned off ( when they are on Dox treatment ) ( Figure 5A ) . The amplitude of phase shift in Scg2::tTA/tetO::ClockΔ19-HA double transgenic mice on Dox treatment is comparable to the phase shift/delay described in WT mice after a 6-h light pulse [73] . During subjective night ( between CT 12 and CT 17 ) , double transgenic mice showed a significantly larger phase delay on water compared to Dox treatment ( Figure 5B ) . The difference between the two treatments ( water versus Dox ) is more clearly demonstrated when the data are presented as a PRC ( Figure 5C ) . Therefore , brain/SCN-specific activation of the ClockΔ19-HA TG affects both the phase resetting , as well as , the period lengthening effects on locomotor activity rhythms , similar to that seen with the original ClockΔ19/+ heterozygous mutation in the whole animal . In our previous report on the rescue of the Clock mutation using bacterial artificial chromosome ( BAC ) TGs , we showed that overexpression of WT Clock allele ( Clockwt ) shortened free-running period length beyond the WT range [66] . Thus , we explored whether a conditional expression of Clockwt , restricted to the SCN and brain , could also generate shortened free-running activity rhythms as observed in transgenic mice that overexpress CLOCKwt ubiquitously . We have crossed one of the tetO::Clockwt-HA target lines to the Scg2::tTA activator line , and progeny from the mating were examined for their locomotor activity rhythm . We find that the Scg2::tTA/tetO::Clockwt-HA double transgenic mice exhibit a shortened circadian period by about 1 h compared to the WT circadian period ( Figure 6 ) . In addition , similar to transcriptional activation of the ClockΔ19-HA TG , we demonstrate temporal control of Clockwt-HA TG expression with rapid switch-off and a rapid reversal . Therefore , conditional expression of the dominant-negative mutant allele , as well as overexpression of the WT allele , restricted to the SCN and brain , is sufficient to alter the period of activity rhythms . Given that the peripheral circadian oscillators in these mice are WT , it will be interesting in future experiments to investigate whether the period of peripheral oscillators in these mice can be regulated as a consequence of period changes induced by the SCN/brain as previously suggested in experiments using transplanted mouse embryonic fibroblasts [20] .
Significant progress has been made in unraveling the molecular mechanism underlying the mammalian circadian system . The core molecular circuitry of opposing interlocking transcriptional feedback loops has been defined as the fundamental basis of the circadian clock [2 , 74]; however , with subsequent discovery of additional molecular components of the circuitry [75–78] , the complexity and network intricacy of the clock system are becoming apparent [2 , 11 , 50 , 79] . Ultimately , we want to understand how these cell-autonomous circadian oscillators interact in multicellular organisms to regulate physiology and behavior [80] . Thus , to elucidate the mechanisms governing the hierarchical nature of mammalian circadian timing , it is necessary to develop genetic tools to manipulate circadian genes in a conditional and tissue-specific manner in vivo . We adapted an in vivo transgenic method , the tTA system , to regulate Clock gene expression in a conditional and reversible manner . This is a significant technical milestone that has not been previously demonstrated in the mammalian circadian system . In this report , we generated an SCN- and brain-enriched tTA-expressing transgenic line , which can transactivate any transcript of choice when crossed to tetO-responsive transgenic lines . We also produced target lines that can express a ClockΔ19 mutant allele or Clockwt allele in a tissue-specific manner when crossed to a transactivator transgenic line . The ClockΔ19 allele is a dominant-negative mutation ( antimorph ) [56] and is an ideal allele to validate the tTA system for circadian experiments for several reasons . First , the Clock gene , with its heterodimeric partner Bmal1/Mop3 , is one of the primary transcriptional activators of the circadian transcriptional autoregulatory feedback loop . Second , the ClockΔ19 mutation in mice perturbs the circadian system and causes a significantly altered running-wheel rhythm that is clearly distinguishable from WT [56 , 57 , 65 , 66] . The antimorphic nature of this mutation , which has a phenotypic effect that is antagonistic with the normal allele , allows manifestation of the mutant phenotype in the presence of the WT allele . Third , we have previously demonstrated not only that transgenic expression of WT Clock can completely rescue circadian period and amplitude in Clock mutant mice but also that overexpression of the Clockwt TG ubiquitously results in period shortening beyond the normal WT values [66] . Fourth , Clock gene expression in the SCN is constitutive , which is easier to mimic with tTA system . Finally , all circadian clock mutants that have been analyzed in mammals thus far have been germline mutations ( either ENU-induced or gene targeted ) , and therefore the developmental consequences of these mutations have not been addressed . For all of these reasons , we set out to test conditional expression of Clock on circadian behavior . Interestingly , the circadian periodicity of Scg2::tTA/tetO::ClockΔ19-HA double transgenic mice provides a real-time readout of the transactivation state of this system . This allowed us to follow the kinetics of Dox regulation on a day-to-day basis and revealed that high doses of Dox ( 2 mg/ml ) required many months of time to washout and reverse . This then gave us the opportunity to find optimal Dox dose treatments for both inactivation and reversal of tet-dependent transactivation in the brain of mice . SCN-directed expression of the dominant-negative ClockΔ19 TG lengthens the circadian free-running period , whereas expression of the Clockwt TG shortens the circadian rhythm . Furthermore , we showed that temporal and spatial control of TG expression can revert the phenotype from a mutant to a WT state within individuals , and vice versa , with a low concentration of Dox treatment , so that experiments can be performed in a longitudinal fashion . This is akin to transplant experiments [17–19] with the added ability to reverse the procedure . Moreover , unlike transplant experiments , which only allow for the receiving of intact humoral signals , our system also retains intact synaptic connections of the SCN . Finally , the expression of the TG affects not only the free-running activity rhythm but also other expected circadian behavioral responses , such as phase shifts to discrete light pulses . Thus , the transgenic mice described in this report are a valuable tool and will facilitate investigation of the functional relationship between central and peripheral clocks . The validation of the tTA system for conditional TG expression in a variety of cell types and tissues has made it the tool of choice for mammalian system research . Arguably , some of the most significant contributions were made by several groups utilizing the Tet-system in vivo to study the effects of conditional TG activation and repression on various neurobiological process [35–39 , 43 , 45 , 81 , 82] . However , our study differs uniquely from previous Tet system applications in several ways . First , this is the first report on the mammalian circadian system where an SCN/brain-driver is used to conditionally regulate clock genes in vivo . Only a handful of studies reported have used the Tet system in the brain and , thus , the availability of brain-specific drivers is very limited . Furthermore , no drivers have previously been shown to function in the SCN . Second , our study demonstrates that circadian locomotor activity records give us a unique opportunity to have a daily readout of the transactivation state in a noninvasive manner . We suspect that this is likely due to a combination of the shorter half-life of the target protein , CLOCK , and our optimization of the Dox dosage used . Thus , we show that CLOCK is an excellent indicator for the kinetics of Dox-dependent induction/suppression in the brain . Third , we show that the standard dose often used in Tet regulation studies ( e . g . , 2 mg/ml ) is excessively high . This leads to very slow kinetics of washout and slow ( weeks to months ) reactivation of the TG after standard doses of Dox treatment [49] . To date , only one study has examined a time course of TG expression in the brain tissues using luciferase activity as an indicator of the TG expression and has suggested administering a 100-fold lower dose ( 50 μg/ml ) [36] . In our study , we demonstrate that even a 200-fold lower dose of 10 μg/ml in drinking water is sufficient to cross the blood-brain barrier and is equally effective in turning off the TG , and subsequently regulate behavioral state . With a lower dose , our results reveal that the washout is rapid and , thus , multiple induction and suppression cycles of the TG can be achieved within subjects with minimal time loss and cost . Furthermore , the effectiveness of the low Dox dosage is not driver specific . We have found that another SCN- and brain-enriched driver to be inducible and reversible using the same low dose administration ( unpublished data ) . Finally , this study provides an important set of transgenic mouse resources for the circadian research community . Exploitation of these transgenic lines along with existing genetic allelic series of circadian genes may yield fundamental insights into the mechanism by which circadian pacemaker systems transmit information to control physiology and behavior . In addition , by using peripheral tissue-specific drivers , manipulations using the tTA system can yield a wealth of knowledge on physiological processes tied to the circadian machinery such as cell division , heme biosynthesis , tumor suppression , metabolism , and bone remodeling [83–87] . For example , we recently reported a dissection of tissue-specific functions of the mammalian clock protein BMAL1 using the SCN- and brain-enriched driver line , which we describe here , and a muscle-specific driver line . We showed that distinct tissue-specific phenotype in Bmal1-null mice can be rescued using the tTA system [88] . Moreover , tetracycline-dependent genetic tools can also assist in elucidating unexpected subtle phenotypes found in knockout mice of some essential clock genes , such as Rev-erbα and Clock [89 , 90] , and address our current criteria for definition of primary clock components [3] . Besides being potentially useful for such studies , the tTA system may be a great resource for the discovery and in vivo validation of novel candidate genes that may be involved in the central SCN oscillator and the output pathway . The successful manipulation of conditional TG expression in the SCN and brain in these studies will lay the groundwork for the development and adaptation of other tools such as the Cre/Lox system for tissue-specific knockout and conditional inactivation of circadian genes . Furthermore , by developing additional SCN subregional-specific drivers , we can begin to decipher the function of the cellular heterogeneity of the mammalian SCN and to understand how these pacemaking neurons are organized to mediate synchrony within the SCN and the whole animal . The flexibility of the tTA system provides a means to dissect the cellular and behavioral networks in the mammalian circadian system .
For the generation of tTA-expressing mice ( transactivator line ) , a 1 , 153-bp fragment upstream of the translation start site of the mouse Secretogranin II gene was amplified by PCR primers 5′-AGTGATTCCTCTTACTAATCCATCTGTGAGAT-3′ ( forward ) and 5′-GTCTTAAAGATTTCCTGAAAACATAGA-3′ ( reverse ) using a mouse BAC clone RP23-470F8 as a template ( Roswell Park Cancer Institute Mouse BAC Library; Invitrogen , http://www . invitrogen . com ) . The PCR product was first cloned into the EcoRV restriction site in pBluescript II SK ( − ) ( pBlue ) ( Stratagene , http://www . stratagene . com ) . This plasmid was then cut with EcoRI and was used to ligate a 9-kb EcoRI promoter fragment isolated from the BAC clone RP23-470F8 , resulting in an intact 9 , 851-bp promoter fragment upstream of the translation start site . This plasmid was named pBlue-Scg2promoter . A 1 , 453-bp EcoRI-XmnI fragment coding tTA was released from the PMY20 plasmid ( kindly provided by Mark Mayford [35] ) and ligated into the SalI restriction site of the pBlue-Scg2promoter plasmid , after T4 DNA polymerase treatment of both fragments . This plasmid was named Scg2::tTA and was linearized with NotI restriction enzyme prior to microinjection . The linearized fragment was isolated using the QIA Quick Gel Extraction Kit ( Qiagen , http://www . qiagen . com ) and dialyzed for 4 h in 10 mM Tris-HCl ( pH 7 . 5 ) /0 . 1 mM EDTA ( pH 8 . 5 ) buffer before pronuclear injection . For the generation of tetO target lines , we first generated the 3′-HA–tagged WT Clock and mutant ClockΔ19 cDNAs from the plasmids pBlue-ClockWT and pBlue-ClockΔ19 , respectively , as described previously [57] . Each full-length cDNA also contained 388 bp of 5′ untranslated region of Clock , which was generated from exons 1b , 2 , and 3 . The following primers were used against the pBlue-ClockWT or pBlue-ClockΔ19 plasmid to generate cDNAs by PCR using Pfu DNA polymerase ( Stratagene ) : 5′-ATAAGAATGCGGCCGCGGGGAGGAGCGCGGCGGTAGCGGTG-3′ ( forward ) and 5′-CCCAAGCTTCTAAAGAGCGTAATCTGGAACATCGTATGGGTACTGTGGCTGGACCTTGGAAGGGTCA-3′ ( reverse ) . Amplified product was digested with NotI and HindIII , gel purified , and ligated into the NotI-HindIII cloning site of the pTRE2 plasmid ( Clontech , http://www . clontech . com ) . These tetO::Clockwt-HA and tetO::ClockΔ19-HA pTRE2 plasmids were linearized using DrdI/XmnI restriction enzymes to exclude most of the vector backbone sequence prior to pronuclear injection . All plasmids were sequence verified using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit and analyzed on an ABI377 automated sequencer ( Applied Biosystems , http://www . appliedbiosystems . com ) . Transgenic mouse lines were generated by pronuclear injection using standard techniques [91] and essentially as described in Antoch et al . [66] . Briefly , the linearized DNA fragment was injected into fertilized mouse oocytes isolated from crosses of WT CD1 matings at a concentration of 1 ng/μl . Transgenic mice were identified by PCR analysis of genomic DNA prepared from tail biopsy samples . PCR amplification of the transactivator TG ( Scg2::tTA ) was carried out using primers ( forward primer 5′-AGACAAGCTTGATGCAAATGAG-3′; reverse primer 5′-CAAGTGTATGGCCAGATCTCAA-3′ ) that generate a 482-bp fragment . Two Scg2::tTA founder lines were obtained and characterized , and one line was chosen for the experiments presented here . For the tetO::Clockwt-HA and tetO::ClockΔ19-HA TGs , genotyping was performed using 5′-ATATGCAAGGCCAGGTTGTC-3′ ( forward primer ) and 5′-TCTGTGGCATACTGGATGGA-3′ ( reverse primer ) , which generates a 258-bp fragment . We produced eight tetO::ClockΔ19-HA lines and 13 tetO::Clockwt-HA target lines . Each founder animal was maintained as a congenic line by backcrossing to the C57BL/6J inbred strain for at least four generations . All animals were raised in a 12/12-h light–dark cycle from birth . All animal studies were conducted in accordance with the regulation of the Committee on Animal Care and Use at Northwestern University . Doxycycline hydrochloride ( Sigma-Aldrich , http://www . sigmaaldrich . com ) was supplied in the drinking water at a concentration of 2 mg/ml or 10 μg/ml . The Dox-containing water was renewed every 2 to 3 d . Mice were supplied with regular food and water with or without Dox ad libitum . Wheel-running activity of singly housed animals was recorded and analyzed essentially as described [65] . Mice were entrained to a 12/12-h light–dark cycle for a minimum of 7 d before they were released into constant darkness . Activity data were collected as number of events per minute and recorded continuously by a PC system ( Chronobiology Kit; Stanford Software Systems , http://www . query . com ) ; data were analyzed using ClockLab software ( Actimetrics , Wilmette , Illinois , United States ) . The free-running period was calculated ( days 1 to 20 in DD ) by using a χ2 periodogram [92] . For the light pulse experiments , a 6-h light pulse ( approximately 100 lux ) was provided at a given CT , where CT 0 denotes the beginning of the subjective day and CT 12 denotes the beginning of the subjective night . The magnitude of phase shift was determined by measuring the phase difference , based on the activity onset as a phase reference point , between regression lines fit immediately before the light pulse and at least seven consecutive activity-onset times after the light pulse ( excluding the four cycles immediately after the pulse ) as described previously [73] . The magnitude of phase shift was corrected for circadian period as estimated before the light pulse for each individual . All other statistical analyses were performed using the Stata Statistical/Data Analysis software ( version Stata/SE 9 . 0; StataCorp , http://www . statacorp . com ) . Tissues were homogenized in a buffer containing 150 mM NaCl , 50 mM Tris ( pH 7 . 4 ) , 1% Triton X-100 , 0 . 1% SDS , and protease inhibitor cocktail ( Complete; Roche Applied Science , http://www . roche . com ) . Homogenates were cleared by centrifugation at 10 , 000g for 10 min at 4 °C , and supernatants were collected and protein concentration was estimated using Bio-Rad DC Protein Assay ( Bio-Rad , http://www . bio-rad . com ) according to the manufacturer's instructions . Total protein ( 40 μg ) was mixed with sample buffer according to the protocol of [93] and resolved on an 8% SDS–polyacrylamide gel by electrophoresis . Thereafter , proteins were electrotransferred onto a Poly Screen PVDF transfer membrane ( Perkin Elmer Life Sciences , http://www . perkinelmer . com ) . The membranes were blocked with PBST ( PBS + 0 . 1% Tween-20 ) containing 5% powder milk for 1 h and then incubated with the 3F10 mouse monoclonal anti-HA–peroxidase antibody ( 1:500; Roche Applied Science ) or anti–actin-peroxidase antibody ( 1:1 , 000; Santa Cruz Biotechnology , http://scbt . com ) according to the manufacturer's protocol . Anti-CLOCK guinea pig antibody ( 1:1 , 000 ) was followed by anti–guinea pig IgG secondary antisera horseradish peroxidase ( 1:1 , 000; Jackson ImmunoResearch Laboratories , http://www . jacksonimmuno . com ) . CLOCK guinea pig antibody was generously provided by Choogon Lee ( University of Florida , Tallahassee , Florida , United States ) . Proteins were visualized with a chemiluminescence detection system ( ECL Western blotting detection analysis system; Amersham Pharmacia , http://www . amersham . com ) and with subsequent exposure to autoradiographic film . In situ hybridization procedures were performed as described [94] . Briefly , animals were sacrificed by cervical dislocation; the brains were removed immediately , frozen on dry ice , and stored at −80 °C . Sectioning , fixation , hybridization , and washing were performed as described . Sections were hybridized using an antisense-HA oligo probe ( 5′- AAGAGCGTAATCTGGAACATCGTATGGGTACTGTGGCTGG-3′ ) , a WT Clock oligo probe ( 5′-GCTCTAGCTGGTCTTTTAGATGCTGCATGGCTCCTAACTGAGCTG-3′ ) , or an Scg2 oligo probe ( 5′-TTCAGCAGCTCCAGGGCGGAGTTGATCACCTTGGACTTGTCCAGGCGGGACAT-3′ ) . Primers were 5′-end-labeled with [γ-33P]ATP by using recombinant terminal deoxynucleotidyl transferase ( Invitrogen ) . Animals were anesthetized with ketamine/xylazine/saline cocktail ( 10 mg/ml ketamine , 2 mg/ml xylazine; Phoenix Scientific , http://www . psiqv . com ) at 0 . 01 ml/g body weight and then perfused intracardially with 50 ml of 0 . 9% saline solution followed by 50 ml of 4% paraformaldehyde ( Sigma Aldrich ) in 0 . 1 M phosphate buffer ( pH 7 . 2 ) . The brains were removed and postfixed for 24 to 48 h at 4 °C in 4% paraformaldehyde in 0 . 1 M phosphate buffer . For immunocytochemistry , 50-μm coronal sections were collected through the entire SCN using a VibroSlice microtome ( World Precision Instruments , http://www . wpiinc . com ) and processed free floating . Sections were incubated with mouse anti-HA . 11 biotin-labeled monoclonal antibody ( BIOT-101L-100 , 1:200; Covance Research Products , http://www . crpinc . com ) followed by anti-biotin , mouse IgG1 , monoclonal 2F5 conjugated with Alexa Fluor 488 secondary antibody ( 1:1 , 000; Invitrogen ) . A primary antibody to either VIP ( 1:5 , 000; ImmunoStar Inc , http://www . immunostar . com ) or AVP ( 1:10 , 000 ImmunoStar Inc ) was followed by Alexa Fluor 568 goat anti-rabbit IgG ( H+L ) secondary antibody ( 1:1 , 000; Invitrogen ) . The Alexa Fluor 488 signal was assigned a green pseudo-color , while the Alexa Fluor 568 signal was assigned a red pseudo-color . Sections were viewed with a Leica DMRXE7 confocal microscope with Ar ( 488 nm ) and green HeNe ( 543 nm ) lasers in the Biological Imaging Facility at Northwestern University ( Evanston , Illinois , United States ) . Images were captured by sequential excitation with each laser to avoid crosstalk between the two wavelengths using the Leica Confocal Software ( version 2 . 61 build 1537 ) . Mice carrying a tetO promoter-lacZ reporter construct ( lines lac1 and lac2 ) were generously provided by Mark Mayford . All procedures were performed as described by Low-Zeddies and Takahashi [80] . Briefly , mice were anesthetized with ketamine/xylazine/saline cocktail and then transcardially perfused with chilled 0 . 1% heparin PBS . Brains were removed and postfixed for 30 min in 4% paraformaldehyde PBS on ice and then stored overnight in 20% sucrose PBS at 4 °C . Brains were frozen on dry ice , embedded in Shandon Lipshaw M1 embedding matrix ( Pittsburgh , Pennsylvania , United States ) , and sectioned coronally at 50-μm thickness . Free-floating sections were collected in a wash buffer ( PBS with 2 mM MgCl2 , 0 . 02% NP-40 [Sigma] [pH 7 . 3] ) and then incubated for 24 h at 37 °C in an X-gal staining solution ( 1 mg/ml 5-bromo-4-chloro-3-indolyl-β-d-galactoside; Gold Biochemical , Long Beach , New York , United States ) dissolved in dimethyl sulfoxide buffer containing 5 mM K3Fe ( CN ) 6 and 5 mM K4Fe ( CN ) 6 ) . Finally , sections were rinsed three times in the wash buffer and then mounted in aqueous mounting medium ( 3:1 glycerol/PBS ) on gelatin-coated glass slides . Stained sections were viewed and photographed under bright-field illumination .
Accession numbers for genes used in this paper are mouse BAC clone RP23-470F8 ( GenBank [http://www . ncbi . nlm . nih . gov/Genbank] AC084213 ) , Clock ( GenBank AF000998 [mRNA] , AF146793 [genomic] , Entrez Gene ID 12753 , UniProt O08785 ) , Bmal1 ( Arntl ) ( Entrez Gene ID 11865 , UniProt Q9WTL8 ) , Scg2 ( Entrez Gene ID 20254 ) , Avp ( Entrez Gene ID 11998 , UniProt P35455 ) , and Vip ( Entrez Gene ID 22353 , UniProt P32648 ) .
|
Although significant progress has been made in unraveling the molecular mechanism of circadian clocks in mammals , previous work has focused on germline mutations and in vitro methods for analysis . To address the function of clock genes , it is necessary to develop tools to manipulate circadian genes in a conditional and tissue-specific manner in vivo . We report such an approach using the tetracycline transactivator system . Despite the development of the “tet” system in transgenic mice over 10 y ago by Bujard and colleagues , there are still relatively few examples of the successful use of the tet system in the central nervous system . Transgenic expression of the Clock gene in the suprachiasmatic nucleus and brain of mice regulated the period length of circadian locomotor rhythms . These effects could be inhibited by low doses of doxycycline in the drinking water . Importantly , low , but not high , doses of doxycycline were completely reversible and led to a rapid reactivation of the Clock transgenes . In summary , the tetracycline-controlled transactivator system can manipulate circadian clock gene expression in a tissue-specific , conditional , and reversible manner in the central nervous system . The specific methods developed here should have general applicability for the study of brain and behavior in the mouse .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"biochemistry",
"physiology",
"mus",
"(mouse)",
"neuroscience",
"genetics",
"and",
"genomics"
] |
2007
|
Inducible and Reversible Clock Gene Expression in Brain Using the tTA System for the Study of Circadian Behavior
|
Ace is an adhesin to collagen from Enterococcus faecalis expressed conditionally after growth in serum or in the presence of collagen . Here , we generated an ace deletion mutant and showed that it was significantly attenuated versus wild-type OG1RF in a mixed infection rat endocarditis model ( P<0 . 0001 ) , while no differences were observed in a peritonitis model . Complemented OG1RFΔace ( pAT392::ace ) enhanced early ( 4 h ) heart valve colonization versus OG1RFΔace ( pAT392 ) ( P = 0 . 0418 ) , suggesting that Ace expression is important for early attachment . By flow cytometry using specific anti-recombinant Ace ( rAce ) immunoglobulins ( Igs ) , we showed in vivo expression of Ace by OG1RF cells obtained directly from infected vegetations , consistent with our previous finding of anti-Ace antibodies in E . faecalis endocarditis patient sera . Finally , rats actively immunized against rAce were less susceptible to infection by OG1RF than non-immunized ( P = 0 . 0004 ) or sham-immunized ( P = 0 . 0475 ) by CFU counts . Similarly , animals given specific anti-rAce Igs were less likely to develop E . faecalis endocarditis ( P = 0 . 0001 ) and showed fewer CFU in vegetations ( P = 0 . 0146 ) . In conclusion , we have shown for the first time that Ace is involved in pathogenesis of , and is useful for protection against , E . faecalis experimental endocarditis .
Enterococci are gram-positive cocci of intestinal origin first reported as a cause of infective endocarditis ( IE ) in 1899 [1] . They were recognized as the 3rd most common cause of IE as early as the 1920's , and have remained the 3rd most common cause of community onset IE since then with Enterococcus faecalis accounting for >90% of isolates from enterococcal IE when identified to the species level [1] , [2] , [3] , [4] , [5] . Over the past 20 years , enterococci have also become the 2nd–3rd most common organisms isolated from nosocomial ( healthcare-associated ) infections including UTIs , bacteremia , intraabdominal and wound infections , endocarditis , sepsis in neonates , among others [1] , [3] . Indeed , among causes of endocarditis , enterococci ( predominantly E . faecalis ) have been variably reported as the #1 and #2 cause [6] , [7] . Since healthcare-associated infections , particularly those caused by antibiotic resistant bacteria , result in enormous increases in hospital stays and costs , enterococci clearly represent an important drain on healthcare dollars . In one study , the attributable mortality of enterococcal bacteremia was 31% [8] , emphasizing the clinical , not just the financial , seriousness of these infections . The first step in infective endocarditis is vascular tissue colonization , which can be mediated by cell-wall anchored adhesins such as MSCRAMMs ( for microbial surface components recognizing adhesive matrix molecules ) [9] of gram-positive bacteria . Our previous in silico analyses of the E . faecalis genome identified a family of genes encoding MSCRAMM-like proteins containing one or more regions of ca . 150 aa segments with deviant Ig-like fold ( s ) , characteristic of the Staphylococcus aureus MSCRAMMs [10] . One of these , called Ace ( for Adhesin to collagen of E . faecalis ) , has been studied in detail . Genetic and biochemical analyses showed that Ace mediates adherence of E . faecalis cells to bovine and rat collagen type I ( CI ) , human collagen type IV ( CIV ) , and mouse laminin [11] , [12] , [13] , as well as human dentin [14] . Crystal structure analysis of the ligand-binding segment of Ace showed that the Ace A domain is composed of two sub-domains , N1 and N2 , each adopting an Ig-like fold [15] . Subsequent point and truncation mutation analyses suggested that Ace binds to collagen by a mechanism called the “Collagen Hug”[15] , a variant of the “Dock , Lock and Latch” ligand-binding mechanism shown for Staphylococcus epidermidis fibrinogen ( Fg ) adhesin SdrG [16] , [17] . The ace gene is ubiquitous [18] in E . faecalis and conserved among diverse isolates albeit with at least four variants due to variation in the number of repeats of the B domain [19] . Conditional in vitro production of Ace ( i . e . , markedly enhanced production after growth at 46°C , growth in brain-heart infusion plus 40% serum ( BHIS ) or growth in the presence of collagen versus growth in BHI broth at 37°C ) by different strains correlates with conditional adherence of these E . faecalis strains to collagens and laminin [19] , [20] . Most sera from patients with E . faecalis IE show reactivity with rAce , indicating that different strains express Ace during human infection and that it is antigenic in vivo [19] . Furthermore , anti-Ace antibodies ( affinity purified from human serum or animals immunized with rAce ) were shown to inhibit in vitro adherence of E . faecalis strains to collagen and laminin [11] , [19] . In a recent study , anti-Ace40 ( ligand-binding A-domain of Ace ) monoclonal antibodies were shown to completely inhibit binding of Ace40 to human CI and collagen type VI and inhibited binding of Ace-coated fluorescent beads to epithelial cell lines , thus suggesting Ace as a potential therapeutic target antigen against E . faecalis infections [21] . In the present study , we have studied the role of Ace in the pathogenesis of E . faecalis endocarditis by generating an ace deletion mutant in E . faecalis strain OG1RF , by complementing this mutant ( OG1RFΔace ) , by comparing these isogenic strains with OG1RF for adherence to various extracellular matrix ( ECM ) proteins and for their ability to infect aortic valves in a rat endocarditis model . Finally , we also determined the importance of Ace as a protective antigen against experimental endocarditis in a rat model by using active and passive immunization .
Our previous disruption mutant of ace was found to be unstable in vivo ( see below ) . We therefore constructed an allelic replacement ace deletion mutant of OG1RF ( TX5467 , OG1RFΔace::cat; resistant to chloramphenicol 10 µg/ml ) . Deletion of ace from OG1RF was verified by sequencing confirming the correct deletion of ace from −23 to + 2200 ( including the RBS , complete ace gene , and 34 bp downstream of ace ) , and by pulsed field gel electrophoresis ( PFGE ) and hybridizations ( Table 1 ) . Growth ( OD600 ) of the Δace mutant was similar to wild-type ( WT ) OG1RF in BHI ( data not shown ) . We have previously shown , using western blotting and RT-PCR , that ace is expressed at higher levels when grown in BHIS at 37°C or in BHI at 46°C [11] than in BHI at 37°C . Here , we assessed surface localization of Ace in OG1RF at 10 h using flow cytometry analyses with affinity-purified anti-rAce Igs . The mean fluorescence intensity levels for different culture conditions increased progressively with cells grown in BHI at 37°C , BHIS at 37°C and BHI at 46°C ( Figure 1A ) , consistent with our previous western and immunofluorescence microscopy data [11] , [20] . The percentages ( % ) of Ace-expressing cells in BHIS cultures of OG1RF , OG1RFΔace , OG1RFΔace ( pAT392 ) ( empty vector control ) , and OG1RFΔace ( pAT392::ace ) ( complementation ) were >70% , <5% , <5% , and >90% , respectively , demonstrating the inability of OG1RFΔace to produce Ace and the efficient complementation of OG1RFΔace by pAT392::ace . In these experiments , pAT392-containing strains were grown without added gentamicin , the same conditions that we used for preparing inocula for the rat endocarditis experiments . When BHIS was supplemented with gentamicin , expression of Ace increased to >95% of cells in OG1RFΔace ( pAT392::ace ) ( Figure 1B ) , likely due to improved plasmid stability ( see below ) . OG1RF and its isogenic Δace mutant as well as the complementation constructs were tested for their ability to adhere to immobilized ECM proteins and BSA . Consistent with our previous demonstration of adherence of OG1RF to CI , CIV and Fg after growth in BHIS at 37°C and to CI and CIV after growth in BHI at 46°C ( but not in BHI [11] at 37°C ) , we observed here that OG1RF adhered to CI and CIV when grown in BHIS at 37°C , unlike OG1RFΔace which showed markedly reduced adherence to CI ( from ∼36 to 15% ) ( Figure 2A ) , and CIV ( 43 to 3% ) ( Figure 2B ) , but no change in adherence to Fg ( ( Figure 2C ) . This corroborates our earlier data with a mutant with an insertional disruption of ace [11] , [20] . Introduction of the ace gene in trans into OG1RFΔace resulted in even greater adherence to collagens than WT ( >1 . 5-fold higher ) , whereas OG1RFΔace electroporated with pAT392 retained its reduced adherence phenotype ( Figure 2 ) ; these results are consistent with Ace expression data from flow cytometry analysis . To determine if Ace is produced during infection , we performed flow cytometry analyses on extracts directly processed from IE vegetations infected with OG1RF grown in BHI at 37°C . Forward and side scatter pattern analyses of particles from processed vegetations and comparisons with those from in vitro grown OG1RF cells indicated that most of the gated particles detected by flow cytometry are likely OG1RF bacterial cells , thus confirming the removal of the majority of host tissue particles from the vegetations during the processing steps described in methods . Sterile processed vegetations from non-infected rats probed with anti-rAce-specific Igs ( negative control ) showed labeling of a minor fraction ( <3% ) of bacterium-sized particles ( Figure 3A ) , while processed vegetations from OG1RF infected rats probed with Igs from an antiserum raised against formalin-killed E . faecalis strain HH22-whole-cells ( positive control ) bound 85% of bacterium-sized particles , further indicating that the majority of these particles were E . faecalis cells ( Figure 3A ) . Affinity-purified anti-rAce-specific Igs bound to ∼40 to 45% of bacterium-sized particles from different rat endocarditis vegetations infected with OG1RF ( Figure 3 ) , demonstrating that Ace is actively expressed in host vegetations during IE . Although our initial mixed-infection competition experiments showed a clear advantage for the WT over an ace disruption mutant TX5256 [11] , [20] to develop IE in rat model ( data not shown ) , subsequent experiments identified instability of this single cross-over ace disruption mutant during in vivo growth . Hence , we generated an OG1RFΔace mutant for further in vivo testing . In an initial mono-infection experiment ( n = 2 ) with our ace deletion mutant ( OG1RFΔace ) , TX5467 , we first determined the expression of cat ( encoding chloramphenicol acetyl transferase ) in OG1RFΔace , which carries this chloramphenicol resistance marker gene in place of ace , by analyzing individual colonies for chloramphenicol resistance and by high stringency hybridization using intragenic cat and ace DNA probes . We found that ∼10% of the colonies recovered from vegetations were chloramphenicol ( 10 µg/ml ) susceptible even though they were cat probe positive and ace probe negative , indicating that , although the cat gene was present , silencing of its expression had occurred in these colonies . Previously , it has been shown that some antibiotic resistance genes of Escherichia coli are silenced in vivo; specifically , expression of an intact antibiotic resistance gene was switched off during the course of gut colonization in pigs , a phenomenon suggested to be helpful for bacterial fitness [22] . Therefore , for our mixed infection animal experiments , all the results reported here are based on high stringency hybridization with ace and cat probes using ∼200 CFUs/rat vegetation for all the rats used . Of interest , we also tested for chloramphenicol resistance but did not observe further cat silencing . In the mixed infection competition assay , all 12 rats were infected with an approximately 1∶1 mixture ( as predicted by OD600 ) of BHI-grown OG1RF ( determined geometric mean ( GM ) CFU 3 . 8×107/rat , representing 47% of the inoculum ) : OG1RFΔace ( GM CFU 4 . 4×107/rat , representing 53% of the inoculum ) ( Figure 4 ) . Bacterial CFUs from vegetations on aortic valves were recovered at ∼72 h from all 12 rats and are shown in Figure 4 . The mean percentage ( % ) of OG1RF in the total CFU of bacteria recovered was 81 . 5% versus 18 . 5% for OG1RFΔace ( P<0 . 0001 ) , thus demonstrating a clear advantage of OG1RF versus OG1RFΔace at 72 h for heart valve colonization in rats . The mean virulence index or competitive index [23] , [24] , which is a sensitive measure of the relative degree of virulence attenuation of a particular mutant in a mixed infection with the WT strain , was calculated using the equation shown in Materials and Methods . The mean virulence index of the ace mutant relative to WT in vegetations was 0 . 077; this indicates that ace has an important role in this endovascular infection . In initial mono-infection experiments with complementation constructs and testing 24 h after inoculation , we observed loss of the plasmid from cells recovered from vegetations , with 94%-98% loss from OG1RFΔace ( pAT392::ace ) ( 7 rats ) and 14%–100% loss from OG1RFΔace ( pAT392 ) ( 8 rats ) ( data not shown ) . We also tried growing both constructs in BHIS supplemented with gentamicin for the preparation of inocula for infection , but in vivo loss of the plasmid still occurred 24 h after inoculation . To minimize in vivo growth time and to determine the role of Ace in the early stage of valve colonization in rats , we tested both OG1RFΔace ( pAT392::ace ) and OG1RFΔace ( pAT392 ) in the rat model 4 h after inoculation . Two independent mono-infection experiments were done and the combined results are shown in Figure 5 . Rats inoculated with OG1RFΔace ( pAT392::ace ) ( n = 12 ) showed 1 . 4±0 . 6 log10 more CFU/gm than OG1RFΔace ( pAT392 ) ( n = 11 ) ( P = 0 . 0417 ) ( Figure 5 ) , thus demonstrating that Ace has a significant role in early colonization of heart valves in E . faecalis rat IE . Reduced time in vivo also resulted in much less loss of the plasmid from each construct . In the case of OG1RFΔace ( pAT392::ace ) , gentamicin susceptible colonies were recovered in only 2/12 rats ( 2/2 colonies from one and 3/3 colonies from the other were gentamicin susceptible ) , while with OG1RFΔace ( pAT392 ) , 3–100% of colonies ( among 8–48 tested ) recovered from 5/11 rats were gentamicin susceptible . These results corroborated the above described complementation of ace surface expression ( Figure 1B ) and restoration of in vitro adherence of OG1RFΔace ( pAT392::ace ) to CI and CIV to similar levels as observed for OG1RF . Since Ace was found to be an important virulence factor in rat experimental IE , an Ace-specific immune response might hinder the development of IE . To study this , rats were vaccinated s . c . thrice with 99% pure 100 µg rAce or PBS or Freund's complete adjuvant – Freund's incomplete adjuvant ( FCA – FICA ) and were challenged with 107 to 109 CFU of E . faecalis OG1RF per rat ( see methods ) . Comparison of anti-Ace antibody levels of 10 immunized and three non-immunized animals by ELISA showed that all 10 immunized rats tested had high levels of anti-Ace titers ( 1: >50 , 000 ) , whereas no anti-Ace antibodies were detected in any of the three control rats ( Figure 6 ) . Sixteen of 16 no-treatment control rats ( 100% ) developed E . faecalis endocarditis after challenge with 10 times the ID50 of BHI-grown OG1RF compared with 5 of 14 rats ( 35% ) in the rAce active-immunization group ( P = 0 . 0001 ) ( Figure 7A ) . The no-treatment control rats showed a mean of 4 . 2±1 . 0 log10 more CFU/gm than the rAce active-immunized ( P = 0 . 0004 ) in vegetations recovered from heart valves . In an independent experiment , in order to mimic in vivo growth conditions more closely and because we had found that Ace is expressed at higher levels by OG1RF when grown in BHIS [11] , we used BHIS-grown OG1RF for the preparation of inocula . rAce ( n = 10 ) rats were inoculated with a higher inoculum of 1 . 4×109 CFU/rat ( ∼100 times the ID50 ) , while non-immunized controls ( n = 18 ) were inoculated with 3 . 1×108 –1 . 1×109 CFU/rat . Fifteen of 18 non-immunized control animal ( 83% ) developed E . faecalis endocarditis compared with 5 of 10 rats ( 50% ) in the rAce-immunized group with a mean ± standard deviation ( SD ) increase of 3 . 2±1 . 3 log10 CFU ( P = 0 . 0231 ) in the non-immunized group ( Figure 7B ) , showing significantly reduced bacterial titers in rAce-immunized group even when given a higher inoculum . Thirty-three sham ( FCA - FICA ) -immunized control group rats inoculated with 1 . 2–3 . 8×108 CFU/rat also showed significantly higher bacterial counts with a mean ± SD increase of 2 . 0±1 . 0 log10 CFU/gm ( P = 0 . 0475 ) versus the rAce active-immunized treatment group ( n = 10 ) ( Figure 7B ) . These results demonstrate reproducibility of in vivo protection by active immunization with rAce against E . faecalis experimental endocarditis in rats using two different growth conditions to prepare the inocula . Five of 6 ( 83% ) of rats administered purified control Igs ( 2 mg/kg ) from pre-immune serum 1 h prior to inoculation of OG1RF developed E . faecalis IE compared with 2/10 rats ( 20% ) of the group given anti-rAce Igs ( 2 mg/kg ) affinity-purified from immune serum ( P = 0 . 0001 ) ( Figure 8 ) . Mean bacterial titers recovered from control rat aortic vegetations showed 3 . 8±1 . 4 log10 more CFU/gm than the group given anti-rAce Igs ( P = 0 . 0146 ) ( Figure 8 ) . Thus , these results corroborated the protection results seen above in active immunization using rAce against E . faecalis IE in rats . In this in vivo model , both OG1RF and OG1RFΔace caused animal mortality at similar rates with all the inocula tested ( data with two inocula were shown in Figure 9 ) showing that OG1RFΔace was not attenuated versus OG1RF .
Infective endocarditis , which affects the endothelial lining of the heart , is among the most severe of the wide range of enterococcal infections encountered in humans , presenting a major therapeutic challenge and resulting in considerable mortality even when treated with antibiotics [5] , [7] , [8] , [25] . Development of endocarditis can be initiated by injury to the valvular endothelium , which disrupts the normal valve structure and exposes underlying tissues , including ECM material . Deposition of host proteins , such as fibrin , as well as platelets at the site of injury then leads to the formation of a sterile thrombotic vegetation . This endovascular lesion may become colonized by circulating bacteria , leading to the growth of an infected vegetation . Valvular and aortic tissues are rich in collagen [26] , and collagen is also found in sterile vegetations [26] . Previous studies have demonstrated that Ace plays a major role in the in vitro adherence of E . faecalis isolates to immobilized collagen [11] , [12] , [13] . Therefore , we reasoned that collagen could serve as a potential adhesion target for enterococci during bacteremia and that Ace could mediate bacterial attachment to these collagen-containing sites . To date , no studies have demonstrated a role for Ace in endocarditis and only very recently has a report appeared showing that Ace is important in a murine urinary tract infection model [27] . However , our previous demonstration of the role of host-derived cues ( i . e . , using moieties typically encountered in the host , such as serum or collagen [19] , [20] ) , for induction of both Ace expression and adherence of E . faecalis cells to collagen , suggested that both of these phenotypes are elicited by close association of this organism with a mammalian host/tissue . Moreover , we found that 90% of patients with prior E . faecalis endocarditis have Ace-specific antibodies in their sera , implying that Ace is expressed in vivo during the infection and that it is immunogenic [19] . For these reasons , we chose an experimental IE model in this study to explore the role of Ace in E . faecalis pathogenesis . We first looked for direct evidence showing that Ace is expressed during E . faecalis infection . E . faecalis OG1RF cells recovered directly from infected vegetations showed surface expression of Ace with a much higher ( ∼3× ) mean fluorescence intensity compared to cells grown in vitro in BHI at 37 °C , indicating that the host environment in vegetations , similar to collagen and serum [19] , [20] , can induce production of Ace and its localization on the cell surface . As anticipated , in vitro ECM protein adherence results with OG1RFΔace corroborated our previous results with an insertionally inactivated ace [11] , [20] . Collagen adherence of the ace deletion mutant was restored by complementation in trans and the adherence of the complemented strain was 1 . 5-fold above the level of the WT parent strain , likely due to the higher number of Ace molecules displayed on the surface of the complemented strain , as determined by flow cytometry . Deletion of ace resulted in significant attenuation in the ability of the mutated E . faecalis OG1RF strain to compete successfully with its isogenic WT parent in infection of vegetations in a mixed-inoculum rat IE model . To the best of our knowledge , this is the first demonstration that Ace contributes to E . faecalis virulence in endocarditis . When we complemented the ace mutant in trans , significantly more colonization of heart valves was observed at 4 h after infection by this strain than by an isogenic strain containing an empty vector . Thus , these results confirmed our 72 h results with the ace deletion mutant and , furthermore , provide evidence that Ace plays an important role during the initial attachment and colonization stage of IE development , possibly by mediating adherence of E . faecalis cells to exposed collagen at the site of endovascular injury . This is consistent with the high surface expression levels of Ace in the complemented strain shown by our flow cytometry analysis . While stably maintained by the majority of E . faecalis cells during early colonization ( 4 h ) , the high instability of the complementation vector after extended growth in vegetations ( 94–98% of cells had lost the plasmid by 24 h ) reduced its effect at later stages of endocarditis . The residual ability of OG1RFΔace to cause endocarditis in some rats indicates that Ace is not absolutely required for E . faecalis to cause endocarditis; this is in agreement with published studies that showed a role for additional factors in causing E . faecalis IE [28] , [29] , [30] . While the precise mechanism of action of Ace for initiating , maintaining and/or propagating IE has yet to be elucidated , we infer that the difference in virulence of OG1RFΔace may be due to its reduced ability to adhere to collagen . However , we cannot exclude the possibility of another ligand or another function of this protein . Interestingly , deletion of ace did not result in observable effects in the mouse peritonitis model in terms of either time to death or total mortality , suggesting that ace is not important for this infection or that the direct administration of a large inoculum of bacteria into the peritoneal cavity may bypass an early infection stage where Ace might be involved . These results also indicate that deletion of ace did not affect growth or survival of OG1RF in vivo in general , consistent with the similar growth rate and viability of the ace deletion mutant and WT when grown in vitro . We have recently shown that Acm , a collagen adhesin from E . faecium , is an important factor for endocarditis caused by that species [23]; this is similar to a previous observation with Cna of S . aureus [31] , which is also involved in other infections , such as septic arthritis [32] . These MSCRAMMs share a large degree of sequence conservation in their collagen-binding domains; similar proteins are present in several other species of gram-positive pathogens , such as Streptococcus equi [33] , Arcanobacterium pyogenes [34] , Bacillus anthracis [35] and Streptococcus gallolyticus [36] , and they possibly share a similar collagen-binding mechanism , called the “Collagen Hug” that has been characterized for Cna and Ace [15] , [37] . Therefore , it seems plausible that this family of proteins has been preserved or acquired across different gram-positive species/genera as a generalized mechanism to provide a binding function , although the ligand in the ecological niches where enterococci are found in nature and the purpose for these adhesins is not known . Recently , we described Ebp pili as another important factor for E . faecalis endocarditis [29] , as well as urinary tract infections and biofilm formation [29] , [38] , a further indication of the significance of surface proteins of the MSCRAMM family for E . faecalis pathogenesis . Our results with active immunization of rats using the collagen-binding domain of Ace showed that only 25% of immunized rats developed endocarditis , while the infection rate in the untreated group was 100% . Protection was also evident when bacterial counts were evaluated . Consistent with these results , prophylactic treatment of rats with affinity-purified anti-Ace antibodies raised against the collagen-binding domain of Ace significantly reduced bacterial numbers in vegetations , demonstrating that passive transfer of Ace-specific antibodies confers significant protection against E . faecalis IE in rat . The differences in pre-infection procedures between the active- and passive-immunized groups preclude direct comparison of results from these two methods . Based on the results presented in this study , it seems likely that these preventive strategies specifically target the initial attachment and colonization stage of endocarditis by blocking collagen adherence of E . faecalis cells . Consistent with this hypothesis , we have previously shown that Ace-specific polyclonal antibodies purified from immunized rabbits or from humans with a prior E . faecalis endocarditis infection were effective in inhibiting adherence of Ace-expressing E . faecalis isolates to collagen [11] , [19] . Furthermore , a recent study that generated monoclonal antibodies against rAce showed that some of the mAbs completely inhibited binding of rAce to collagen and Ace-coated fluorescent beads to epithelial cell lines [21] . The ace gene is ubiquitously present among isolates of E . faecalis and its encoded amino acid sequence , especially within the collagen-binding domain , is highly conserved [19] . Therefore , targeting ace could potentially offer protective immunization against a large spectrum of genetically diverse E . faecalis isolates , an advantage over other virulence-associated factors , such as aggregation substance , hemolysin and gelatinase , which were found to be produced by <45% of endocarditis isolates [39] and for which protective efficacy has not been shown [40] . So far , only one E . faecalis antigen , the capsular polysaccharide , has shown promise as a potential vaccine candidate , as passive and active immunization against it lowered bacterial counts in kidneys , spleens and livers in a mouse i . v . infection model [41] . To our knowledge , our study is the first report of an immunization strategy that reduces E . faecalis colonization of aortic valves and shows protection against the development of E . faecalis endocarditis , thus , suggesting Ace as a promising alternative target for prophylaxis of E . faecalis endocarditis in high risk patients . However , the ability of OG1RF to cause IE in some of the rAce-immunized rats and also in some anti-Ace antibody-treated rats indicates that targeting multiple MSCRAMMs may be necessary for a robust protection of E . faecalis IE . Consistent with this , a recent study [42] showed full vaccine protection against abscess formation or lethal challenge with S . aureus strains when a combination of four MSCRAMM antigens were used versus a moderate reduction in bacterial load when used as individual vaccine antigens . In summary , we have demonstrated here that i ) deletion of the ace gene resulted in significant attenuation of the ability of E . faecalis to colonize aortic valves and cause endocarditis in an experimental rat IE model , coinciding ( ii ) with reduced in vitro adherence by the ace deletion mutant to collagen types I and IV; we have also shown that ( iii ) Ace is actively expressed within host vegetations during endocarditis and that ( iv ) both active and passive immunization against the collagen-binding domain of Ace conferred significant protection against endocarditis and reduced the numbers of bacteria found in vegetations . Taken together , these results demonstrate that Ace is an important virulence-associated factor and a likely target for prophylactic and therapeutic strategies against E . faecalis endocarditis . Since Ace-like proteins are widespread among streptococci and staphylococci , future cross-protection studies may reveal novel opportunities for the development of vaccines or immunotherapeutics that may be useful for the prevention and treatment of gram-positive infective endocarditis .
The rat endocarditis model and surgical procedures were performed in accordance with the institutional policies and the guidelines stipulated by the animal welfare committee , University of Texas Health Science Center at Houston ( AWC , UTHSC ) . This study was reviewed and approved by the University Institutional Review Board ( AWC approval # HSC-AWC-08-067 ) . E . coli and E . faecalis strains and all plasmids used in this study are listed in Table 1 . All constructs were given TX numbers and plasmids from these constructs were assigned respective pTEX numbers ( Table 1 ) . E . coli strains were grown in Luria-Bertani media ( Difco Laboratories , Detroit , Mich . ) . Enterococci were grown either in BHI , BHIS , Todd-Hewitt ( TH ) broth/agar ( Difco Laboratories ) or Enterococcosel™ Agar ( EA ) ( Becton Dickinson ) at 37°C , unless a different growth temperature is specified . The following antibiotic concentrations were used with E . faecalis: chloramphenicol 10 µg/ml , kanamycin 2000 µg/ml , rifampicin 100 µg/ml and gentamicin 125 µg/ml . With E . coli , the concentrations used were chloramphenicol 10 µg/ml , kanamycin 50 µg/ml , and gentamicin 25 µg/ml . Resistance of enterococci to gentamicin and kanamycin was defined as MICs >500 and >2000 µg/ml , respectively [18] . All antibiotics were obtained from Sigma ( St . Louis , Mo . ) . Tran 35S-label and bovine serum albumin ( BSA ) were purchased from MP Biomedicals Inc . ( Irvine , Calif . ) . C I and CIV were from Sigma and Fg was from Enzyme Research Laboratories ( South Bend , Ind . ) . Oligonucleotide primers were purchased from Invitrogen ( Carlsbad , Calif . ) or IDT ( Coralville , Iowa ) or Sigma and their sequences are provided in Table 2 . Restriction enzymes and DNA modification enzymes were mostly from Invitrogen and New England BioLabs , Inc . ( Beverly , Mass . ) . All other chemicals used in the investigation were of molecular biology grade . Chromosomal DNA from E . faecalis isolates was prepared following the hexadecyltrimethyl ammonium bromide method described earlier [43] . Plasmid DNA was isolated from E . coli using the Wizard Plus SV minipreps DNA purification system ( Promega Corporation , Madison , Wis . ) and , from E . faecalis , by a previously described method [44] . General recombinant DNA techniques such as ligation and agarose gel electrophoresis were performed using standard methods [45] . When necessary , DNA fragments were purified with low melting temperature agarose gel followed by purification using QIAquick-gel extraction kit ( Qiagen Inc . , Valencia , Calif . ) . PCR reactions were performed with a Perkin-Elmer GeneAmp PCR system 9700 using the optimized buffer B ( 1 × buffer: 60 mM Tris-HCl [pH 8 . 5] , 15 mM ammonium sulfate and 2 mM MgCl2 ) obtained from Invitrogen . PCR-generated fragments were purified using the Wizard PCR DNA Cleanup System ( Promega Corporation ) . Recombinant plasmids were generated in E . coli DH5α or XL1-blue . Electroporation of E . coli and E . faecalis was carried out using a Gene Pulser ( Bio-RAD Laboratories , Richmond , Calif . ) as described previously [46] , [47] . Agarose plugs containing genomic DNA were digested with SmaI and PFGE was performed using a previously described method [1] . Southern blotting was performed using Hybond-N+ nylon membrane and 0 . 4 N sodium hydroxide solution . Preparation of colony lysate blots was described elsewhere [48] . The RadPrime DNA Labeling System ( Invitrogen ) was used for labeling DNA probes with [α-32P] dCTP ( GE Healthcare , Piscataway , N . J . ) and hybridizations were carried out using high stringency conditions [48] , [49] . DNA sequencing reactions were performed by the Taq dye-deoxy terminator method and an automated ABI Prism sequencer ( Applied Biosystems , Foster city , Calif . ) . An E . faecalis ace mutant ( OG1RFΔace::cat ) was constructed by allelic replacement using pTEX5500ts as described earlier for E . faecium [50] . We used a replacement strategy in this study to facilitate distinguishing between WT and OG1RFΔace during in vivo animal experiments with mixed cultures; bioinformatics and mRNA analyses of the ace locus predicts the absence of a polar effect of ace deletion by the cat replacement ( unpublished data ) . E . faecalis OG1RFΔace was constructed by allelic replacement using pTEX5500ts as described earlier for E . faecium [50] . A 1027-bp DNA fragment ( AceDelUp ) encompassing the region upstream of ace was amplified from OG1RF genomic DNA template using primers AceDelF1 and AceDelR1 ( Table 2 ) , digested with BamHI and HindIII , and ligated with similarly digested pTEX5500ts . Similarly , a 989-bp DNA fragment ( AcedelDn ) encompassing the region downstream of ace was amplified from the same genomic DNA template using primers AceDelF2 and AceDelR2 ( Table 2 ) . The AceDelDn PCR product digested with PstI and EcoRI was ligated to similarly digested pTEX5500ts::AceDelUp and was then transformed into E . coli DH5α to obtain TX5428 . The plasmid from this construct , pTEX5428 ( pTEX5500ts::AceDelUp+AceDelDn ) , was introduced into electrocompetent cells of OG1RF and cells were then plated on gentamicin plates at the permissive temperature ( 28°C ) . A single gentamicin and chloramphenicol resistant colony from these plates was grown overnight at 42°C , then plated on chloramphenicol plates and incubated at 37°C . After confirming the specific single crossover integration ( OG1RFaceUp::pTEX5428 ) by PCR ( with primer sets AceUpF1 and CmR as well as AceDnR1 and CmF ) , one of the integrants was picked , grown for eight overnight serial passages at 42°C , and then plated on BHI to select for plasmid excision by double crossover recombination . The colonies from these BHI plates were then replica plated to chloramphenicol plates and gentamicin plates to identify colonies that retained the cat gene but not the vector . To complement OG1RFΔace in trans , an ∼2 kb fragment containing the ace open reading frame plus its ribosome-binding site ( amplified using primers aceComF1 and aceComR1; Table 2 ) was cloned under the control of the P2 promoter of the shuttle vector , pAT392 [51] . This in vitro-ligated construct for complementation ( designated as pTEX5646 ) was transformed into E . coli XL1-Blue to obtain TX5646 and was then introduced into electrocompetent cells of TX5467 to obtain TX5647 ( OG1RFΔace ( pAT392::ace ) . Surface expression of Ace in OG1RFΔace ( pAT392::ace ) was determined by flow cytometry ( see below ) . Overnight cultures were inoculated into BHI broth at a dilution of 1:100 . The cultures were then grown at 37°C with shaking in an orbital shaker and aliquots were removed hourly from 0 to 12 h and at 24 h , for determining the absorbance at 600 nm ( OD600 ) with a spectrophotometer . Adherence of E . faecalis to CI , CIV , Fg and BSA was determined in four independent experiments using Tran 35S-labeled bacteria by a previously described assay [11] . Construction of the recombinant plasmid pTEX5254 ( complete ace A domain of OG1RF cloned into pBAD/HisA expression vector ) was described previously [11] . Expression cultures of TX5254 were induced with arabinose and the N-terminally His6 tagged proteins were purified using nickel affinity chromatography and anion exchange chromatography , as described previously [11] , [52] . Protein concentrations were determined by absorption spectroscopy at 280 nm using calculated molar absorption coefficient values [53] . Expression and purification of ( His ) 6-tagged recombinant Ace A domain was done using a previously described construct and methods [11] . Goat polyclonal serum against recombinant rAce A domain ( rAce ) was generated by Bethyl Laboratories ( Montgomery , TX ) . Ace A-domain specific antibodies were eluted from rAceA coupled to cyanogen bromide-activated Sepharose 4B , according to the manufacturer's protocol ( Amersham Biosciences , Piscataway , N . J . ) . The antibodies were concentrated by ultrafiltration with a 10 , 000-Da molecular-weight-cutoff filter ( Millipore , Bedford , Mass . ) , dialyzed against PBS and concentrations were determined by absorption spectroscopy . Surface protein extracts from E . faecalis isolates were prepared using mutanolysin ( Sigma ) as described earlier [11] . Protein extracts were electrophoresed in 4–12% NuPAGE Bis-Tris gels ( Invitrogen ) under reducing conditions in MOPS buffer , and transferred to a polyvinylidene difluoride ( PVDF ) membrane . Membranes were then incubated with either affinity-purified anti-Ace A-domain specific immunoglobulins ( Igs ) [11] or pre-immune rabbit serum Igs followed by horseradish peroxidase-conjugated anti-goat antibodies . The blots were then developed with Supersignal West Pico Chemiluminescent substrate ( PIERCE , Rockford , Ill . ) . Purified recombinant Ace A-domain was used as a positive control . Anti-Ace antibody titers in rat sera were determined by ELISA as described by [19] with some modifications . Briefly , 96-well plates ( Immulon 4HBX , Thermo Fisher Scientific , Waltham , Mass . ) were coated with 1 μg of rAce in 0 . 05 M carbonate-bicarbonate buffer , pH 9 . 6 . Rat sera were tested in duplicate with serial dilutions from 1∶100 to 1∶240 , 800 , followed by detection with peroxidase-conjugated anti-rat secondary antibody ( Jackson ImmunoResearch Laboratory , West Grove , Pa . ) and TMB peroxidase substrate ( Bethyl Laboratories , Montgomery , Tex . ) . The reaction was stopped with 2 M H2SO4 . Antibody titers were expressed as the highest serum dilution with an A450nm ≥0 . 10 at 3 min after addition of the substrate [55] . Aortic valve endocarditis was produced in rats by following previously published methods [23] , [30] , [55] , [56] , [57] , [58] . In brief , for induction of endocarditis , white Sprague-Dawley rats ( ∼200 gm ) were used . The animals were anesthetized with isoflurane for placement of intravascular catheters . The right carotid artery was exposed and a sterile polyethylene catheter was inserted through a small incision and advanced to ∼4 cm into the left ventricle . Proper positioning was assured by sensing resistance and vigorous pulsation of the line . Randomly picked colonies recovered from vegetations of infected rats were also tested by PFGE to reconfirm the infecting organism [49] . In brief , agarose plugs containing genomic DNA were digested with SmaI and PFGE was performed using a previously described method with ramped pulse times of 5 s and 45 s . OG1RF and OG1RFΔace were tested following our previously published method [60] . In brief , mice were injected intraperitoneally with appropriate dilutions of bacteria ( BHI or BHIS ) , premixed with sterile rat fecal extract ( SRFE ) and were observed for 5 days for survival . Two-fold inocula ( range of ∼1×108–1×109 CFU/ml ) of both test bacteria were used to compare animal survival/mortality . LD50 was determined using six mice per group and by the method of Reed and Muench [59] . To compare the mean ± SD values of the adherence results , an unpaired t test was used . Percentages ( % ) of OG1RF versus OG1RFΔace present in mixed infection vegetations were analyzed by the paired t test . Similar to the method previously described for E . faecalis and E . faecium endocarditis using a mixed infection [23] , [29] , the mean virulence index of the mutant relative to WT was calculated using the following equation: Mean virulence index for the mutant should be 1 . 0 , if the WT and the mutant have the same level of virulence , and lower values would indicate increasing levels of attenuation . Differences in bacterial log10 CFU ( geometric mean ) from vegetations of rAce-immunized versus non-immunized and FCA-FICA-immunized controls were analyzed by the unpaired t test . Fisher's exact test was used for comparing the total number of infected/non-infected rats in the rAce-immunized group versus control groups . Graph Pad Prism version 4 . 00 for Windows ( GraphPad Software , San Diego , Calif . ) was used for statistical analysis .
|
Enterococcus faecalis was recognized as a common cause of infective endocarditis ( IE ) by the early 1900s . It is still third in community-onset IE , but is the second most common cause of hospital-associated IE . Complications due to E . faecalis IE include congestive heart failure , septic emboli and death and current management involves a combination of antimicrobials , often with surgery . Emergence of antimicrobial resistance has created the need for alternative strategies ( such as immunoprophylaxis ) that target in vivo expressed virulence-associated surface proteins . One such E . faecalis protein is Ace , which is antigenic during human IE and mediates attachment of E . faecalis cells to host extracellular matrix proteins collagen and laminin . Using a rat model , we now show that ace contributes to E . faecalis IE pathogenesis and demonstrate that Ace is expressed at high levels during IE even though produced at low levels under laboratory conditions; both active and passive immunization based on the collagen-binding domain of Ace conferred significant protection against IE . These observations , along with data that human antibodies against Ace inhibit collagen adherence of E . faecalis , indicate that Ace is an important virulence-associated factor and a promising target for prophylactic and possibly therapeutic strategies against E . faecalis IE .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/bacterial",
"infections",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"microbiology/immunity",
"to",
"infections"
] |
2010
|
Importance of the Collagen Adhesin Ace in Pathogenesis and Protection against Enterococcus faecalis Experimental Endocarditis
|
Pharmacologic stimulation of innate immune processes represents an attractive strategy to achieve multiple therapeutic outcomes including inhibition of virus replication , boosting antitumor immunity , and enhancing vaccine immunogenicity . In light of this we sought to identify small molecules capable of activating the type I interferon ( IFN ) response by way of the transcription factor IFN regulatory factor 3 ( IRF3 ) . A high throughput in vitro screen yielded 4- ( 2-chloro-6-fluorobenzyl ) -N- ( furan-2-ylmethyl ) -3-oxo-3 , 4-dihydro-2H-benzo[b][1 , 4]thiazine-6-carboxamide ( referred to herein as G10 ) , which was found to trigger IRF3/IFN-associated transcription in human fibroblasts . Further examination of the cellular response to this molecule revealed expression of multiple IRF3-dependent antiviral effector genes as well as type I and III IFN subtypes . This led to the establishment of a cellular state that prevented replication of emerging Alphavirus species including Chikungunya virus , Venezuelan Equine Encephalitis virus , and Sindbis virus . To define cellular proteins essential to elicitation of the antiviral activity by the compound we employed a reverse genetics approach that utilized genome editing via CRISPR/Cas9 technology . This allowed the identification of IRF3 , the IRF3-activating adaptor molecule STING , and the IFN-associated transcription factor STAT1 as required for observed gene induction and antiviral effects . Biochemical analysis indicates that G10 does not bind to STING directly , however . Thus the compound may represent the first synthetic small molecule characterized as an indirect activator of human STING-dependent phenotypes . In vivo stimulation of STING-dependent activity by an unrelated small molecule in a mouse model of Chikungunya virus infection blocked viremia demonstrating that pharmacologic activation of this signaling pathway may represent a feasible strategy for combating emerging Alphaviruses .
The innate immune system includes an array of sentinel proteins termed pattern recognition receptors ( PRRs ) that sense and react to microbe- and danger-associated molecular patterns ( reviewed in [1] ) . These patterns are often constituents or replication intermediates of intracellular ( especially viral ) pathogens . PRRs respond to this engagement by initiating signaling pathways that bring about the expression or processing of cytokines , chemokines , and effector molecules that both directly block microbial replication and facilitate related adaptive immune processes . As such , PRRs represent an essential first line of immunological defense against infection and are the target of both microbial inhibitory phenotypes as well as pharmacologic manipulation for therapeutic purposes ( reviewed in [2] ) . Synthesis and secretion of interferon ( IFN ) proteins is often a primary outcome of PRR-mediated signaling . This includes multiple subtypes of IFNα and β ( type I IFN ) as well as IFN λ1–3 ( type III IFN ) . IFNs act via cognate cell surface receptors by triggering a phosphorylation cascade involving Janus and tyrosine kinases ( Jak1 , Tyk2 ) and signal transducer and activator of transcription 1 and 2 ( STAT1/2 ) transcription factors that amplify the expression of antiviral effector and other immune stimulatory genes conventionally termed IFN-stimulated genes ( ISGs ) . PRR-mediated expression of IFNβ is particularly well characterized and requires phosphorylation of the transcription factor IFN regulatory factor 3 ( IRF3 ) by serine kinases TANK Binding kinase 1 ( TBK1 ) and I Kappa B kinase ε ( IKKε ) [3] . This occurs primarily through pathways that utilize specific adaptor proteins acting as integration points for upstream PRRs . TIR-domain-containing adaptor-inducing IFNβ ( TRIF; also called TICAM1 ) is required for signals initiated by Toll-like receptors ( TLRs ) 3 and 4 [4 , 5] . IFN promoter stimulator 1 ( IPS-1; also called MAVS , VISA , Cardif ) is employed by RIG-I and MDA5 , that both sense cytoplasmic dsRNA [6–9] . Stimulator of IFN genes ( STING; also called MITA , TMEM173 , MPYS , ERIS ) [10–12] is actually both a PRR for cyclic dinucleotides ( CDN ) via a binding pocket in its C-terminal cytoplasmic domain ( CTD ) [13–15] as well as an adaptor molecule for multiple cytoplasmic receptors of dsDNA [16–18] . Given the importance of these pathways for innate immune activation and antimicrobial protection they have been the focus of broad and intense research aimed at both understanding their physiological effects and harnessing their potential for contributions to immune-based therapeutics . Given the ability of the IFN system to render cells and tissues refractory to replication of a wide array of virus types as well as its role in coordinating adaptive immune responses , pharmacologic IFN stimulation has been suggested as a broad spectrum antiviral strategy [2 , 19–22] . Moreover , factors capable of yielding therapeutic effects via activation of IRF3-mediated responses have been identified and biologically validated . This includes agonists of TLRs shown to block replication of some chronic viruses [23–25] as well as enhance vaccine immunity ( reviewed in [26] ) . Similarly , stimulation of the RIG-I/MDA5/IPS-1 by synthetic nucleic acids can be employed for antiviral outcomes against diverse acute viruses [27 , 28] . Intriguingly , two synthetic small molecules , 10-carboxymethyl-9-acridanone ( CMA ) [29] and the chemically unrelated 5 , 6-dimethylxanthenone-4-acetic acid ( DMXAA ) [30] are each capable of activating the STING pathway . Both molecules block multiple , even drug-resistant viruses [29 , 31–33] . Intriguingly , DMXAA exhibits other immunotherapeutic effects including vaccine adjuvanticity [34 , 35] , anti-angiogenic vascular disruption promoting tumor necrosis [36 , 37] , and immune-mediated clearance of solid tumors [38] . Unfortunately , CMA and DMXAA were found to only function in mouse , not human cells and tissues [39–41] and thus were not effective in clinical trials . While analogs of cross-specific stimulatory CDNs have been synthesized [38] , to our knowledge there exists no published biological characterization of novel synthetic molecular entities that activate human STING-dependent innate responses , despite the high and multi-pronged therapeutic potential of exploiting this important immunological protein . Members of the Alphavirus genus include mosquito-transmitted agents that are re-emerging worldwide and can lead to significant morbidity and mortality ( reviewed in [42] ) . Among these is Chikungunya virus ( CHIKV ) , which , despite its evolutionary origin in the Old World , is currently experiencing a severe outbreak in the Caribbean , Central , and South America . Since it first arrived in the Western hemisphere in December 2013 over one million suspected and confirmed cases are estimated to have occurred [43] . CHIKV disease is characterized by severe joint pain that can persist for months to years . Venezuelan Encephalitis virus ( VEEV ) is a related virus belonging to the New World clade that has experienced numerous outbreaks in South and Central America as well as southern Texas [44] . VEEV is a much more deadly agent with fatality rates at approximately 20% but that can reach up to 35% in children ( reviewed in [45] ) . Currently no FDA-approved antiviral drugs or vaccines exist for either virus . Interestingly , however , both viruses are extremely sensitive to type I IFN [46–48] . Moreover , being RNA-based viruses their infection triggers IRF3/IFN activation via the IPS-1 pathway [49] and as such may not exhibit evasion phenotypes directed at the cytoplasmic DNA-based STING pathway . In light of this pharmacologic activation of IRF3/IFN via STING may represent an efficacious therapeutic strategy . Herein we describe the identification and characterization of a novel small molecule capable of stimulating IRF3 phosphorylation and IFN production in human cells that prevents replication of Alphaviruses . Through reverse genetic studies using CRISPR/Cas9-mediated gene editing we also show that this molecule requires STING for its innate gene induction and antiviral activity and thus it represents the first synthetic compound definitively capable of activating this pathway in human cells . Moreover , in vivo stimulation of the STING pathway was also shown to prevent replication of CHIKV demonstrating the potential therapeutic application of pharmacologically targeting activation of this protein .
We sought to discover novel small molecules capable of stimulating innate immune signaling and effector activity in human cells . For this we employed previously described human fibroblasts stably transfected with constitutively expressed human telomerase reverse transcriptase ( termed THF ) as well as luciferase ( LUC ) from Phytonis pyralis downstream of a promoter element that is reactive to type I IFN-dependent as well as IRF3-dependent transcription ( termed THF-ISRE ) [18] . Using these cells in a 384-well high-throughput in vitro screening platform we examined 51 , 632 chemically diverse compounds in duplicate for their ability to significantly stimulate expression of LUC . Sixteen positive control ( 1000U/mL IFNβ ) and negative control ( 1% DMSO ) LUC readings were obtained for each plate ( μP and μN averages , respectively ) . Readings for individual compounds ( R ) on a single plate were designated as significant if R > ( μP-μN ) *0 . 5 . Fig 1A illustrates the distribution of raw LUC readings for all molecules that exceeded this threshold on duplicate plates . A compound that exhibited the third highest signal of all those screened was 4- ( 2-chloro-6-fluorobenzyl ) -N- ( furan-2-ylmethyl ) -3-oxo-3 , 4-dihydro-2H-benzo[b][1 , 4]thiazine-6-carboxamide , which we termed G10 ( Fig 1B ) . The two molecules eliciting stronger responses are currently being characterized . As shown in S1 Fig G10 exhibited low cytotoxicity even at high concentrations and was thus selected for more comprehensive examination . We next validated the G10-mediated induction of IFN/IRF3-dependent LUC in reporter cells by exposing them to a range of concentrations . As shown in Fig 2A , the compound was able to induce LUC expression in a dose-dependent manner . We therefore aimed to verify that the molecule could trigger expression of endogenous genes transcribed in response to IRF3- and IFN-dependent processes . For this we employed semi-quantitative RT-PCR ( qPCR ) to examine induction of mRNA characterized as dependent on either IRF3 or IFN ( ISG54 , ISG56 , ISG15 , Viperin ) . As shown in Fig 2B , exposing cells to G10 led to transcription of cellular genes triggered by IRF3/IFN-dependent signaling in a manner similar to that induced by exposure to human cytomegalovirus rendered replicationally inactive by UV irradiation ( UV-CMV ) ; a stimulus of IRF3 and IFN that occurs via STING and ZBP1/DAI [18 , 50–52] . Since molecular patterns that culminate in IRF3 activation and synthesis of type I IFN often simultaneously induce pathways leading to activation of the transcription factor NF- κB we next examined whether G10 also stimulated this response . For this we employed telomerized human fibroblasts ( THF ) stably transduced with LUC driven by an NF-κB-dependent promoter . As shown in Fig 2C we detected no G10-associated stimulation of NF-κB-dependent transcription even when using high concentrations of the compound . This contrasts with what is observed in these cells for other NF-κB-inducing stimuli such as Sendai virus ( SeV ) , TNFα , and LPS . Moreover , the compound also failed to stimulate mRNA synthesis of endogenous NF-κB-dependent genes ( IL8 , IL1β , and MIP1α Fig 2D ) . Overall these results suggest that G10 activates IRF3 , but not canonical NF-κB pathways in human fibroblasts . Our results indicate that exposure of cells to G10 stimulates the expression of genes that are dependent on IRF3- and/or IFN-dependent signaling . Numerous such genes have been characterized as antiviral effectors that act either via direct molecular or indirect immunological mechanisms ( see [53] ) . We therefore examined whether G10 was correspondingly capable of generating a cellular state refractory to virus replication in vitro , presumably through the initial activity of IRF3 . For this we pre-exposed THF cells for 6h to concentrations of G10 that fell well within a nontoxic range ( S1 Fig ) yet that were observed to trigger innate gene transcription ( Fig 2 ) . We next quantitated the replication of West Nile Virus ( WNV ) , Vaccinia Virus ( VACV ) , and Chikungunya virus ( CHIKV ) on these cells . As shown in Fig 3 , the highest concentration of G10 was only able to reduce growth of VACV by approximately one log and WNV by less than one log . Intriguingly , however , the cellular state induced by G10 was much more inhibitory to the growth of CHIKV , reducing replication by over three logs ( IC90 = 8 . 01μM; Fig 3 ) . In light of this we examined replication of another clinically relevant Alphavirus species Venezuelan Equine Encephalitis Virus ( VEEV ) . As shown in Fig 3 , G10 also potently blocked replication of this virus ( IC90 = 24 . 57μM ) . While related , these species represent highly diverse Alphavirus clades with CHIKV deriving from the Old World and VEEV the New World . Nevertheless G10 was highly effective at similarly impairing replication of both viruses . The relative inability of G10 to render cells as resistant to WNV or VACV replication may reflect the differential susceptibilities of the virus types to the specific antiviral genes induced by G10 or innate evasion phenotypes exhibited by the different viruses ( see [54 , 55] ) . Since LUC expression in THF-ISRE reporter cells can be activated directly by IRF3 alone or following IFN-mediated Jak/STAT1/2 signaling we next examined whether either or both transcription complexes were required for this effect . We therefore constructed derivative THF-ISRE cells from which either the STAT1 or IRF3 protein was stably removed via disruption of the respective coding regions by lentivirus-delivered CRISPR/Cas9 components [56–59] . As shown in Fig 4A cells lacking IRF3 ( THF-ISRE-ΔIRF3 ) or STAT1 ( THF-ISRE-ΔSTAT1 ) protein as detectable by immunoblot ( IB ) were obtained . Absence of these proteins was functionally validated by the elimination of STAT1- ( Fig 4B ) or IRF3-dependent ( Fig 4C ) LUC expression following treatment with control stimuli IFNβ and UV-inactivated cytomegalovirus ( UV-CMV ) , respectively . Using these cells , G10 was found to induce LUC expression in the absence of STAT1 . However , G10-induced LUC expression was abrogated in cells lacking IRF3 . These results suggest that the innate immune stimulation observed in response to G10 is dependent on the IRF3 transcription factor and does not require direct activation of Jak/STAT-dependent signaling . Since transcription of a reporter gene by G10 is abolished in the absence of IRF3 this strongly implies that G10 stimulates the activation of IRF3 , which involves phosphorylation of C-terminal serine residues and subsequently allows its dimerization , nuclear translocation and DNA binding . To verify that IRF3 activation does occur in response to G10 we performed IB using an antibody reactive to phosphorylated IRF3 residue S386 with whole cell lysates harvested from THF exposed to G10 or control stimuli . As shown in Fig 4D G10 stimulates phosphorylation of IRF3 in a manner similar to that triggered by UV-CMV and SeV . We next examined whether IRF3 was involved in establishment of the observed G10-mediated anti-Alphaviral state ( Fig 3 ) . To answer this we utilized THF-ISRE-ΔIRF3 by exposing them to DMSO , 100μM G10 ( a concentration over twelve times the IC90 for CHIKV and over four times the IC90 for VEEV ) , or 1000U/mL IFNβ and examined growth of CHIKV and VEEV after a period that enables peak viral titers . Fig 4E shows that while a strong anti-Alphaviral state can still be established in these cells by pre-exposure to IFNβ , the ability of G10 to block replication of CHIKV and VEEV is lost . In addition , we also examined Sindbis virus ( SINV ) , another Old World Alphavirus species , that grows poorly on wild type human fibroblasts ( S2 Fig ) but can replicate when the IPS-1-IRF3-IFN response is impaired . As shown in Fig 4E cells lacking IRF3 are permissive for SINV replication . However , in these cells IFNβ , but not G10 , is capable of inhibiting virus replication . These results indicate the anti-Alphaviral activity elicited by G10 requires IRF3-dependent cellular responses . An array of PRRs reacting with multiple classes of pathogen-associated molecules is capable of initiating signaling pathways that terminate in IRF3 activation . As discussed above these conventionally employ an adaptor protein to activate the IRF3-directed kinases TBK1 and IKKε . IFNβ promoter stimulator 1 ( IPS-1 , also called MAVS ) is utilized by RIG-I and MDA5 , cytoplasmic sensors of ( typically virus-associated ) dsRNA . In an effort to characterize the cellular pathway targeted by G10 we first asked whether IPS-1 is important for the molecule’s effect on innate cellular activation . To address this we again employed lentivirus—delivered CRISPR/Cas9 to construct THF cells lacking the protein . As shown in Fig 5A disruption of the IPS-1 coding region in this manner correspondingly results in undetectable protein . Furthermore , IRF3 activation in response to poly ( I:C ) , ppp-dsRNA , or SeV infection , processes requiring IPS-1 [8 , 60 , 61] , are accordingly abrogated in these cells . In contrast , IRF3 activation by UV-CMV or 2’3’-cGAMP , which occur via the STING pathway in these cells [13 , 52 , 62] and are independent of IPS-1 remains intact . Likewise , G10-induced IRF3 phosphorylation is also functional in these cells ( Fig 5A ) indicating that the molecule stimulates a response that does not require components of the IPS-1 signaling apparatus . We next asked whether the anti-Alphaviral activity elicited by G10 similarly occurred independently of IPS-1 as described above . As shown in Fig 5B , pretreatment of cells with either G10 or IFNβ diminishes replication of all three virus types by multiple logs; a magnitude that parallels what is observed in wild type cells ( Fig 3 ) . Together these results indicate that the IRF3-dependent anti-Alphaviral activity conferred by G10 does not require IPS-1 or , by extension , any upstream PRRs ( i . e . RIG-I , MDA5 , POL3 ) associated with this signaling pathway . We next examined whether the IRF3-terminal adaptor protein STING is critical to G10-mediated innate activation . For this we constructed THF-ISRE cells from which the STING protein is eliminated via CRISPR/Cas9-mediated gene disruption as described above . Knockout of the protein was confirmed visually by IB of whole cell lysates and functionally by demonstrating the absence of IRF3 S386 phosphorylation following treatment with UV-CMV or transfection with 2’3’-cGAMP , both of which are STING-dependent cellular reactions ( Fig 6A ) . IPS-1-dependent signaling remains operational in these cells , however , as evidenced by IRF3 phosphorylation in response to SeV exposure and poly ( I:C ) transfection . Intriguingly , STING deletion resulted in elimination of G10-induced IRF3 S386 phosphorylation . We next examined whether the G10-induced transcriptional response observed in wild type cells ( Fig 2 ) was similarly inactive in STING-deficient cells . For this we also included THF-ISRE-ΔIPS-1 cells as a control for any potential off-target effects of CRISPR/Cas9 genome editing or lentivirus transduction . As shown in Fig 6B , treatment with G10 leads to strong LUC induction in cells lacking IPS-1 , as does exposure to UV-CMV or LPS ( a TRIF-dependent process ) but not SeV . However , in the absence of STING both G10 and UV-CMV fail to activate appreciable LUC whereas SeV and LPS induce substantial expression thus validating intact IRF3-terminal signaling . Transcriptional induction of IRF3-dependent endogenous host genes was also examined by qPCR . Fig 6C shows that , similar to LUC expression , ISG54 , ISG15 , and Viperin mRNAs are highly transcribed in response to UV-CMV and G10 exposure in cells lacking IPS-1 but remain comparatively unstimulated in cells lacking STING . In contrast , SeV-induced transcription of these genes is only observed in cells lacking STING . From these results we conclude that the IRF3 activation and IRF3-dependent transcriptional activity in response to G10 occurs via a STING-dependent pathway . Based on our observations above that G10-elicited anti-Alphaviral activity requires IRF3 ( Fig 4 ) , we hypothesized that STING-dependent IRF3 signaling is also essential to this process . We therefore examined replication of CHIKV and VEEV in THF-ISRE-ΔSTING cells following treatment with DMSO , G10 , or IFNβ as described above ( these cells are not permissive for SINV replication since IPS-1-IRF3-IFN signaling is intact [49] ) . As shown in Fig 6D the ability of G10 to block virus replication is absent in cells lacking STING despite the fact that an antiviral state can effectively be established as indicated by treatment with IFNβ . Overall these data clearly indicate that the STING pathway is crucial to the IRF3-dependent innate antiviral activity induced by G10 . STING behaves as a PRR of cyclic dinucleotides ( CDN ) by way of a direct interaction between them and the protein’s C-terminal ( ligand-binding ) domain [13 , 63] . We therefore asked whether G10 was a direct ligand of human STING , similar to the mouse STING-specific small molecules DMXAA [40 , 41 , 64 , 65] and CMA [39] , and sought evidence supporting this hypothesis . For this we used differential scanning fluorimetry to examine changes in thermal stability of purified STING-CTD in the presence of G10 . Thermal stability of the protein is expected to increase with binding affinity of protein-ligand complexes [66 , 67] . However , as shown in S3 Fig , incubating purified mouse or human STING-CTD with G10 did not increase the protein’s thermal stability , as does a validated ligand such as 2’3’-cGAMP or DMXAA ( in the case of mouse protein ) . These observations are inconsistent with G10 binding directly to human or mouse STING-CTD . STING also behaves as an adaptor molecule [68] required for activating IRF3-targeting kinases by multiple upstream cytoplasmic DNA-sensing PRRs including ZBP1/DAI [69] , IFI16 [16] , DDX41 [70] , and IFI203 [17] . Given that we were unable to detect evidence of direct interaction with STING it is possible that G10 engages one or more of these ( or an as yet unknown; see [71] ) PRRs to initiate STING-dependent activity . In light of this it is interesting to note that G10 does not induce IRF3 activation or IRF3-dependent gene expression in the immortalized promonocytic cell line THP-1 despite the fact that these cells express phenotypically active STING [16 , 70 , 72–77] ( S4 Fig ) . These results are potentially informative with respect to identity of the direct cellular target of G10 since it likely rules out cGAS [14] , DDX41 [73] , and IFI16 [16] based on the observation that these are present and functional in THP-1 cells [78] . Currently we are examining whether other known STING-dependent receptors ( IFI203 and ZBP1/DAI ) are required for G10-mediated activity . Our data indicate that G10 induces expression of cellular antiviral effector genes and that this process ultimately requires IRF3 and STING . However , transcription of these genes ( ISG15 , Viperin , ISG54 , ISG56 ) can be triggered in response to either activated IRF3 or IFN-dependent ( Jak/STAT ) signaling [79–81] . Therefore , we next aimed to discern the respective roles of these pathways in G10-mediated innate and antiviral activation . For this we first examined whether G10 is able to stimulate expression of type I or III interferons , both of which are known to induce Jak/STAT-dependent signaling via type I and type III IFN receptor complexes , respectively . As shown in Fig 7A , exposure of THF-ISRE to G10 leads to transcription of both IFNβ and IFNλ1 mRNAs . However , induction of IFNλ1 by G10 does not approach the levels observed for UV-CMV and SeV suggesting the involvement of different , potentially G10-independent , cellular factors in this process . We next examined whether genes known to be induced solely by Jak/STAT- ( as opposed to either Jak/STAT or IRF3 ) dependent signaling were also induced following G10 exposure since this would be indicative of autocrine/paracrine signaling following release of type I or type III IFN from treated cells . In contrast to the 7h treatments used above ( Figs 2 and 6 ) we exposed cells to indicated stimuli for 18h in order to allow for completion of the full sequence of IFN transcription , synthesis , secretion , and autocrine/paracrine signal transduction . Fig 7B illustrates substantial induction of known IFNα/β-dependent genes Mx2 and OAS in response to exposure to G10 as well as other control stimuli including IFNβ . These results are consistent with the secretion of IFN in response to treatment with G10 . Numerous subtypes of type I and III interferons exist and thus demonstrating the presence of secreted molecules requires type-specific immunoassays ( ELISA ) . We chose to examine secreted interferon of all subtypes by employing a cell-based reporter assay that reacts with any bioactive type I or III interferon species . For this we utilized THF-ISRE-ΔIRF3 cells described above ( Fig 4 ) . Since these cells cannot make type I IFN ( due to the absence of IRF3 ) they do not generate an autocrine LUC reporter signal in response to exogenous IFN-inducing stimuli and only react to IFN itself ( Fig 4 ) . We treated THF-ISRE , THF-ISRE-ΔIPS-1 , and THF-ISRE-ΔSTING with DMSO , SeV , UV-CMV , or G10 for 18h and transferred media from these cells to THF-ISRE-ΔIRF3 for 8h . Next , IFN-dependent luciferase expression was measured by luminescence . In agreement with our previous observations , IFN secretion was detected by wild type THF cells exposed to UV-CMV , SeV , or G10 . In cells lacking IPS-1 secretion of IFN was abolished in response to SeV but not UV-CMV or G10 . Moreover , cells lacking STING failed to secrete IFN in response to UV-CMV or G10 but SeV-induced secretion was intact . Interestingly , G10-induced secretion by cells lacking IPS-1 was significantly diminished relative to wild type cells ( p = 0 . 00013 ) perhaps indicating pivotal crosstalk between STING and IPS-1 signaling , a phenomenon previously alluded to [11] . However , as indicated by Fig 6 the role of IPS-1 does not appear to be essential for G10’s anti-Alphaviral activity . Overall these data indicate that G10 triggers the STING-dependent transcription , synthesis , and secretion of IFN species capable of initiating Jak/STAT signaling and ISG expression . G10 exposure elicits secretion of bioactive IFN in cells that contain STING and IRF3 ( Fig 7C ) but also the expression of two classes of ISGs: Those that are induced by either IRF3- or IFN-dependent signaling ( ISG15 , ISG54 , ISG56 , Viperin ) and those whose expression is activated only by IFN-mediated Jak/STAT signaling ( Mx2 , OAS ) . Since many IRF3-dependent genes are direct antiviral effectors ( see [79 , 82] ) we asked whether the G10-mediated anti-Alphaviral cellular state we observe could be elicited in the absence of canonical STAT1-mediated , IFNα/β-induced signaling by examining replication of CHIKV , VEEV , and SINV on cells lacking STAT1 . As shown in Fig 8A replication of these viruses in cells treated with G10 from which STAT1 was deleted is similar to that seen in the presence of DMSO alone . Intriguingly , IFNβ appears to induce detectable yet significant ( in the case of VEEV and CHIKV ) and profound ( in the case of SINV ) antiviral effects even in the absence of STAT1 . These data suggest the expression of antiviral effectors that are IFN-induced yet independent of STAT1 . Cells lacking STAT1 can still react to IFN-exposure via the activity of STAT2 , STAT3 , STAT4 , or STAT5 . In fact , expression of antiviral ISGs ( including those examined here ) has been detected in the absence of STAT1 , likely via STAT2 homodimers [83–85] . In light of these results we examined whether IFN or G10 could stimulate expression of either class of ISG in the absence of STAT1 . As shown in Fig 8B SeV , UV-CMV , IFNβ , and G10 were all capable of triggering synthesis of Mx2 and ISG56 proteins . Surprisingly , however , in STAT1-deficient cells Mx2 and ISG56 proteins were still detectable following exposure to SeV , UV-CMV , and IFNβ but not G10 . These results are consistent with previous studies showing that some IFN-responsive antiviral effectors are inducible in the absence of STAT1 [83–87] Our observations indicating that STING is required for G10-mediated induction of IRF3-dependent cellular activity led us to examine how the response to the molecule compares to that of a canonical STING ligand such as 2’3’-cGAMP . For this we examined the kinetics of IRF3 phosphorylation and levels of IRF3-dependent gene induction following exposure to each of the molecules . As shown in Fig 9A phosphorylation of IRF3 S386 is detectable at 30m post exposure to G10 and at 1h post-exposure to 2’3’-cGAMP . It is important to note that 2’3’-cGAMP was transfected into THF and this may affect timing of observable IRF3 phosphorylation . We next examined the transcriptional induction of IRF3- and IFN-dependent genes in response to a range of G10 and 2’3’-cGAMP concentrations . Fig 9B illustrates that 2’3’-cGAMP elicits stronger mRNA synthesis of IFIT1 , IFIT2 , ISG15 , Viperin , OAS , and Mx2 at lower concentrations than G10 . Oddly , transcription of IFNβ appeared to show a somewhat different pattern with dose-dependent induction being roughly similar between the molecules . These results indicate that innate immune reactivity may occur more quickly in cells exposed directly to G10 relative to 2’3’-cGAMP ( following transfection ) but that higher concentrations of the molecule are needed to trigger similar levels of IRF3/IFN-dependent gene expression . To this point our examination of G10-mediated innate activation has focused on human fibroblasts that , while not strictly immortalized , are life-extended through the introduction of telomerase reverse transcriptase . To determine whether G10 is immunostimulatory to a similar degree in more physiologically relevant primary cell types we examined the transcription of IRF3- , IFN- , and NF-κB-dependent genes in human peripheral blood mononuclear cells ( PBMCs ) . As shown in Fig 10 , G10 triggered expression of IFIT1 , IFIT2 , IFNβ , ISG15 , OAS , and Viperin to degrees that were proportional to compound concentration . Interestingly , NF-κB-dependent genes MIP1α and IL1β were also transcriptionally induced by G10 , a phenomenon not observed in THF cells ( Fig 2 ) . Respective IPS-1- and STING-inducing control stimuli ppp-dsRNA and 2’3’-cGAMP were similarly capable of activating transcription of these genes in these cells ( Fig 10B ) . As shown in S5 Fig primary human umbilical microvascular endothelial cells also responded to G10 exposure by expressing IRF3- , IFN- , and NF-κB-dependent genes . These results clearly demonstrate that innate immune induction by G10 is not an effect specific to the cell model and suggests that in vivo stimulation by G10 is likely to be feasible . While G10-induced innate immune activation is observed in multiple human cell types , we were unable to detect similar activity in murine myeloid-derived RAW264 . 7 cells ( S6 Fig ) . Multiple molecular analogs of G10 were thus constructed in an attempt to identify one that is active in both human and mouse cells . While this allowed characterization of essential and nonessential moieties within the molecule ( S7 Fig ) , no derivatives were identified that were active across species . However , 5 , 6-dimethylxanthenone-4-acetic acid ( DMXAA ) is a small molecule that has been examined extensively and shown to trigger STING-dependent IRF3 and interferon activity in murine but not human cells ( [40 , 41] and S3 Fig and S6 Fig ) . As such , the compound has been explored for multiple immunotherapeutic uses including anti-angiogenesis [88] , vaccine adjuvanticity [89] , anti-tumor immunology [38] and antiviral activity [32 , 33 , 90] . We therefore explored patterns of DMXAA-stimulated innate immune activation in murine cells to evaluate their resemblance with those induced by G10 in human cells . Fig 11A illustrates that DMXAA-induced IRF3 phosphorylation in RAW264 . 7 macrophage-like cells is detectable by 1h post-treatment . Furthermore , DMXAA also elicits dose-dependent transcription of key innate antiviral genes IFNβ , ISG15 , IFIT2 , and Viperin in a manner similar to that observed for G10 in human cells ( Fig 11B ) . The physiological effects of DMXAA thus appear to resemble those we observe for G10 . Fortunately , wild type mice represent a commonly used model of CHIKV infection that manifests viremia , pathogenesis , and immune responses resembling those seen in human patients [46 , 91–93] . We therefore decided to use this model to ask whether artificial stimulation of STING-dependent signaling was sufficient to block CHIKV replication in vivo . DMSO or DMXAA ( 25mg/kg ) were administered to mice intraperitoneally at 3h pre-infection with CHIKV ( 1000 PFU ) . As shown in Fig 11 treatment with DMSO resulted in an average titer of 1 . 47 x 104 PFU/mL serum at 72h post infection . In contrast , serum-associated virus was undetectable by plaque assay in mice pre-treated with DMXAA . We next examined whether viral titers were influenced by DMXAA given post-inoculation . For this we administered DMXAA at 6h and 24h post CHIKV infection . Fig 11 illustrates that virus was still undetectable in serum from mice treated at 6h post infection . However , mice treated at 24h post infection showed viral titers that were diminished relative to DMSO-treated animals but this difference was not statistically significant ( p = 0 . 1386 ) . Overall these data indicate that artificial stimulation of STING-dependent signaling in vivo within an appropriate temporal window can block CHIKV replication . It is likely that the antiviral efficacy of STING activation will diminish as time after acute viral replication increases and viral innate evasion phenotypes appear . It is also possible that recurrent STING activation could have an effect on persistence of CHIKV genomic RNA in joint tissues [94] and we are currently examining this possibility .
Pharmacologic activation of STING-dependent signaling represents a potentially high-impact therapeutic strategy with applications in diverse clinical areas such as broad-spectrum antivirals , vaccine adjuvants , vascular disruption , and antitumor immunology . This is represented by multiple successes of the utilization of this approach in mouse models of virus infection [29 , 32 , 33 , 90 , 95] , enhancement of vaccine immunogenicity [89 , 96–98] , immune-mediated tumor necrosis [36 , 38 , 99] , and inhibition of solid tumor angiogenesis [100 , 101] . Unfortunately , synthetic small molecules identified thus far have only exhibited suitable efficacy in mouse models due to their strict specificity for the murine STING ortholog [39–41] . Here we employed high-throughput screening to identify a novel compound ( G10 ) capable of triggering IRF3/IFN-dependent responses and subsequently blocking replication of CHIKV , VEEV , and SINV in human cells . Follow-up work seeking to pinpoint cellular targets essential to the phenotypic responses utilized a reverse genetics approach by way of CRISPR/Cas9-mediated genome editing . This enabled identification of the STING protein as required for G10’s biological activity thus indicating that the compound is the first described human-specific synthetic small molecule STING agonist . G10 triggers innate immune responses that involve expression of IRF3-dependent genes including type I and III interferons . This was observed in telomerized foreskin fibroblasts as well as primary cells such as PBMCs and endothelial cells . Unexpectedly , however , G10 did not induce expression of genes associated with the activity of NF-κB in fibroblasts even though such genes were induced in PBMCs and endothelial cells . Given the central role of NF-κB in generation of pro-inflammatory states that can lead to pathogenic consequences , especially under chronic circumstances , ( reviewed in [102 , 103] ) , it is perhaps desirable that the activity of G10 is more transcriptionally focused to IRF3-dependent responses in certain cell types . It is also interesting that G10 induces type I IFN synthesis in the absence of detectable NF-κB activity given the reported requirement of the transcription factor for this process [104–107] . Activation of noncanonical NF-κB subunits may play a role in this case . Undertaking a more thorough molecular investigation of NF-κB subunit activation ( e . g . nuclear localization , phosphorylation , DNA binding ) will be required to understand this with greater clarity . Importantly , G10 induced the phosphorylation of IRF3 and the protein’s deletion led to elimination of reporter gene transcription as well as the compound’s anti-Alphaviral activity . As such the innate biological effects of G10 examined here require IRF3-driven gene expression . Deletion of the adaptor molecule IPS-1/MAVS did not eliminate G10-induced IRF3 phosphorylation or affect the molecule’s antiviral effect ( Fig 5 ) . Furthermore , G10-associated transcription of IRF3-dependent genes was also intact in the absence of IPS-1 ( Fig 6 ) . In light of these results , it is intriguing that G10-induced IFN secretion was diminished following IPS-1 deletion ( Fig 7 ) . This result is consistent with data showing interaction between IPS-1 and STING during RIG-I-mediated stimulation ( reviewed in [54] ) although whether STING strictly requires this interaction for full signaling has not been shown . Nevertheless , IPS-1 does not appear to play a substantial role in G10-mediated anti-Alphaviral activity and thus upstream IPS-1-dependent PRRs such as RIG-I and MDA5 are unlikely to be engaged by G10 or relevant for these effects . Our results clearly establish an essential role for the signaling molecule STING . Deletion of the STING protein resulted in complete inactivation of G10-mediated IRF3 phosphorylation , IRF3-dependent transcription , IFN secretion , and antiviral activity ( Figs 6 and 7 ) . These results plainly signify that STING-dependent function ( s ) are necessary for the innate phenotypic response elicited by G10 . Whether G10 represents a directly binding and activating synthetic ligand of human STING ( as are DMXAA and CMA for mouse STING ) was examined using thermal shift assays of purified protein . These revealed no increase in the thermal stability of STING-CTD in the presence of G10 as was seen for 2’3’-cGAMP ( S3 Fig ) . If the molecule bound directly to STING-CTD we would expect a net increase in the melting temperature of the protein dimers [66] as observed when a bona fide ligands such as 2’3’-cGAMP is co-incubated . Additionally , the inability of G10 to induce IRF3-dependent transcription in THP-1 cells ( S4 Fig ) is also not consistent with the molecule behaving as a direct ligand since these cells express biologically functional STING , as described in numerous studies [16 , 70 , 72–77] . The identity of the protein or factor engaged by G10 that ultimately stimulates the STING-dependent response is currently under investigation . IRF3-activating , STING-dependent sensors such as IFI16 , DDX41 , and cGAS are also present and functional in THP-1 cells [17 , 73–75 , 78 , 108] . As such , it is probable that G10 either engages an alternative STING-dependent PRR such as ZBP1/DAI [18] or IFI203 [17] or an as yet uncharacterized STING-dependent PRR [71] . ZBP1/DAI is a particularly attractive target since we have previously shown it to be expressed and biologically active in THF cells [18] . We are currently addressing this question using CRISPR/Cas9-mediated genome editing and ZBP1/DAI and IFI203 overexpression studies . Given that G10 stimulates innate cellular effects that require STING we decided to compare the dose dependence of these effects to 2’3’-cGAMP , an established STING ligand . Our results indicate that while G10 may trigger earlier IRF3 phosphorylation than 2’3’-cGAMP , perhaps due to its smaller size and cell permeability , it triggers levels of IRF3-dependent gene expression with overall less potency than 2’3’-cGAMP ( Fig 9 ) . More precisely , 2’3’-cGAMP induces higher levels of IRF3- and IFN-dependent mRNA expression at lower concentrations than G10 . It would be interesting to establish whether these dissimilarities are causally linked to differences in the molecules’ cellular targets , especially whether their proximity in the signaling cascade to IRF3-directed kinases is important . Alternatively , differences in physico-chemical properties between the molecules and how those relate to solubility and permeability may also impact stimulatory potency [109] . G10 induces synthesis and secretion of bioactive type I and III IFNs and generates an antiviral state in fibroblast cells positive for STING , IRF3 , and STAT1 proteins . Based on these results our model for the elicitation of anti-Alphaviral activity by G10 first involves STING-dependent induction of IRF3 followed by IRF3-mediated synthesis and secretion of type I and III IFNs and subsequent IFN-stimulated , STAT1-dependent expression of antiviral effectors . Detection of STAT1-independent ISG expression in response to IFN exposure and IFN-inducing stimuli was unexpected but not unprecedented and has been reported in multiple studies [83–85 , 110 , 111] . Blaszcyk and colleagues attribute this to IFN-induced transcriptional complexes composed of IRF9 and STAT2 homodimers [84] although homo- and heterodimers of other Jak/Tyk2-phosphorylated STAT proteins may also play roles ( reviewed in [112] ) . Interestingly , IFNβ was able to stimulate some antiviral activity in cells lacking STAT1 and to a degree that varied between viruses with SINV replication being undetectable . The full assortment of STAT1-independent ISGs expressed cannot be inferred from two proteins ( Mx2 and ISG56 ) and as such the differential susceptibilities of CHIKV , VEEV , and SINV to ISG-encoded proteins in general cannot be known based on these results . Yet it is clear that SINV is highly sensitive to STAT1-independent ISGs relative to the other Alphaviruses . Intriguingly , however , while other IRF3-activating , IFN-inducing stimuli were capable of triggering expression of Mx2 and ISG56 in the absence of STAT1 , G10 was not . This likely explains the reliance on STAT1 of G10-mediated anti-Alphaviral activity . Why this disparity in STAT1-dependence occurs between SeV , UV-CMV , and G10 is not clear . It is possible that each stimulus triggers the secretion of unique signatures of type I and type III IFN subtypes that subsequently elicit distinct gene expression patterns [113 , 114] . Elucidation of the importance of the various IFN proteins in G10’s antiviral effects will require more detailed examination , for instance by comparative transcriptomics , by using subtype-specific neutralizing antibodies or reverse genetics via gene editing . While the majority of our investigation employed fibroblast cells , it is evident that G10 elicited innate immune activation in primary human cells such as PBMC’s ( Fig 10 ) and umbilical endothelial cells ( S5 Fig ) . Unexpectedly , however , induction of NF-κB-dependent transcription by G10 was observed in primary cells but not fibroblasts . Moreover , no G10-induced IRF3-dependent activity was detected in THP-1 cells ( S4 Fig ) . These disparities may be related to differences in cell type-specific expression of PRRs or innate signaling molecules , especially between stromal versus myeloid-derived cells [115] and between transformed and untransformed cells ( [116] and references therein ) . Understanding the biological bases for these divergent effects will require additional experimentation . However , demonstrating efficacy of G10 on primary human cells is obviously crucial to assessing the therapeutic potential of the compound . Unfortunately G10 was unable to stimulate similar activation in murine cells . As such , evaluating the in vivo efficacy of G10 using a well-established mouse model of Alphavirus ( CHIKV ) infection was not directly practical . Yet IFN-inducing STING agonists ( e . g . DMXAA , CMA ) have been described that are murine specific [39–41] . We therefore examined whether DMXAA triggers IRF3 activation and IRF3-dependent gene induction in a manner comparable to G10 . While comparisons of absolute responses are complicated by the fact that different species , cell types , and reagents are employed , DMXAA does trigger rapid IRF3 phosphorylation and dose-dependent IFNβ and ISG transcription in mouse cells ( Fig 11 ) as does G10 in humans cells . In light of this we used the mouse model of acute CHIKV infection to ask whether activation of the STING pathway is feasible as an in vivo anti-Alphaviral strategy . We demonstrate that DMXAA clearly blocks viremia but that this is related to the timing of administration with early ( 3h pre- or 6h post-infection ) being more effective than late ( 24h post-infection ) treatment . It is probable that these kinetics correlate with the appearance of CHIKV-encoded IFN/ISG evasion phenotypes and as such STING-dependent antiviral efficacy diminishes with time post infection . However , whether STING activation represents an effective approach for diminishing persistent ( e . g . >6 weeks ) CHIKV infection [94 , 117] is an enticing possibility that warrants examination since this could lead to alleviation of chronic virus-associated arthralgia and is currently being examined in our laboratory . In summary we have identified a novel synthetic small molecule capable of inducing expression of type I and III IFNs as well as IFN-dependent antiviral effector genes . Using a reverse genetics approach based on CRISPR/Cas9-mediated genome editing to identify cellular targets of the molecule we shown that this effect requires STING , IRF3 , and STAT1 proteins . These molecules are likewise essential to the ability of G10 to elicit a cellular state refractory to replication of Alphavirus species . Furthermore , given the pivotal role of STING we also show that pharmacologic activation of the molecule represents an effective anti-Alphaviral strategy in vivo . Given the demonstrated role of STING pathway stimulation in numerous immunological processes , it is being pursued as a therapeutic target for many diseases . Our work demonstrates the feasibility of identifying molecules that activate STING-dependent signaling and yield therapeutic outcomes as well as a strategy for characterizing cellular effects and essential modulatory proteins via genome editing .
Dimenthyl sulfoxide ( DMSO ) was obtained from Thermo-Fisher . Puromycin was obtained from Clontech and used at 3μg/mL in cell culture medium . Lipopolysaccharide ( LPS ) and polybrene were obtained from Sigma-Aldrich . Human recombinant IFNβ and tumor necrosis factor α ( TNFα ) were obtained from PBL . ONE-Glo cell lysis/luciferin reagent was obtained from Promega . Lucia luciferin reagent was obtained from Invivogen . Lipofectamine LTX was obtained from Life Technologies . Poly ( I:C ) was obtained from Amersham ( 27–4729 ) . 2’3’-cGAMP and ppp-dsRNA were purchased from Invivogen ( tlrl-cga23 and tlrl-3prna , respectively ) . Unless otherwise indicated cells were exposed to ppp-dsRNA at 12 . 5μg/mL based on a dose response of innate immune activity performed on THF cells . Stocks of G10 were purchased from ChemDiv . DMXAA was purchased from ApexBio . Antibodies used against the following antigens are indicated in parentheses: GAPDH ( Santa Cruz SC-51906 ) ; STAT1 ( Santa Cruz SC-346 ) IRF3 ( Santa Cruz SC-9082 ) ; human S386 phospho-IRF3 ( Epitomics 2562–1 ) ; mouse S396 phospho-IRF3 ( cell Signaling 4947 ) ; STING ( Cell Signaling 3337 ) ; IPS-1 ( Bethyl A300-782A ) ; IFIT1/ISG56 ( Thermo Fisher PA3 848 ) ; and Mx2 ( Sigma HPA030235 ) . Human foreskin fibroblasts originally obtained from the American Type Culture Collection were stably transduced with constitutively expressed human telomerase reverse transcriptase and the IRF3/IFN-responsive pGreenFire-ISRE lentivector and were maintained in DMEM containing 10% fetal calf serum ( FCS ) and antibiotics as described previously [52] . Vero , BHK-21 , and C6/36 cells were obtained from Alec Hirsch ( Oregon Health and Science University ) and were grown as described [49] . RAW264 . 7 cells were obtained from Jay Nelson ( Oregon Health and Science University ) and transduced with a lentivector that contains firefly luciferase under the control of the type I IFN responsive element obtained from SA Biosciences . THP1-ISG-Lucia cells were obtained from Invivogen and maintained in RPMI containing 10% FCS and antibiotics . These cells were differentiated in 100nM phorbol 12-myristate 13-acetate ( PMA ) for 24h before stimulation . Human peripheral blood mononuclear cells were obtained from StemCell Technologies and maintained in RPMI containing 10% FCS and antibiotics . Human umbilical microvascular endothelial cells were obtained from Patrizia Caposio ( Oregon Health and Science University ) and maintained as described [118] . All cells were grown at 37°C and 5% CO2 . Sendai virus ( SeV ) was obtained from Charles River Laboratories and used at 16 HA units/mL . Cytomegalovirus was grown , titered , UV-inactivated , and exposed to cells as described previously [51 , 52] . West Nile Virus ( WNV ) was obtained from Alec Hirsch ( Oregon Health and Science University ) and used as previously described [119] . Vaccinia Virus ( VACV ) strain Western Reserve was obtained from Klaus Früh ( Oregon Health and Science University ) and used as previously described [120] . Sindbis virus ( SINV ) strain Ar-339 was obtained from ATCC . Venezeulan encephalitis virus ( VEEV ) strain TC83 and Chikungunya virus ( CHIKV ) strain MH56 were obtained from Michael Diamond ( Washington University ) . CHIKV was derived from an infectious clone as follows . RNA was transcribed from the linearized clone using the T7 mMessage mMachine kit ( Ambion ) and transfected using Lipofectamine LTX into BHK-21 cells . Resultant virus was propagated in C6/36-insect cells for 48h to produce high titer viral stocks after pelleting through a 20% sucrose cushion by ultracentrifugation ( 22 , 000 rpm , 825206g for 1 . 5 hrs ) . In all cases infectious virus was quantified by serial dilution plaque assays on Vero cells with a carboxymethylcellulose overlay . Unless otherwise indicated experimental infections were carried out in triplicate using a multiplicity of infection ( MOI ) of 1 plaque forming unit ( PFU ) per cell . Cell viability was examined by quantitating ATP using the Cell Titer GLO assay according to the manufacturer’s instructions ( Promega ) . Sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) immunoblots were performed as follows . After trypsinization and cell pelleting at 2 , 000 x g for 10 min . whole-cell lysates were harvested in 2% SDS lysis buffer ( 50 mM Tris-HCl , 20% glycerol ) . Lysates were electrophoresed in 8% polyacrylamide gels and transferred onto polyvinylidene difluoride membranes ( Millipore ) using semidry transfer at 400 mA for 1h . The blots were blocked at room temperature for 2h or overnight using 10% nonfat milk in 1x PBS containing 0 . 1% Tween 20 . The blots were exposed to primary antibody in 5% nonfat milk in 1x PBS containing 0 . 1% Tween 20 for 18 h at 4°C . The blots were then washed in 1x PBS containing 0 . 1% Tween 20 for 20 , 15 , and 5 min , followed by deionized water for 5 min . A 1h exposure to horseradish peroxidase-conjugated secondary antibodies and subsequent washes were performed as described for the primary antibodies . The antibody was visualized using enhanced chemilumi- nescence ( Pierce ) . Total RNA was isolated from cells and DNased using a DNA Free RNA Isolation kit according to the manufacturer’s protocol ( Zymo Research ) and quantified by UV spectrometry . Single-stranded cDNA for use as a PCR template was made from total RNA using random hexamers to prime first-strand synthesis by Superscript III reverse transcriptase ( Life Technologies ) as described in the manufacturer’s protocol . Comparison of mRNA expression between samples ( e . g . , treated versus untreated ) was performed using semiquantitative real-time RT-PCR ( qPCR ) with the Applied Biosystems sequence detection system according to the ΔΔCT method [121] . For IFNβ , IFNλ1 , and mouse and human GAPDH ( housekeeping gene ) pre-validated PrimeTime FAM qPCR primer/probe sets obtained from IDT were used . For all other genes Maxima SYBR Green qPCR master mix ( Thermo Fisher ) was used . Primers for human ISG15 , ISG56 , ISG54 , and Viperin were described in [18 , 51 , 52 , 122] . Other human primer sequences were as follows: Mx2-For , 5’-ACTTCAGTTCAGAATGGAG-3’ Mx2-Rev , 5’-TATTCTGTGAAGGCGTCC-3’ OAS-For , 5’- CAGCGCCCCACCAAGCTCAA-3’ OAS-Rev , 5’- TGCTCCCTCGCTCCCAAGCA-3’ IL8-For , 5’-GACTTCCAAGCTGGCCGT-3’ IL8-Rev , 5’-GAATTCTCAGCCCTCTTCA-3’ IL1β-For , 5’- AACAGGCTGCTCTGGGATTCTCTT-3’ IL1β-Rev , 5’- TGAAGGGAAAGAAGGTGCTCAGGT-3’ MIP1α-For , 5’- GCTGCCCTTGCTGTCCTCCTC-3’ MIP1α-Rev , 5’- GGTCAGCACAGACCTGCCGG-3’ . Mouse primers were as follows: IFIT2/ISG54-For , 5’-TCCAGCCCCTACAGGATTGA-3’ IFIT2/ISG54-Rev , 5’-TTCGGGTCCTTTTCCAGAGC-3’ IFNβ-For , 5’-CTGGAGCAGCTGAATGGAAAG-3’ IFNβ-Rev , 5’-CTTCTCCGTCATCTCCATAGGG-3’ Viperin-For , 5’-AGCAGGTGTGTGCCTATCAC-3’ Viperin-Rev , 5’-TCAGCCAGCAGAACAGGATG-3 . NF-κB-responsive luciferase reporter cells were made using a commercially available replication incompetent lentivirus ( Qiagen ) . Telomerized human fibroblasts were exposed to virus inoculum in the presence of DMEM plus 5μg/mL polybrene and rocked at 37°C for 8h . At two days post inoculation cells were exposed to 3μg/mL puromycin . After cells were fully resistant to puromycin they were verified for responsiveness to NF-κB-inducing stimuli ( e . g . TNFα , SeV , LPS ) . Genome editing using lentivector-mediated delivery of CRISPR/Cas9 components was performed generally as described previously [56] . Briefly , 20nt guide RNA ( gRNA ) sequences targeting protein-coding regions were inserted into the lentiCRISPRv2 vector ( AddGene # 52961 ) . These sequences are as follows . IRF3: GAGGTGACAGCCTTCTACCG; IPS-1: AGTACTTCATTGCGGCACTG; STAT1: AGAACACGAGACCAATGGTG; STING: CCCGTGTCCCAGGGGTCACG . Lentivirus was made by transfecting specific lentiCRISPRv2 plasmid along with packaging ( psPAX2; AddGene # 12260 ) and VSV-G pseudotyping ( pMD2 . G; Addgene # 12259 ) plasmids into Lenti-X 293T cells ( Clontech ) using Lipofectamine-LTX ( Life Technologies ) . Media was harvested at 48h and 72h post transfection , centrifuged ( 3 , 000 x g for 10 min . ) and filtered through a 0 . 45-μm-pore-size filter to remove cell debris . Subconfluent target cells were exposed to lentivirus for 8h in the presence of 5 μg/mL polybrene . After the cells reached confluence they were split into DMEM plus 10% FCS containing 3μg/mL puromycin . Transduced cells were passaged in the presence of puromycin for 7–10 days before protein knockout was examined by immunoblot . Cells were next serially diluted twice in 96 well plates to obtain oligoclonal lines purified for gene deletion . Protein knockout was additionally verified functionally by measuring phenotypic responsiveness to relevant stimuli as discussed below . Confluent reporter cells were plated at 20 , 000 ( THF-ISRE ) or 100 , 000 ( THP1-ISG-Lucia ) cells per well in a white 96 well plate 24h before stimulation . Treatments were performed in quadruplicate in 50μL DMEM plus 2% FCS for 7h unless otherwise indicated . One-GLO lysis/luciferin reagent ( Promega ) was added at 1:1 to each well and luminescence measured on a Synergy plate reader ( BioTek ) . Coding sequences for human STING-C-terminal domain ( CTD; AA 137–379 ) and mouse STING-CTD ( AA 137–378 ) were cloned into pRSET-B vector ( Invitrogen ) and contained a 6xHIS tag for protein expression in E . Coli strain BL21 ( DE3 ) pLysS ( Promega ) . Sequences were verified before transforming bacteria , which were then grown in LB media at 37°C until the OD600 reached 0 . 8 . Protein expression was induced with 1mM IPTG at 16°C for 18h . After induction , the culture was centrifuged and the pellet resuspended in 50 mM NaH2PO4 , 150 mM NaCl ( pH 7 . 5 ) and 10% glycerol after which the cells were lysed by sonication . The recombinant soluble STING-CTD was purified by nickel-affinity chromatography ( Clontech laboratories ) after which it was further purified by gel-filtration chromatography using a HiPrep 16/60 Sephacryl S-100 HR column ( GE Healthcare Life Sciences ) . Protein was eluted in 50 mM NaH2PO4 , 150 mM Nacl ( pH 7 . 5 ) and the eluted fractions containing STING-CTD concentrated using an Amicon centrifugal filter ( 10 Kd molecular weight cut-off; Millipore ) . Aliquots of concentrated STING-CTD were immediately stored at -80°C . For thermal shift assay , 1 μg of recombinant human or mouse STING-CTD was used combined with various concentrations of G10 , 2’3’-cGAMP , or DMXAA along SYPRO Orange dye ( 1:1000 dilution ) in a 20μL reaction ( in triplicate ) . A StepOne Plus Real-time PCR system was used to acquire fluorescence . The samples were subjected to a temperature gradient of 25 to 99°C . The melting curves were plotted and Tm values determined by fitting the curves to Boltzmann sigmoidal equation using the GraphPad Prism 6 software . Three independent experiments were performed . C57Bl/6J mice ( 5–7 weeks of age , Jackson Laboratories ) were housed in cage units in an animal BSL3 facility , fed ad libitum , and cared for under USDA guidelines for laboratory animals . 25mg/kg DMXAA ( or DMSO alone ) was prepared in 50μL DMSO and injected intraperitoneally . Mice were challenged with 1000 PFU CHIKV via footpad injection in 20μL RPMI under isoflurane-induced anesthesia . Animals were euthanized at 72h post infection by isoflurane overdose . Blood was collected by cardiac puncture and serum viral loads titered on Vero cells in duplicate as described above . All animal procedures were conducted in accordance with and approved by the Oregon Health and Science University Institutional Animal Care and Use Committee ( IACUC ) under protocol 0913 . The Oregon Health and Science University IACUC adheres to the NIH Office of Laboratory Animal Welfare standards ( OLAW welfare assurance # A3304-01 ) .
|
STING is a pattern recognition receptor of cyclic dinucleotides as well as an innate immune adaptor protein that enables signaling from cytoplasmic receptors to the transcription factor interferon regulatory factor 3 . Initiation of these pathways leads to the expression of type I interferons and proteins associated with antiviral and antitumor immunity . Small molecules capable of triggering STING-dependent cellular processes are effective at blocking virus replication , enhancing vaccine efficacy , and facilitating immune response to cancer cells . Here we describe the first synthetic small molecule capable of activating STING-mediated signaling in human cells . In addition , we show that exposure of cells to the compound renders them refractory to replication by interferon-sensitive emerging Alphaviruses . In addition , in vivo stimulation of STING-dependent activity also blocks viremia of Chikungunya virus . Ultimately this work may lead to the utilization of STING as a target for multiple immune-mediated therapies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Characterization of a Novel Human-Specific STING Agonist that Elicits Antiviral Activity Against Emerging Alphaviruses
|
Highly social insects are characterized by caste dimorphism , with distinct size differences of reproductive organs between fertile queens and the more or less sterile workers . An abundance of nutrition or instruction via diet-specific compounds has been proposed as explanations for the nutrition-driven queen and worker polyphenism . Here , we further explored these models in the honeybee ( Apis mellifera ) using worker nutrition rearing and a novel mutational screening approach using the clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 ( CRISPR/Cas9 ) method . The worker nutrition-driven size reduction of reproductive organs was restricted to the female sex , suggesting input from the sex determination pathway . Genetic screens on the sex determination genes in genetic females for size polyphenism revealed that doublesex ( dsx ) mutants display size-reduced reproductive organs irrespective of the sexual morphology of the organ tissue . In contrast , feminizer ( fem ) mutants lost the response to worker nutrition-driven size control . The first morphological worker mutants in honeybees demonstrate that the response to nutrition relies on a genetic program that is switched “ON” by the fem gene . Thus , the genetic instruction provided by the fem gene provides an entry point to genetically dissect the underlying processes that implement the size polyphenism .
Highly social insects are characterized by caste dimorphism , with morphologically and physiologically distinct reproductive queens and more or less sterile workers [1–3] . In honeybees , the development of two distinct phenotypes is controlled by different nutrition , and it is a prominent example of developmental plasticity and polyphenism [4 , 5] . One major concern for the study of caste development involves explaining how a usually sterile worker and a queen that lays up to 2 , 000 eggs per day develop from different diet and feeding regimens [4 , 6 , 7] . Worker-destined larvae receive restricted amounts of a reduced sugar content diet ( worker jelly [WJ] ) , while queen-destined larvae receive large quantities of a sugar-rich diet ( royal jelly [RJ] ) [8–11] . WJ and RJ drive the development of female larvae in two distinct morphs . Workers have a five-day longer developmental time , lower body mass , two small ovaries containing few ovarioles , and mid- and hind-leg structures adapted for pollen collection and transport . Queens have a five-day shorter developmental time , larger body mass , and two large ovaries that contain many more ovarioles , and they lack the pollen collection structures on the legs . Two types of models have been proposed to explain how diets and feeding regimens mediate worker/queen development . The Nutrition/Growth model suggests that queen/worker development is driven by the amount of food and balance of nutrition [7 , 11 , 12] , which modulate a developmental program . Queen-destined larvae have abundant nutrition , and organ growth is only limited by the intrinsic program . Worker-destined larvae have a shortage of nutrition that restricts growth and influences metabolic parameters accordingly . In contrast , the Instruction model proposes that the RJ has a compound ( or compounds ) that instruct the development of queens [13–15] . In support of the Instruction model , research over the past decades has attempted to identify a single compound from RJ [12 , 14] that can determine queen development . A recent study provided evidence that the protein royalactin has queen-determining activity [15] . However , follow-up experiments in another laboratory were unable to repeat these results [7] , questioning the existence of a single determinant for queen development [4] . Gradually increasing the sugar levels of WJ and altering the composition of RJ-containing diets produced workers , intercastes , and eventually queens [9–11 , 16] , but it failed to rear only queens . The more continuous caste characteristics resulting from different feeding regimes [17] have been proposed in support of the Nutrition/Growth model . The RJ and the WJ produce different reaction norms of the general developmental program that determines the caste polyphenism . An alternative explanation is that the essential higher sugar levels for queen-destined larvae are a secondary effect and reflect the higher energy requirements for the faster and larger-growing queen organs of an otherwise instructed queen program . The rearing of larvae at day 5 in queenless colonies yielded bees with ovariole numbers that were discontinuous ( either more worker or queen-like distributed ) , while other queen and worker traits were either absent or present in a noncorrelated fashion [18] , suggesting two distinct states of the developmental program and the possible existence of regulatory switches [19] . One possible mechanism by which nutrients are sensed by bee larvae is the insulin/IGF signaling ( IIS ) and target of rapamycin ( TOR ) pathways , which link the abundance of nutrition with worker and queen differential gene expression [20–23] . Indeed , nutritional input can also influence growth and metabolic programs via the IIS and TOR pathways in mammals and other insects [24–26] . However , whether regulation of the IIS and TOR pathways drives caste differentiation or whether the regulation is a response to the activation of a queen developmental program is currently unknown . Consistent with the faster and larger growth of queens , gene expression studies have revealed the up-regulation of physiometabolic genes in queens , reflecting their higher metabolic rate [27 , 28] . Chromatin modifications and DNA methylation analyses have indicated distinct epigenetic states in worker- and queen-destined larvae , suggesting another level of regulatory control associated with caste-specific gene expression [29–31] . Here , we explored whether nutrition is the only factor directing size polyphenism and whether further genetic instruction from the sex determination pathway is required . To do so , we introduced a method to screen mutations directly in worker bees using the clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 ( CRISPR/Cas9 ) technique .
According to the Nutrition/Growth model , nutrition is the only driver of reduced reproductive organ size , the most prominent trait in caste development . Males , like queens , receive high amounts of sugar during larval development [32] and develop large reproductive organs unlike sterile worker bees . Gradually increasing the sugar levels of WJ produces intercaste development [9 , 10 , 16] . Hence , if a shortage of nutrition in the worker diet ( and reduced sugar levels ) is the only driving component , we would expect that this diet would also mediate the size reduction of reproductive organs in males . We manually reared genetic females and males on worker nutrition [16 , 33] and compared their phenotypes with those of workers and genetic males reared in the colony ( Fig 1 and S1 and S2 Tables ) . The reproductive organs of genetic female bees raised on worker nutrition either inside the colony ( n = 14 ) or manually outside ( n = 15 ) were equivalent in size ( Fisher’s exact test , df = 1 , P = 1 ) . In both laboratory- and colony-reared genetic females , there were few ovarioles , and the size of each ovary was small compared with the size of the heads ( Fig 1 and S1 Table ) . This contrasts with the large ovaries of the female larvae fed a queen diet in the hive ( queens alone cannot be consistently reared under laboratory conditions [7]; see Fig 4A and 4B as an example of a queen phenotype ) . This result indicates that our manual feeding regime mirrors the effect of a worker diet in the hive [16 , 33] . To examine whether only the balance and amount of nutrition ( low amount of sugar ) determine small reproductive organs , we reared genetic male larvae on worker nutrition in the laboratory and compared these with males that received high amounts of sugar in the colony [32] . Genetic males that were reared on the worker nutrition diet had large male reproductive organs ( Fig 1 and S2 Table ) . They were equivalent in size ( n = 20 ) to the males obtained from the colony ( n = 8 ) that were reared on drone nutrition ( Fisher’s exact test , df = 1 , P = 1 ) . These results indicate that worker nutrition ( and a shortage of sugar ) is not the only requirement for the size polyphenism , suggesting input from the sex determination pathway . We next established a method that enables the mutational screening of sex-determining genes directly in worker bees using the CRISPR/Cas9 method [34–36] . Following traditional mutant approaches , we would need to produce mutant queens and drones that need to be crossed to generate double-mutant worker bees . If we could mutate all nuclei in the embryo , we would be able to directly rear mutated worker bees without maintaining colonies and performing crossings . To examine whether we could mutate worker bees entirely using the CRISPR/Cas9 method , we tested different embryonic injection conditions . To determine the robustness of this approach , we studied at least two sites for three genes , the doublesex ( dsx ) , fruitless ( fru ) , and loc552773 genes ( S1 Fig ) . Only the dsx gene was used later on for phenotyping . We injected into the anterior embryos of very young female embryos ( 0 to 1 . 5 hours after egg deposition ) [37] . We tested a set of single guide RNAs ( sgRNAs; S3 Table ) at different concentrations and observed that we repeatedly mutated each injected embryo . The fragment length ( FL ) and sequence analyses of the amplicons in larval stage 1 larvae revealed that up to 100% of the fru and dsx and 60% of the loc552773 target embryos were mutated ( Tables 1 and S4 and S5 and Fig 2 ) . The wild-type ( WT ) allele was consistently not detected in 30 of the 39 mutated larvae ( 77% ) , suggesting that all nuclei ( to the level of detection ) and both alleles in the larvae were mutated ( generating double mutants ) . More than two mutated sequence variants were detected in a single larva ( 3% ) , while singly mutated sequences together with the WT allele were detected in 8 larvae ( 20% ) ( S4 and S5 Tables ) . Indels occurred most frequently between the 5 bp to 1 bp range , with 44% of mutations being deletions and 20% resulting in insertions ( S5 and S6 Tables ) . All mutations occurred at the designated target site . Therefore , our results on the adjustments demonstrate that nearly 80% of the injected embryos had mutations on both alleles ( double mutants ) affecting the bee entirely ( absence of mosaicism ) . This high proportion enabled us to screen for mutant effects of the sex-determining genes directly in the injected bees . To examine whether the feminizer ( fem ) gene is required for small size polyphenism , we mutated the gene in genetic females and reared them with worker nutrition . The fem gene instructs female development and maintains the female signal during development , as revealed from fem interference RNA ( RNAi ) knockdown and mosaic studies using a non–worker-specific diet for bee rearing [19 , 38] . The Fem protein is encoded by female-specific spliced fem transcripts but not the male spliced variant , which harbors an early stop codon [19] ( Fig 3A ) . The female splicing of fem is directed by the complementary sex determiner ( csd ) gene when the genotype is heterozygous ( Fig 3A ) [39] . If the fem gene is required for small size polyphenism , we would expect that worker nutrition cannot drive size reduction when fem is inactive . If the fem gene is dispensable , worker nutrition would drive size reduction even when the fem gene is inactive . We induced mutations at two target sites in the first half of the female open reading frame ( ORF ) of the fem gene with fem-sgRNA1 and fem-sgRNA2 ( S1 and S2 Figs ) and reared genetic females with worker nutrition to larval stage 5 . Fifteen percent of the reared and injected genetic females ( heterozygous for the csd gene; S7 Table ) were double mutants for nonsense mutations as revealed from the sequenced amplicons ( S8 Table and S2 Fig ) . These double mutants ( n = 4 ) had large gonads ( Fig 3B and 3D ) compared with the small gonads of WT genetic females reared on worker nutrition ( n = 38 , Fisher’s exact test , df = 1 , P < 0 . 001 , S9 Table ) . The large gonads in the mutants were of the male type . They consisted of packed layers of multiple testioles of the same size as those of the males reared on worker nutrition ( Fig 3B ) and those of the males in the colony ( Fig 1 ) . The female fem mutants lost the female dsx transcript and only displayed the male dsx transcript ( Fig 3C ) , demonstrating that the mutant bees entirely switched in their development from female to male identity . These results indicate that fem is required for size polyphenism or that size polyphenism relies on the intrinsic program of the female differentiating tissue induced by fem . To examine the role of female dsx on size polyphenism of the reproductive organ , we mutated the dsx gene in genetic females and reared them on worker nutrition . If dsx is dispensable , we would expect small size polyphenism even when dsx activity is compromised . In Drosophila melanogaster , the dsx gene essentially controls , beside the reproductive organs , all aspects of somatic sexual differentiation [40 , 41] , and it controls at least reproductive organ development in other insects that belong to different insect orders , including hymenopteran insects [42–45] . The dsx transcripts in honeybees are sex-specifically spliced by the presence of the Fem protein in females and the absence of the Fem protein in males [19] ( Fig 3A ) . The sexual splice variants encode a transcription factor with an intertwined zinc-containing DNA binding ( DM ) domain and male- and female-specific termini at the carboxyl end [46–50] . We mutated the dsx gene at two target sites in the non–sex-specific expressed N-terminal portion . dsx-sgRNA2 targeted the DM domain , whereas dsx-sgRNA6 targeted a downstream region in exon 3 ( S1 Fig ) . The treated genetic females were reared on worker nutrition and were examined for morphological changes of the reproductive organ and head . Genotyping of the mutated bees with morphological changes via next-generation sequencing ( NGS ) of the amplicons revealed that they were regularly double mutants with an approximate ratio of 1:1 , suggesting that the mutations belong to the two chromosomes of the diploid set . If we detected more than two sequence variants per bee , we excluded these bees from further phenotype analysis as they were genetic mosaics ( e . g . , a mosaic of differently mutated cells ) . Eleven ( 17% ) of the adult or pupal bees had intersex morphology in the reproductive organs compared with the WT genetic females ( S10 Table ) . No effect was observed for the heads . The following mutations were the most common ones in the genetic females: ( i ) different nonsense mutations that introduced new stop codons at various positions in exons 2 and 3 , ( ii ) deletions of amino acids in the DM domain mainly the histidine codon at amino acid position 68 ( ΔH68 ) , and ( iii ) deletion of the alanine codon ( ΔA191 ) at amino acid position 191 ( Fig 4 with the deduced amino acid sequences and S3 Fig with the detected nucleotide sequences ) . The ΔH68 mutation removes a histidine of the DM domain that is essential for the zinc binding and DM domain functions [47 , 51] and that is conserved between vertebrates and invertebrates ( S4 Fig ) . The intersex reproductive organs were all of the same small size ( n = 11 ) as the worker reproductive organs in WT genetic females that were manually reared on worker nutrition ( n = 17 , Table 2 , Fisher’s exact test , df = 1 , P = 1 ) . The small intersex reproductive organs displayed either male gonads with poorly or non–sex-specifically differentiated duct systems ( n = 4 ) , as observed in stop200/stop202 and ΔH68/stop91 genetic females ( arrows in Figs 4 and S5 ) . The potentially earlier developmental stage of some of these mutant bees cannot explain why these male-like gonads are so small because the distinct size differences of male and worker gonads are also present at earlier pupal stages ( S6 Fig ) . In other cases , the reproductive organs were underdeveloped ( n = 7 ) , and the oviducts were consistently misshaped while the ovarioles were repeatedly missing , as identified in ΔH68/ΔH68 , ΔH68/stop73 , ΔH68/stop75 , and ΔA191/stop202 genetic females ( Figs 4 and S5 ) . The heads of the mutant genetic females with intersex reproductive organs were all of worker type ( n = 11 , Fig 4 and S10 Table ) , suggesting that dsx is not required for sexual development of the head . The results of the consistently small , intersex reproductive organs with varying degrees of masculinization suggest that dsx is not required for size polyphenism .
Caste polyphenism in honeybees is determined by different nutrition with the size of the reproductive organ as an important trait . Most studies suggest that the balance and amount of nutrition ( Nutrition/Growth model ) drive the size polyphenism between queens and workers . Our genetic and rearing results now suggest that the response to nutrition relies on a genetic program that is switched on by the fem gene . The genetic females with a mutant fem gene show large size reproductive organ ( large polyphenism ) , while WT genetic females ( Fig 5A ) reared on the same worker nutrition have only small reproductive organs ( small polyphenism ) . Genetic females that have a mutated dsx gene ( operating downstream of fem ) do show small reproductive organs ( small size polyphenism; Fig 5A ) . dsx mutants produce intersex reproductive organs and male-like gonads that are all of small size , demonstrating that small size does not rely on female development of the tissue . The small size polyphenism also did not result from dsx malfunction because ( i ) small phenotypes were consistently observed irrespective of the different degrees of dsx malfunctions we introduced by missense and nonsense mutations ( Fig 4 ) and ( ii ) dsx mutations in other insects did not influence the size of the reproductive organs [42 , 52 , 53] . Thus , the results together suggest that the fem gene is required for the small size polyphenism . We conclude that the fem gene must be switched “ON” so that size polyphenism can be executed ( Fig 5B ) . The essential role of the fem gene in small size polyphenism assigns a further key function to the fem gene . Previous studies demonstrated that the fem gene is also required to ( i ) induce entire female development in response to the primary signal csd [19 , 38] and to ( ii ) maintain the female signal during development via a positive regulatory feedback loop [19] . Whether fem also instructs the large size polyphenism of queens needs further functional testing once a queen-only rearing protocol has been developed for the laboratory [7] . The genetic instruction via the fem gene provides an entry point to dissect nutrition-mediated control . Our results suggest that the fem gene switches “ON” the machinery that is required for sensing the worker nutrition and for implementing the size polyphenism . Because the fem gene encodes a serine arginine rich ( SR ) -type protein , the direct targets of the fem gene involved in size polyphenism may also be activated by sexual splicing . The fem-controlled candidate genes can be functionally tested by determining whether they affect the size polyphenism . The function will be directly tested in mutated genetic females as demonstrated in this study . Our mutant analysis further demonstrate that dsx controls female differentiation of the reproductive organs . The mutant honeybee phenotypes of the reproductive organs in honeybees yielded similar phenotypes as in female D . melanogaster . Female dsx-mutant fruit flies have reproductive organs of varying intersex phenotypes . The organs are often underdeveloped with occasionally developed ovaries , but are frequently of the “male type” [52 , 54 , 55] . The internal duct system can develop into a mixture of female/male or single poorly differentiated ducts [52] . RNAi-mediated knockdown studies on the beetle Tribolium molitor , housefly Musca domestica , and sawfly Athalia rosae , as well as conditional expression and CRISPR/Cas9 experiments on the silkworm Bombyx mori , have revealed sex-related effects on internal reproductive organ development [42–46 , 53] . Our results support a conserved role for dsx in the sexual development of the reproductive organ . However , in honeybees there is a nutrition-driven size control of reproductive organ development that operates upstream of or in parallel with dsx-regulated sexual development . The first CRISPR/Cas9-induced morphological mutants in honeybees introduced a new genetic screening method for worker bees . We efficiently induced mutations in injected embryos using the CRISPR/Cas9 method [34 , 35] and directly screened for somatic mutations in the reared honeybees ( somatic mutation approach ) . Up to 100% of the embryos were mutated , and mosaicism among the mutated embryos was rare ( <10% ) . The previous studies in honeybees using CRISPR/Cas9-induced mutations report on 1 out of 2 queens with only 12% and 2 out of 4 queens with only 5% and 10% mutant drone offspring , suggesting that the previously published method has a substantial lower rate and produced strong mosaicism in the queens [36 , 56] . These previous studies generated no worker bees that would require further crossing experiments . With very early embryonic injections [37] and a selection step to identify the most efficient sgRNAs and Cas9 concentrations , we generated mutation rates of up to 100% and no mosaicism in worker bees directly . The rearing of the mutated embryos to worker bees was performed under controlled conditions in the laboratory [16 , 33] . This required no rearing of queens and drones and crossing experiments . The procedure was demonstrated for mutations at two target sites for two genes and their morphological changes ( Figs 3 and 4 ) . The absence of mosaicism and completeness of mutagenesis of this procedure were shown by the results that most mutated bees lost the WT allele ( they were double mutants; Figs 2 , 3D and 4 ) and that double fem nonsense mutations produced an entire female to male switch , including dsx splice products ( Fig 3C ) . This somatic mutation approach does not require further crossing experiments and laborious maintenance of hundreds of colonies and therefore offers the prospect of larger genetic screens in honeybees . In other insects in which somatic mutation approaches have been applied [57 , 58] , the adults were genetic mosaics in which parts of the butterfly wing were WT while other parts were mutated . Enhancing the efficiency of mutagenesis can thus provide an opportunity for somatically testing gene functions in insects that are not yet genetically trackable .
Cas9 mRNA was synthesized from the Cas9 gene [59] ( Vector MLM3613 , ID #42251 , Addgene , Cambridge , MA ) using a linearized plasmid via the T7 promoter and the mMESSAGE mMACHINE Kit ( Ambion , Darmstadt , Germany ) . mRNAs were polyadenylated using the Poly ( A ) Tailing Kit ( Ambion ) . Target sites for the sgRNAs were identified via Optimal Target Finder software ( http://tools . flycrispr . molbio . wisc . edu/targetFinder/ ) . sgRNAs were 20 nt long with a G nucleotide at the 5´ end . sgRNAs with no off-target effects or with at least three nucleotide mismatches to alternative target sites were selected . sgRNAs were generated via PCR without a template using two overlapping oligonucleotide sequences containing the sequence of the T7 RNA polymerase transcription start site , the gene-specific target site and the Cas9 protein-binding site . sgRNAs were synthesized using a RiboMax Kit ( Promega , Madison , WI ) according to the manufacturer’s instructions . RNAs were purified using the MEGAclear Kit ( Ambion ) . Embryos were microinjected 0 to 1 . 5 hours after egg deposition [19 , 37 , 60] using 53-mm injection pipettes ( Hilgenberg , Malsfeld , Germany ) . Cas9 mRNA or protein ( New England Biolabs , Ipswich , MA ) was applied at 400 to 2 , 000 ng/μl and mixed with sgRNAs using a molar ratio of 1:2 to 1:0 . 75 . The number of injected embryos that hatch can vary greatly between experiments and sgRNAs ( 5% to 40% ) . Rearing was performed using a mass rearing technique for the worker bees [16 , 33] . Freshly hatched larvae were provisioned only once with the worker larval diet ( 50%–53% RJ , 4% glucose , 8% fructose , 1% yeast extract , and 30%–34% water ) , approximately 120 to 170 mg of which was consumed [16 , 33] . The larvae were incubated at 34°C and 90% humidity until the larval stage 5 or to adults . For pupal rearing we also used a slightly different diet for larvae at stage 5 ( 50 mg diet 2 [50% RJ , 12% fructose , 6% glucose , 2% yeast extract , and 30% water] ) . For genotyping , genomic DNA was isolated from freshly hatched L1 or L5 larvae [61] using the peqGOLD Tissue DNA Mini Kit ( VWR , Darmstadt , Germany ) . RNA was isolated using the TRIZOL method ( Thermo Scientific , Braunschweig , Germany ) , and cDNA was synthesized using the RevertAid First Strand cDNA Synthesis Kit ( Thermo Scientific ) . Second-strand cDNA synthesis was performed by adding 10 μl of 10× DNA Polymerase Buffer , 40 U DNA Polymerase I , 0 . 8 U Ribonuclease H , and 65 . 68 μl of dH2O to 20 μl of the cDNA first-strand synthesis product . Double-stranded cDNA was purified using the EZNA Cycle Pure kit ( Omega Bio-Tek Inc . , Norcross , GA ) . All mutant bees were genotyped by sequencing the amplicons of the targeted site . PCR amplifications were performed using standard conditions [62] and GoTaq polymerase ( Promega ) . Oligonucleotide sequences were synthesized at Eurofins ( Ebersberg , Germany ) . Amplicons were either cloned and sequenced ( Sanger sequencing [Eurofins] ) or sequenced via NGS . NGS index PCR was performed using the Nextera XT Index Kit ( Illumina , San Diego , CA ) , and purification of the Index PCR products was performed using Agencourt AMPure XP beads ( Beckman Coulter , Brea , CA ) . NGS was performed on an Illumina MiSeq system using the MiSeq Reagent Kit version 2 ( 500 cycles; Illumina ) , generating 800 , 000 paired-end reads with a read length of 2 × 250 bp , resulting in approximately 15 , 000 paired-end reads per sample . We removed contamination by removing sequences that were less frequent than 5% . The FLs of hexachlorofluorescein ( HEX ) -labeled amplicons were determined using an ABI 3130XL Genetic Analyzer ( Applied Biosystems , Darmstadt , Germany ) and Peak Scanner software ( Thermo Scientific ) . For the fem mutants , we conducted fragment and sequence analysis on the amplicons of the cDNAs to ensure that the many fem-related sequences observed at the genomic fem locus ( derived from duplication events ) [63] were not amplified .
|
In honeybees , nutrition drives dimorphic size development of reproductive organs in fertile queens and sterile workers . The first induced morphological mutants in honeybees demonstrate that this developmental plasticity requires a genetic program that is switched “ON” by the feminizer ( fem ) gene .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"honey",
"bees",
"sociology",
"social",
"sciences",
"animals",
"reproductive",
"physiology",
"social",
"systems",
"developmental",
"biology",
"mutation",
"nutrition",
"nonsense",
"mutation",
"sequence",
"motif",
"analysis",
"bees",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"hymenoptera",
"bioinformatics",
"life",
"cycles",
"insects",
"arthropoda",
"eukaryota",
"physiology",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"larvae",
"organisms"
] |
2019
|
A genetic switch for worker nutrition-mediated traits in honeybees
|
Nucleotide-binding oligomerization domain ( NOD ) 2 is a cytosolic protein that plays a defensive role in bacterial infection by sensing peptidoglycans . C5a , which has harmful effects in sepsis , interacts with innate proteins . However , whether NOD2 regulates C5a generation during sepsis remains to be determined . To address this issue , cecal ligation & puncture ( CLP ) -induced sepsis was compared in wild type and Nod2−/− mice . Nod2−/− mice showed lower levels of C5a , IL-10 , and IL-1β in serum and peritoneum , but higher survival rate during CLP-induced sepsis compared to wild type mice . Injection of recombinant C5a decreased survival rates of Nod2−/− mice rate during sepsis , whereas it did not alter those in wild type mice . These findings suggest a novel provocative role for NOD2 in sepsis , in contrast to its protective role during bacterial infection . Furthermore , we found that NOD2-mediated IL-10 production by neutrophils enhanced C5a generation by suppressing CD55 expression on neutrophils in IL-1β-dependent and/or IL-1β-independent manners , thereby aggravating CLP-induced sepsis . SB203580 , a receptor-interacting protein 2 ( RIP2 ) inhibitor downstream of NOD2 , reduced C5a generation by enhancing CD55 expression on neutrophils , resulting in attenuation of polymicrobial sepsis . Therefore , we propose a novel NOD2-mediated complement cascade regulatory pathway in sepsis , which may be a useful therapeutic target .
Sepsis is a complex dysregulated inflammatory response in infection , which causes multiple organ dysfunction and coagulopathy , often resulting in death [1] , [2] . Pro-inflammatory cytokines contribute to an overwhelming inflammatory immune response in the early phase of sepsis , whereas anti-inflammatory cytokines are involved in the late-phase immune response [1] , [3] . Studies have shown that C5a , a complement protein , has harmful effects during sepsis , although the complement is crucial for clearance of infectious agents [4]–[8] . During sepsis , C5a causes multiple organ failure , cardiomyopathy , and imbalanced coagulation [9] . Therefore , C5a is generally accepted as a crucial target for therapeutic approaches in sepsis . Nevertheless , the mechanisms by which C5a is regulated in sepsis remain unclear . Nod-like receptors and toll-like receptors ( TLRs ) are a family of innate proteins that trigger innate immune activation by recognizing pathogen-associated molecular patterns [10] . NOD2 genetic mutations in humans have been implicated in Crohn's disease , Blau syndrome , sarcoidosis , and graft-versus-host disease [11]–[14] . With respect to bacterial infection , NOD2-mediated peptidoglycan sensing regulates mononuclear cell recruitment and chemokine production , which promotes clearance of Streptococcus pneumonia [15] . Moreover , NOD2 activates the autophagy process , thereby confining intracellular bacteria within intracellular autophagosomes and subsequently restricting the infection [16] , [17] . Consistent with these data , Nod2-deficient mice are susceptible to Listeria monocytogenes , indicating that NOD2 plays a defensive role during bacterial infection [18] . Moreover , very recent study suggests that polymorphisms in the NOD2/Card15 gene might be related with susceptibility to sepsis in children [19] . However , it remains unclear whether NOD2 play crucial role in the pathogenesis of sepsis . It has been demonstrated that various TLR ligands regulate cytokine production in a complement-dependent manner , suggesting that crosstalk between innate proteins and the complement system makes a crucial contribution to immune response regulation in vivo [20] . Thus , we hypothesize that NOD2 regulates C5a generation via crosstalk with the complement system during sepsis . To address this hypothesis , we investigated whether NOD2 regulates C5a generation during sepsis . Our results indicate that NOD2-mediated signals enhance C5a generation by suppressing CD55 expression on neutrophils through IL-1β-dependent or IL-1β-independent IL-10 production during polymicrobial sepsis .
To investigate whether NOD2 regulates C5a generation during sepsis , we performed CLP in wild-type ( WT ) and Nod2−/− mice . Serum and peritoneal C5a levels were lower in Nod2−/− than in WT mice during sepsis , whereas C3a levels were similar ( Fig . 1A ) . All Nod2−/− mice were alive up to 10 days after CLP , whereas all WT mice died within 2 days . However , ELISA system for C5a might detect C5 in some cases , although cross reactivity for C5a and C5 in detection system does not usually occur . Thus , to confirm harmful effects of NOD2-mediated C5a generation on sepsis , we injected WT and Nod2−/− mice with recombinant ( r ) C5a during CLP-induced sepsis . Injection of Nod2−/− mice with rC5a decreased survival rates during sepsis , whereas rC5a had no effect on survival of WT mice ( Fig . 1B ) . These findings suggest that NOD2-mediated C5a generation contributes to sepsis development and severity . NOD2 was recently reported to affect composition of the host microbiota in mice and humans [21] . To rule out differences in the cecal bacterial composition of WT and Nod2−/− mice might affect sepsis , cecal contents obtained from WT or Nod2−/− mice were injected i . p . into WT and Nod2−/− mice after their ceca had been ligated but not punctured and WT and Nod2−/− mice were cohoused for 4 weeks . The survival rates after injection of Nod2−/− or WT cecal contents were lower in WT mice than Nod2−/− mice ( Fig . S1 ) . Moreover , survival rates of cohoused WT mice were lower than cohoused Nod2−/− mice during CLP-induced sepsis ( Fig . S1 ) , indicating that difference in intestinal microbiota between two mouse groups minimally contributes to septic responses because gut microbiota of cohoused mice are replaced by each other [22] . Collectively , these findings suggest that NOD2-mediated signals enhance generation of C5a , but not C3a , thereby enhancing the systemic inflammatory response . To determine the effect of C5a on immune responses in WT and Nod2−/− mice during sepsis , we investigated neutrophil dysfunction by measuring the responsiveness of immune cells to LPS , phagocytic activity , CFUs , and serum D-dimer levels after CLP . Total peritoneal cells obtained from Nod2−/− mice 24 h after CLP produced higher IL-6 and TNF-α levels to LPS than did WT peritoneal cells ( Fig . 1C ) . Moreover , serum D-dimer levels in Nod2−/− mice were lower than those in WT mice ( Fig . 1D ) . However , total peritoneal cells obtained from Nod2−/− mice engulfed more FITC-conjugated Escherichia coli than did WT peritoneal cells ( Fig . 1E ) . Consistent with these findings , culturable bacterial CFU levels in blood and liver homogenates were higher in WT than in Nod2−/− mice ( Fig . 1F ) . rC5a administration to Nod2−/− mice with CLP reversed the response to LPS by peritoneal cells and serum D-dimer levels , but not phagocytosis activity or bacterial CFUs . These findings suggest that NOD2-mediated signals trigger immune cell dysfunction and coagulopathy by enhancing C5a levels during sepsis , whereas the NOD2-mediated immune response regulates bacterial phagocytic activity and CFU levels in a C5a-independent manner . To explore the mechanism by which NOD2 enhances C5 generation during sepsis , the serum and peritoneal levels of various cytokines in WT and Nod2−/− mice were estimated after CLP . The serum and peritoneal IL-1β and IL-10 levels of WT mice were significantly higher than those of Nod2−/− mice , whereas IL-6 and IFN-γ levels in WT mice were similar to those in Nod2−/− mice ( Fig . 2A ) . Serum TNF-α levels were higher in WT than Nod2−/− mice , whereas its peritoneal levels were similar in two mouse groups . To estimate IL-1β and IL-10 production by peritoneal cells , we obtained total peritoneal cells from WT and Nod2−/− mice 4–6 h after injection with thioglycollate . Upon MDP , an NOD2 agonist treatment , WT peritoneal cells produced IL-1β and IL-10 , whereas NOD2-deficient cells produced minimal IL-1β and IL-10 ( Fig . 2B ) . A kinetic analysis revealed that IL-1β and IL-10 levels in peritoneal fluid peaked 4 and 12 h after CLP , respectively , and then decreased gradually ( Fig . 2C ) . Serum IL-1β levels peaked 12 h after CLP and were significantly higher in WT than Nod2−/− mice at 24 h , whereas serum IL-10 levels in WT mice increased continuously from 4 to 24 h after CLP , and were significantly higher than those in Nod2−/− mice . Subset analysis revealed that F4/80−Ly-6G+ neutrophils and F4/80+Ly-6G− macrophages were major cells infiltrated into peritoneum during sepsis ( Fig . S2 ) and the numbers of these cells was similar in Nod2−/− and WT mice ( data not shown ) . The percentages of neutrophils peaked and macrophages showed the lowest percentages in Nod2−/− and WT mice 4 h after CLP , and the percentages of these cells were similar in WT and Nod2−/− mice before , 4 , 12 , and 24 h after CLP ( Fig . S2 ) . Based on these findings , NOD2 expression was investigated in sorted F4/80+Ly-6G− macrophages and F4/80−Ly-6G+ neutrophils from peritoneum of WT mice with CLP ( Fig . 2D ) . Real-time PCR analysis revealed that NOD2 expression in F4/80−Ly-6G+ neutrophils was constitutive and sustained during sepsis , whereas F4/80+Ly-6G− macrophages showed low expression levels of NOD2 before , and 4 and 12 h after CLP , but highly expressed NOD2 24 h after CLP , which was consistent with expression pattern of NOD2 in blotting assay ( S3A and B ) . These findings indicate that F4/80−Ly-6G+ neutrophils rather than F4/80+Ly-6G− macrophages in the peritoneum predominantly express NOD2 during early and intermediate stages of sepsis . Consistent with the kinetics of the IL-1β and IL-10 levels , peritoneal F4/80−Ly-6G+ neutrophils from WT mice showed high IL-1β mRNA levels at 4 and 24 h , but low transcriptional levels at 12 h ( Fig . 2E ) . In contrast , peritoneal F4/80+Ly-6G− macrophages produced high levels of IL-1β at 24 h . Unlike IL-1β , peritoneal F4/80−Ly-6G+ neutrophils from WT mice predominantly produced IL-10 at 12 h . Although the kinetics of IL-1β and IL-10 production were similar in Nod2−/− and WT mice , individual cytokine levels were much lower in the former , which were consistent with results in intracellular staining ( Fig . 2F ) . To investigate whether peritoneal neutrophils produce IL-1β and IL-10 during sepsis , peritoneal F4/80−Ly-6G+ neutrophils and F4/80+Ly-6G− macrophages from Nod2−/− and WT mice with CLP were obtained , sorted , and cultured for 24 h without stimulation . F4/80−Ly-6G+ neutrophils from WT mice produced larger amount of IL-1β and IL-10 than WT F4/80+Ly-6G− macrophages did ( Fig . S4A ) . In contrast , F4/80−Ly-6G+ neutrophils and F4/80+Ly-6G− macrophages from Nod2−/− mice minimally produced IL-1β and IL-10 . Furthermore , neutrophil depletion using anti-Ly-6G mAb in WT mice reduced the levels of IL-1β and IL-10 in serum and peritoneum , which was dependent on depletion time points ( Fig . S4B and C ) . Combined in vitro and depletion experiments , it is suggested that peritoneal neutrophils rather than macrophages directly produce IL-1β and IL-10 at different time points during sepsis , although neutrophils might interact with macrophages or monocytes to produce various cytokines during CLP-induced sepsis . To estimate cytokine-mediated effector functions of immune cells in sepsis , we measured the expression of IL-1β and IL-10 receptors on peritoneal cells . Both IL-1β and IL-10 receptors were similarly expressed on total peritoneal cells of WT and Nod2−/− mice with CLP ( Fig . 3A ) . Next , to determine whether NOD2-mediated IL-1β and IL-10 production plays a critical role in C5a generation during sepsis , we administered rIL-1β or rIL-10 to WT or Nod2−/− mice 4 h or 12 h after CLP , respectively . Administration of rIL-1β or rIL-10 enhanced serum and peritoneal C5a , but not C3a , levels ( Fig . 3B ) . Furthermore , rIL-1β or rIL-10 injection into Nod2−/− mice reduced survival rates during sepsis , whereas these recombinant cytokines did not affect the survival of WT mice ( Fig . 3C ) . These findings indicate that NOD2-mediated IL-1β and IL-10 production by neutrophils contributes to the pathogenesis of sepsis by enhancing C5a generation . The IL-1β autocrine loop amplifies the NOD2-mediated induction of pro- and anti-inflammatory cytokines in human monocyte-derived macrophages [23] . Moreover , our experiments and other studies demonstrated that neutrophils were a major subset to produce IL-1β and IL-10 and highly expressed NOD2 during CLP-induced sepsis [24] , [25] . These findings led us to hypothesize that IL-1β-dependent IL-10 production by neutrophils may occur in the NOD2-mediated immune response during sepsis . To address this , total peritoneal cells obtained from WT , Nod2−/− , or Il-1r−/− mice with CLP were incubated with rIL-1β . IL-10 levels in culture fractions of WT and Nod2-deficient cells were increased in an IL-1β dose-dependent manner , whereas those of Il-1r-deficient cells were not altered ( Fig . 4A ) . Administration of rIL-1β to Nod2−/− mice enhanced serum and peritoneal IL-10 levels during sepsis ( Fig . 4B ) . Furthermore , Il-1r−/− mice showed reduced serum and peritoneal IL-10 levels compared to WT mice with CLP , whereas IL-1β levels were similar ( Fig . 4C ) . These results suggest that peritoneal neutrophils produce IL-10 via IL-1β-dependent signaling during sepsis . Next , to explore whether IL-10 production regulates C3a and C5a levels during sepsis , we measured those levels and survival rates in Il-10−/− and Il-1r−/− mice with CLP . Both groups showed higher survival rates and lower C5a , but not C3a levels than WT mice during sepsis ( Fig . S5 , Fig . 4D , E ) . The serum and peritoneal C5a levels were not altered by administration of rIL-1β to Il-10−/− mice ( Fig . 4F ) , whereas injection of recombinant C5a reduced survival rates in Il-10−/− and Il-1r−/− mice during CLP-induced sepsis ( Fig . S5 ) . Moreover , rIL-1β did not alter the responsiveness to LPS by peritoneal immune cells from Il-10−/− mice with CLP ( Fig . 4G ) . These findings suggest that NOD2-mediated IL-1β-dependent IL-10 production by neutrophils regulates C5a generation during sepsis , although it is completely ruled out that other peritoneal immune and non-immune cells might contribute to C5a generation . Several immune molecules , such as CD55 and CR1/2 , on the surface of immune cells regulate the complement network by inhibiting complement generation [8] . Thus , to functionally link the expression of these molecules to NOD2-mediated C5a generation during sepsis , CD55 and CR1/2 expression levels on gated peritoneal F4/80−Ly-6G+ neutrophils and F4/80+Ly-6G− macrophages were measured . Peritoneal F4/80+Ly-6G− macrophages minimally expressed CD55 and CR1/2 in WT and Nod2−/− mice with CLP ( Fig . 5A ) , suggesting that expression modulation of these molecules on macrophages minimally contributes to NOD2-mediated C5a generation during sepsis . In contrast , CD55 expression levels on peritoneal F4/80−Ly-6G+ neutrophils from Nod2−/− and Il-10−/− mice were higher than those of WT mice 24 h after CLP , whereas CR1/2 was not detected on peritoneal F4/80−Ly-6G+ neutrophils from WT and Nod2−/− mice ( Fig . 5A , B ) . Moreover , rIL-10 or rIL-1β administration to Nod2−/− mice decreased CD55 expression on peritoneal F4/80−Ly-6G+ neutrophils during sepsis ( Fig . 5A ) . However , IL-1β administration to Il-10−/− mice did not decrease CD55 expression on peritoneal F4/80−Ly-6G+ neutrophils ( Fig . 5B ) . Furthermore , anti-IL-10 receptor mAb increased CD55 expression on peritoneal F4/80−Ly-6G+ neutrophils of WT and Nod2−/− mice administered rIL-10 ( Fig . 5C ) . These findings suggest that NOD2-mediated IL-1β-dependent IL-10 production decreases CD55 expression on peritoneal neutrophils , which regulates C5a generation during sepsis . However , CD55 expression levels on peritoneal F4/80−Ly-6G+ neutrophils in Il-1r−/− mice were intermediate between those on cells of WT and Nod2−/− mice with CLP , indicating that NOD2-mediated IL-10 suppresses CD55 expression on peritoneal F4/80−Ly-6G+ neutrophils in both an IL-1β-dependent , and an IL-1β-independent manner during sepsis ( Fig . 5D ) . In complement system , CD55 inhibits complement convertase activity by dissociating Bb factor from convertase attached on cell membrane [26] . Bb factor expression was minimally detected in Nod2-deficient total peritoneal cells , whereas Bb factor was highly expressed in WT cells ( Fig . 5E ) . Upon incubation with WT mouse serum , total WT peritoneal cells generated more C5a than Nod2-deficient cells ( Fig . 5F ) . These findings suggest that NOD2-mediated suppression of CD55 expression on peritoneal neutrophils enhances C5a generation during sepsis . To confirm this suggestion in vivo , soluble CD55 protein was administered to WT , Nod2−/− , WT mice depleted neutrophils , or Nod2−/− mice given rIL-10 during sepsis . Soluble CD55 protein decreased serum and peritoneal C5a levels in WT and Nod2−/− mice given rIL-10 , resulting in high survival rates ( Fig . 5G and H ) . Neutrophil depletion using anti-Ly-6G mAb increased serum and peritoneal C5a levels in Nod2−/− mice during CLP-induced sepsis , which was reduced by administration of soluble CD 55 ( Fig . S6 ) . Taken together , these data suggest that NOD2-mediated IL-1β-dependent and/or IL-1β-independent IL-10 production enhances C5a generation by suppressing CD55 expression on neutrophils , thereby aggravating sepsis . Upon activation , NOD2 oligomerizes and recruits RIP2 via CARD-CARD interaction , triggering IκB phosphorylation and NF-κB activation [27] , [28] . SB203580 , an inhibitor of RIP2 and P38 [29] , inhibited MDP-mediated IL-1β and IL-10 production by total peritoneal cells from WT mice ( Fig . 6A ) . Furthermore , Nod2−/− mice showed lower levels of RIP2 expression and phosphorylation , and P38 phosphorylation in total peritoneal cells during sepsis than did WT mice ( Fig . 6B ) . Upon SB203580 injection , WT mice exhibited reduced RIP2 expression and phosphorylation , and P38 phosphorylation in total peritoneal cells and serum and peritoneal IL-1β , IL-10 , and C5a levels during sepsis , whereas these were unaffected in Nod2−/− mice ( Fig . 6B and C ) . Moreover , SB203580 administration to WT mice increased CD55 expression levels in F4/80−Ly-6G+ neutrophils , and increased survival rates during sepsis ( Fig . 6D and E ) . These findings suggest that NOD2 blockade inhibits C5a generation by enhancing CD55 expression on neutrophils , depending on IL-1β and IL-10 production by neutrophils and resulting in increased survival rates .
Our experiments demonstrated that serum and peritoneal levels of C5a , but not C3a , were lower in Nod2−/− mice than WT mice during sepsis , while Nod2−/− mice showed higher survival rates than did WT mice , which was reversed by administration of rC5a . NOD2 agonists induce IL-1β production in mice by activating proIL-1β transcription and triggering the release of bioactive IL-1β [30] . Moreover , the 3020insC NOD2 mutant protein in patients with Crohn's disease actively inhibits IL-10 production by impairing hnRNP-A1 phosphorylation and hnRNP-A1 binding to the IL-10 locus [31] . Therefore , it is feasible that NOD2-mediated signals induce IL-1β and IL-10 production by immune cells during sepsis . Consistent with this suggestion , our experiments demonstrated that C5a generation was regulated via NOD2-mediated IL-1β and IL-10 production by peritoneal neutrophils rather than macrophages . This regulation pattern is intriguing in that prototypical pro- and anti-inflammatory cytokines exert a similar effect on C5a generation in vivo . Several lines of evidence in our experiments support this regulatory pattern . First , in vitro experiments revealed that WT and Nod2-deficient , but not Il-1r-deficient peritoneal cells produced IL-10 in an IL-1β dose-dependent manner . Second , Il-1r−/− mice exhibited lower serum and peritoneal IL-10 and C5a levels during sepsis , although IL-1β levels were similar in Il-1r−/− and WT mice . Third , rIL-1β did not enhance serum and peritoneal C5a levels in Il-10−/− mice , which showed minimal C5a levels compared to WT mice with sepsis , whereas rIL-1β or rIL-10 administration to Nod2−/− mice enhanced serum and peritoneal C5a . Fourth , rIL-1β administration to Nod2−/− mice increased IL-10 and C5a levels during sepsis . Fifth , neutrophil depletion in WT mice reduced the levels of IL-1β , IL-10 , and C5a , although these depletion effects were dependent on time points of sepsis . Therefore , the NOD2-mediated IL-1β-IL-10 regulatory loop in neutrophils helps enhancement of C5a generation , which may partially account for the different kinetics of pro- and anti-inflammatory cytokines in sepsis [1] , [3] . Furthermore , NOD2-mediated IL-1β and IL-10 production also suppressed LPS-mediated cytokine production by peritoneal immune cells during CLP-induced sepsis ( Fig . S7A ) . Consistent with our results , IL-1 receptor blockade attenuates CLP-induced sepsis [32] , [33] . In contrast , injection of human recombinant IL-1α protects against experimental sepsis in a time-dependent manner [34] . Consistent with this suggestion , time points of recombinant IL-1β injection was critical to exert harmful effects on sepsis in Nod2−/− mice ( data not shown ) . Thus , these findings suggest that IL-1 might play diverse functions in sepsis , depending on different time points . Meanwhile , several studies demonstrate that macrophages also produce IL-1β , TNF-α , and IL-6 during sepsis [35] , [36] . However , our experiments demonstrated that the levels of IL-6 and TNF-α in serum and peritoneum , and cytosolic IL-1β and IL-10 expression in macrophages were similar between WT and Nod2−/− mice during sepsis . Thus , it is less likely that macrophages might be a main subset to produce IL-1β and IL-10 for regulation of NOD2-mediated C5a generation , although macrophages play a critical role in the regulation of septic responses . Our experiments demonstrated that NOD2-mediated IL-10 production suppresses CD55 expression on neutrophils in an IL-1β-dependent manner . However , IL-1β regulates CD55 expression only minimally . Furthermore , considering the CD55 expression levels on neutrophils from WT , Il-1r−/− , Il-10−/− , and Nod2−/− mice with CLP , it is conceivable that NOD2-mediated IL-10 production suppresses CD55 expression on peritoneal neutrophils during sepsis in both IL-1β-dependent and IL-1β-independent manners . Soluble CD55 administration reduced C5a generation and increased the survival rates of WT and Nod2−/− mice injected with rIL-10 during sepsis . Moreover , soluble CD 55 administration also decreased serum and peritoneal C5a levels in Nod2−/− mice depleted neutrophils during CLP-induced sepsis . This suggests that the reduced CD55 expression by neutrophils caused by NOD2-mediated IL-10 production directly regulates C5a generation . This appears to be reasonable because neutrophils represent a major subset of cells in the peritoneum during sepsis . However , the altered CD55 expression on neutrophils affected C5a , but not C3a generation in the NOD2-mediated pathogenesis of sepsis , although CD55 inhibits both C3a and C5a convertase [37] . Thus , it is questionable how reduced CD55 expression on neutrophils inhibits the generation of C5a rather than C3a , in NOD2-mediated pathogenesis of sepsis . CD55 expression on APCs regulates local generation of C5a following cognate interactions between APCs and T cells [38] , suggesting that CD55 expressed on immune cells regulates generation of C5a , rather than C3a . Therefore , we postulated that CD55 expressed on neutrophils might inhibit C5a to a greater extent than C3a convertase in the septic microenvironment . Alternatively , compensatory generation of C3a in the complement cascade may account for the relative lack of a change in C3a level during sepsis , even though altered CD55 expression inhibits C3a and C5a convertase equally in vivo . Moreover , unknown mechanisms operating during sepsis might explain this unusual situation . To the best of our knowledge , this study provides the first demonstration that IL-1β-dependent and/or IL-1β-independent IL-10 production enhances C5a generation by suppressing CD55 expression on neutrophils during sepsis . IL-1β and IL-10 concentrations are significantly higher in patients with septic shock than in those with severe sepsis [39] . It generally accepted that IL-1β-mediated systemic inflammatory responses and cardiac dysfunction , and IL-10-mediated immune suppression account for the high mortality in sepsis [40]–[42] . In our experiments , NOD2-mediated IL-1β production exerted not only indirect modulation of C5a generation via the IL-1β-IL-10 loop , but also direct the regulation of septic response by decreasing immune cell phagocytosis and elevated culturable bacterial CFU levels in a C5a- and IL-10-independent manner ( Fig . S7B , C ) . Several studies have reported that IL-10 suppresses immune responses during sepsis by activation-induced apoptosis of T cells , reducing MHC class II expression on APCs , decreasing IFN-γ production , and deactivating monocytes [41]–[43] . However , no differences between WT and Nod2−/− mice were detected in terms of T cell apoptosis and IFN-γ production , whereas the expression levels of MHC class II , CD80 , and CD86 on APCs in Nod2−/− mice were lower than those in WT mice during sepsis ( Fig . 2A and Figs . S8 , S9 ) . Furthermore , administration of rIL-1β or rIL-10 to Nod2−/− mice did not increase CD4+ or CD8+ T cell apoptosis in the spleen and thymus ( Fig . S4 ) . These results suggest that NOD2-mediated IL-10 production minimally modulates T cell apoptosis , activation , and differentiation during sepsis . Therefore , the high mortality of patients during hypo-inflammatory phase of sepsis might be attributable to the effect of IL-10 on both NOD2-mediated C5a generation and immune suppression , which leads to primary and/or secondary hospital-acquired infection [44] . In contrast to protective role in single bacterial infections , NOD2-mediated signals aggravate polymicrobial sepsis . Considering that polymicrobial infection and the septic microenvironment appear to differ from those in monomicrobial infections , it is conceivable that NOD2 plays diverse roles in innate immune responses against bacteria , depending on the in vivo microenvironment . Thus , we propose that NOD2 has both protective and provocative functions in immunity to bacterial infection . With regard to NOD2-targeted therapeutics for sepsis associated with bacterium , inhibition of NOD2 signals might be useful . Consistent with this suggestion , our experiments demonstrate that blockade of NOD2 signals via RIP2 and P38 inhibition using SB203580 attenuated sepsis by reducing C5a generation , suggesting that inhibitors of RIP2 and/or its downstream molecules may be therapeutically useful for treatment of patients with sepsis . Moreover , a recent study demonstrated that EGFR tyrosine kinase inhibitors such as gefitinib and erlotinib , already used clinically as chemotherapy for non-small cell lung cancer , inhibited RIP2 tyrosine phosphorylation and MDP-induced cytokine release , but not in an EGFR-dependent manner [45] . Therefore , inhibitors of RIP2 phosphorylation may be effective therapeutic agents against sepsis and NOD2-related immune diseases . However , it has been reported that clinical trials targeting C5a to treat sepsis have failed [46] . Thus , it is need to be circumspect to develop therapeutic approach for sepsis based on NOD2-mediated C5a regulation pathway . In conclusion , NOD2-mediated signals increase C5a levels by suppressing CD55 expression on neutrophils via IL-1β-dependent or IL-1β-independent IL-10 production by neutrophils , thereby aggravating sepsis .
This study was performed in strict accordance with Korean law ( ANIMAL PROTECTION LAW ) . The experimental protocol was approved by the Institutional Animal Care and Use Committee of Biomedical Research Institute of Seoul National University Hospital ( SNUH-IACUC No . 12-0130 ) . Seven- to eight-week-old C57BL/6 mice were purchased from the Orient Company Ltd ( Seoul , Korea ) . Nod2−/− , Il-10−/− , and Il-1r−/− mice were purchased from the Jackson Laboratory ( Bar Harbor , ME , USA ) . The mice were bred and maintained under specific pathogen-free conditions at the Biomedical Research Institute Seoul National University Hospital . To perform CLP-induced sepsis , the mouse cecum was exposed through an 1 cm incision , and the cecum was ligated below the ileocecal valve using a 5-0 Ethilon suture ( Ethicon , Somerville , NJ , USA ) without causing bowel obstruction . Then the cecum was punctured with a 26-gauge needle at two different spots . In neutrophils depletion experiments using anti-Ly-6G antibody in vivo , the cecum was punctured at one spot . Mouse recombinant ( mr ) IL-1β 40 µg/mouse ) or IL-10 ( 30 µg/mouse ) ( ProSpec-Tany TechnoGene , Rehovot , Israel ) in PBS was i . p . injected into mice 4 or 12 h after CLP , respectively . mrC5a ( 5 µg/injection ) ( R&D Systems Inc . , Minneapolis , MN , USA ) was i . p . injected into mice 4 and 12 h after CLP . To block IL-10 signaling in vivo , WT mice were intravenously injected anti-IL-10R mAb ( 200 µg/mouse ) ( BD Bioscience , Sparks , MD ) , 1 day prior to CLP . mrCD55 20 µg/mouse ) ( R&D systems ) was i . p . injected into mice 12 h after CLP . To deplete neutrophils in vivo , WT and Nod2−/− mice were i . p . injected with anti-Ly-6G antibody ( 150 µg/mouse ) ( Bioledgend , San Diego , CA , USA ) 0 and 6 h after CLP . SB203508 ( 0 . 1 µmol/mouse ) ( Sigma Aldrich ) in 200 µl 0 . 5% PBST was i . p . injected into WT mice 1 , 5 , 16 h after CLP . Bacterial CFUs were counted by plating serial dilutions of blood and liver homogenates onto blood agar plates ( Hanil Komed , Seoul , Korea ) , which were incubated in 5% CO2 at 37°C overnight . The number of colonies was counted after incubation for 18 h . To obtain total peritoneal cells , we injected 4 ml of RPMI media containing 2 . 5% FBS into peritoneal cavity of mice and collected cells from peritoneal fluids . Peritoneal cells ( 5×105 ) from mice were cultured with MDP ( 20 µg/ml ) ( Sigma Aldrich ) in the presence or absence of SB203580 ( 100 nM ) ( Calbiochem , San Diego , CA . ) for 20–24 h . To estimate the responsiveness of immune cells to LPS , cells ( 5×105 ) were cultured with LPS ( 1 µg/ml ) for 7 h and various cytokine concentrations were measured . All cytokines and complements were measured using a BD Bioscience ELISA kit according to the manufacturer's instructions . ELISA assay for C5a is not detecting C5 . Cells were incubated with antibodies on ice for 30 min in 100 µl staining buffer ( 0 . 5% BSA ) . FITC- or PE Cy7-conjugated anti-Ly-6G , PE-conjugated anti-CD55 , anti-IL-10R , anti-IL-1R , anti-CD4 , anti-CD8 , anti-MHC class II mAb , FITC-conjugated anti-CD80 , anti-CD86 , anti-annexin V mAbs , and 7-amino-actinomycin D were purchased from BD Biosciences . A PE Cy5- or Alexa 647 ( eBioscience , San Diego , CA , USA ) -conjugated anti-F4/80 mAb was used . CD55 expression was estimated on gated F4/80+Ly-6G− and F4/80−Ly-6G+ peritoneal cells . To perform intracellular staining , peritoneal cells were obtained from mice 4 h and 12 h after CLP and incubated 4 h with 1 µl/ml GolgiStop ( BD Bioscience ) . Cells were surface stained with Cy7-conjugated anti-Ly-6G and Alexa 647-conjugated anti-F4/80 mAb , and incubated with fixation/permeabilization solution of Cytofix/Cytoperm Kit ( BD Biosciences ) for 20 min . After washing , the cells were stained with PE-conjugated anti-IL-10 or FITC-conjugated anti-IL-1βmAb ( BD Biosciences ) . Cells were run on an FACS LSR II or FACS caliber ( BD Bioscience ) , and analyzed using the Flowjo software ( Treestar , Ashland , OR , USA ) . Peritoneal fluid cells obtained from mice were stained with FITC-conjugated anti-Ly-6G mAb ( BD Biosciences ) and PE Cy5-conjugated anti-F4/80 mAb ( eBioscience ) . Then , stained cells were sorted on a BD FACSAria flow cytometer ( Franklin Lakes , NJ , USA ) . Sorted Ly-6G+F4/80− and Ly-6G−F480+ cells were isolated at 98% purity . An RNeasy Mini kit ( Qiagen , Courtaboeuf , France ) was used to isolate mRNA from sorted peritoneal Ly-6G+F4/80− and Ly-6G−F480+ cells . RNA ( 3 µg ) was reverse-transcribed into cDNA using M-MLV Reverse Transcriptase ( Promega , Madison , WI , USA ) . PCR was performed using cDNA as a template with primers and probes from Applied Biosystems ( Foster City , CA , USA ) and Biosource ( Camarillo , CA , USA ) for GAPDH , NOD2 , IL-1β , and IL-10 ( TaqMan pre-developed Assay Reagent ) . Gene expression levels were normalized to that of GAPDH . Mouse peritoneal cells were cultured with FITC-labeled E . coli ( Invitrogen , Carlsbad , CA , USA ) for 15 min . Attached cells were washed with warm PBS three times and then treated with 0 . 2% trypan blue for 1 min at room temperature . Then the cells were fixed with 4% formalin for 15 min and cultured with FITC-labeled E . coli to estimate nonspecific binding of E . coli to the cell surface . These cells , which were cultured with FITC-labeled E . coli , were run on the FACs caliber and analyzed with the Flowjo software . Western blotting was performed as described previously [47] . Antibodies against phosphor–p38 , p38 ( Cell Signaling Technology , MA , USA ) , phosphor-Rip2 ( Thermo scientific , Rockford , USA ) , Rip2 ( Santa Cruz Biotechnology , CA , USA ) , Bb ( Santa Cruz Biotechnology ) , NOD2 ( Santa Cruz Biotechnology ) , and a horseradish peroxidase-conjugated goat anti-rabbit IgG ( Thermo scientific ) were used . Survival data were plotted as Kaplan-Meier survival curves and analyzed using the log-rank test . Statistical significance was analyzed using Prism ver . 5 . 0 ( GraphPad Software Inc . , San Diego , CA , USA ) . One-way and two-way analyses of variance ( ANOVA ) and t-tests were performed , and a post hoc test was used if P<0 . 05 . Data are expressed as the mean ± standard error of the mean ( SEM ) .
|
Nucleotide-binding oligomerization domain ( NOD ) 2 is a cytosolic protein that senses peptidoglycans of bacteria and exerts a defensive effect on bacterial infection . Sepsis is a complex dysregulated inflammatory response in bacterial infection , causing multiple organ dysfunction , coagulopathy , and fatal outcome . C5a , which has harmful effects in sepsis , interacts with innate proteins . However , it remains unclear the mechanism by which NOD2 affects sepsis responses in vivo by regulating C5a generation . Here , we demonstrate that NOD2 enhances C5a generation by IL-10-mediated suppression of CD55 expression on neutrophils , thereby aggravating polymicrobial sepsis . These findings suggest a provocative role for NOD2 in sepsis , in contrast to its protective role during bacterial infection . Therefore , we propose that a novel NOD2-mediated complement cascade regulatory pathway in neutrophils may be a useful therapeutic target for sepsis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases",
"bacteremia",
"sepsis",
"infectious",
"disease",
"modeling",
"critical",
"care",
"and",
"emergency",
"medicine"
] |
2013
|
NOD2-mediated Suppression of CD55 on Neutrophils Enhances C5a Generation During Polymicrobial Sepsis
|
Infectious disease transmission is an inherently spatial process in which a host’s home location and their social mixing patterns are important , with the mixing of infectious individuals often different to that of susceptible individuals . Although incidence data for humans have traditionally been aggregated into low-resolution data sets , modern representative surveillance systems such as electronic hospital records generate high volume case data with precise home locations . Here , we use a gridded spatial transmission model of arbitrary resolution to investigate the theoretical relationship between population density , differential population movement and local variability in incidence . We show analytically that a uniform local attack rate is typically only possible for individual pixels in the grid if susceptible and infectious individuals move in the same way . Using a population in Guangdong , China , for which a robust quantitative description of movement is available ( a travel kernel ) , and a natural history consistent with pandemic influenza; we show that local cumulative incidence is positively correlated with population density when susceptible individuals are more connected in space than infectious individuals . Conversely , under the less intuitively likely scenario , when infectious individuals are more connected , local cumulative incidence is negatively correlated with population density . The strength and direction of correlation changes sign for other kernel parameter values . We show that simulation models in which it is assumed implicitly that only infectious individuals move are assuming a slightly unusual specific correlation between population density and attack rate . However , we also show that this potential structural bias can be corrected by using the appropriate non-isotropic kernel that maps infectious-only code onto the isotropic dual-mobility kernel . These results describe a precise relationship between the spatio-social mixing of infectious and susceptible individuals and local variability in attack rates . More generally , these results suggest a genuine risk that mechanistic models of high-resolution attack rate data may reach spurious conclusions if the precise implications of spatial force-of-infection assumptions are not first fully characterized , prior to models being fit to data .
The spatial heterogeneity of infectious disease incidence at large scales presents numerous intervention opportunities and challenges . Maps of malaria prevalence [1] have been used to target additional surveillance and to prioritize countries and geographical regions for additional intervention investment , resulting in substantial decreases in numbers of infections [2] . Over shorter timescales , spatial asynchrony in the northern hemisphere during the 2009 influenza pandemic likely led to variable effectiveness of vaccination when eventually deployed because of prior infections [3] . The epidemiological implications of substantial spatial heterogeneity in both incidence and transmission are topics of active research for most human pathogens [4] . These spatial heterogeneities must be influenced by two key human behaviours: where people choose to live and how they move . Because the home location of an individual is primarily used as the geographic location when cases are recorded , absolute spatial incidence is driven by population density: where more people live in a given unit area , there is greater potential for cases . Accurate high resolution estimates of population density [5 , 6] and travel [7] have helped refine global absolute estimates of disease incidence and prevalence [8–11] . In order for a directly transmitted human pathogen to move through space , at least one person must travel away from home and meet another person . Even for vector borne pathogens such as malaria and Zika virus , typical distances traveled by the vector are much shorter than those traveled by human hosts . Human movement is captured by survey data on journeys to work [12] , questionnaire-based surveys [13] and location logging of mobile devices [14–16] . Although spatial heterogeneity has been measured at larger scales ( e . g . serological attack rates for influenza [17] ) , modern pathogen surveillance enables more finely resolved incidence data sets , with details such as precise geographical location captured with increasing frequency by modern digital and biological technology . For example , the full genome of a pathogen can be made available in almost real time directly from clinical samples taken in the community [18] , and the home location of everyone attending a health care facility can be extracted from clinical episode data [19] . Because this level of geographical precision for high quality incidence data has not previously been available , both epidemiological and disease-dynamic studies of infectious disease have focused on predicting and explaining incidence patterns measured at larger spatial scales , often with all cases within an administrative unit reported together . Additional insights are likely being lost during this aggregation process . Available evidence and intuition suggests that infectious and non-infectious individuals have different social interactions during an outbreak [20] , with plausible scenarios in which either one or the other may be more connected in space . For example , susceptible individuals are more likely to travel more than are infectious individuals with mild symptoms [21] . However , family members and friends providing care for infectious individuals may often not behave in the same way as an average susceptible individual . Also , infectious individuals themselves may travel long distances away from transmission hotspots to seek medical care during outbreaks of highly pathogenic infections [22] . Disease dynamic models are often used to study infection incidence and are defined primarily by their force-of-infection ( FOI ) term: a precise mathematical specification of how the risk of infection experienced by a susceptible individual is driven by the number of currently infectious individuals and by their characteristics . For example , the ages of infectious and susceptible individuals must sometimes affect the risk of infection , as must the distance between their home addresses . Disease dynamic models that represent space [23] are now used routinely to understand large-scale spatial heterogeneity in incidence: to estimate the relative effectiveness of spatially heterogeneous interventions ( given the observed incidence ) ; to reveal underlying social mechanisms of transmissions; and , with increasing frequency , to forecast future spatial incidence patterns [24] . All transmission models that represent space include some kind of spatial kernel—a formal definition of the way in which individuals from different locations distribute their influence over the whole of geographical space . However , there is substantial variability in the underlying FOI assumptions made in these models , which are often not discussed explicitly and have likely only rarely made material differences to model-based results aggregated at larger spatial scales . Nonetheless , we hypothesise that these different FOI assumptions represent important alternate hypotheses for the mechanisms of transmission and may lead to substantial structural biases in the predictions of attack rates at smaller spatial scales . Here , we propose a general theoretical framework for the study of infectious disease incidence at arbitrarily small spatial scales and , in particular , we look at the relative mobility of infectious individuals relative to susceptible individuals as a potential driver of heterogeneity in incidence .
Algebraic analyses show that differential spatial connectivity of susceptible and infectious individuals can lead to variability in local attack rates ( S1 Protocol ) . Firstly , we showed that if susceptible and infectious individuals are assumed to be connected in the same way across all points in space , then local attack rates are uniform for any population density distribution or grid resolution . For lower resolution grids with large individual spatial elements , where the amplitude of connectivity of individuals outside their home pixel is small , the impact of differential connectivity between susceptible and infectious individuals is still negligible , even to the point that it is reasonable to assume that infectious individuals have no connectivity at all outside their home location . However , as the resolution of the grid increases and pixels become smaller , individuals have a substantial number of connections outside their home pixel . Under this scenario , it was no longer possible to prove analytically that differences in the connectedness of susceptible and infectious individuals would not lead to local variation in attack rates . These analytical results were not affected by the presence of age stratification in the transmission process , so long as the behavior and distribution of age groups was assumed to be uniform across space . We established a baseline numerical scenario consistent with a 1918-like influenza pandemic by implementing the underlying transmission model ( see Methods ) as ordinary differential equations ( ODEs ) . Using: a 1km by 1km gridded population density ( 55km by 33km to the east and north of Guangzhou , China ) ; a spatial contact kernel estimated in the same population [25]; a basic reproductive number R0 = 1 . 8 [26] and recovery rate 1/2 . 6 days−1 [27]; we recovered a global uniform attack rate of z = 0 . 73 , consistent with the homogeneous mixing model SIR model [28] . We also introduced age-stratified populations and transmission using parameters estimated in this population [13] . For this population , accurate high-resolution data on local age distributions were not available , therefore , we assumed that all pixels had populations with the same age distribution , even though the total number of individuals in a single pixel varied substantially . This addition of age effects in the transmission process did not introduce spatial variation but did reduced the uniform global attack rate to z = 0 . 43 , consistent with analysis of the 2009 influenza pandemic [29] . We validated the precision of attack rates obtained from the ODEs using age- and space-stratified refinements [23] of the standard implicit equation relating attack rate ( final size ) z to R0: z = 1 − e−R0z [28] . We hypothesized that both population density and the gradient of population density may influence small-scale attack rates in these models . Fig 1A and 1B show the uniform attack rate when mobility is independent of infection status ( henceforth referred to as “dual mobility” ) with four age classes , plotted against log of population density and gradient of log population density respectively ( with log gradient defined as the average difference between the log of a location’s resident population and that of its 8 immediate neighbors ) . When only non-infectious individuals were assumed to be mobile ( S-mobility ) , location-specific attack rates were positively correlated with log population density , correlation coefficient c = 0 . 75 ( Fig 1C ) . Attack rates varied between a minimum of 33 . 72% to a maximum of 45 . 76% , an absolute range of 12 . 04% . Location-specific attack rates were slightly less correlated with the log gradient of population density ( correlation coefficient c = 0 . 73 , Fig 1D ) . Locations with higher attack rates tended to be densely populated relative to neighboring locations ( Fig 2A and 2B ) . Note that the term “S-mobility” includes mobility in the recovered population . Conversely , when only infectious individuals were assumed to be mobile ( I-mobility ) , pixel attack rates were negatively correlated with log population density ( c = -0 . 7707 , Fig 1E ) and even more strongly negatively correlated with log density gradient ( c = -0 . 8816 , Fig 1F ) . Attack rates varied over a greater range than for susceptible-only mobility: from a minimum of 32 . 61% to a maximum of 90 . 73% , with an absolute range of 58 . 12% . High attack rate pixels tended to be sparsely populated relative to neighboring locations ( Fig 2A and 2C ) . The reader is referred to the discussion for an evaluation of the applicability of this assumption to epidemic models . Measures of spatial variation are inherently dependent on the resolution of the model grid and even the strong variability outlined above would be missed by most surveillance systems . The absolute range of attack rates for the susceptible-only movement was reduced to 1 . 67% when aggregated to 8km by 8km pixels . Even though the effect of infectious-only movement was stronger than for susceptible-only mobility , it was rapidly hidden by the aggregation of pixels , with the absolute range dropping to 3 . 78% when aggregated to 8km by 8km pixels . Results of aggregation using S-mobility is shown in Fig 3 , and the corresponding result using I-mobility is shown in S1 Fig . The direction of association between FOI assumptions and local attack rate was preserved and the amplitude remained substantial for intermediate scenarios in which both susceptible and infectious individuals were mobile but to differing degrees . If infectious individuals had any more contacts than susceptible individuals then attack rates were negatively correlated with population density , and vice versa ( Fig 4 ) . When infectious individuals reduced their travel by a factor of 0 . 5 , the absolute range of attack rates was 5 . 38% and when susceptible individuals reduced their mixing by the same degree ( with infectious agents fully mobile ) , the absolute range was 12 . 89% . The underlying mobility choice kernel K was defined by the relative probability of making a contact in a population at a distance r and of population size N . It was parameterized by an offset distance a , a distance power p and destination population power α; K = Nα ( 1 + r/a ) −p , with values obtained by fitting to data from this population [25] . Qualitatively , our conclusions about the impact of differential contact rates by susceptible individuals were not sensitive to values for the offset distance a nor the distance power p ( Fig 5A–5D ) . However , they were sensitive to values of the destination power α for which we have used the best fit value of 0 . 53 ( for results up to this point ) ( Fig 5E&5F ) . Intriguingly , with the often-assumed default value α = 1 , the correlation between local attack rates and population density or gradient have the opposite sign ( S2 and S3 Figs ) . Moreover , α = 1 induces weaker correlations with local population gradient . It is therefore essential to provide an accurate estimate for α , which does not require infection-related data , before attempting to infer infection-dependent mobility . Stochastic solutions to the meta-population models suggest that attack rate variation driven by asymmetric mobility would not be dominated by demographic stochasticity ( Fig 6 ) . Variation in attack rate for the extreme cases of S- and I-mobility was dominated by stochastic effects only in sparsely populated areas . For pixels with the smallest population , the amplitude of variation expected to arise from asymmetric mobility is similar to that which may arise by chance due to stochastic effects . However , the expected amplitude of stochastic variation diminishes as population density increases , and variation in attack rate due to mobility assumption becomes apparent ( S4 Fig ) . For example , using susceptible-only mobility for 1km by 1km pixels with populations between 1 and 85 , 163 , the standard deviation in attack rate due to stochasticity is 9 . 45% while the standard deviation of expected attack rates due to asymmetric mobility is 2 . 61% . These results are robust to our choice of illustrative population density and to alternate natural history parameters . The same effects are observed when using population density of Puerto Rico with influenza natural history parameters ( S5 Fig ) and with parameters that approximate vector-borne transmission , such as those of Zika or Chikungunya ( S6 Fig ) . Summary statistics for these and all other deterministic model variants we have presented in this study are shown in S1 Table .
We have shown that , under the assumption that an individual’s total contact is independent of home location and where they travel , substantial heterogeneity in local attack rates could arise if mobility is dependent on infection status . Moreover , the direction of the relationship between attack rate and population density is dependent on the contribution of population density to the relative attractiveness of a location . For the estimate of that scaling for our sample population ( α = 0 . 52 ) , and when susceptible individuals are more mobile than infectious individuals , attack rates are positively correlated with population density . Conversely , when using the often implicit assumption that the kernel is directly proportional to population density ( α = 1 ) , this correlation is negative . Though increased mobility in infectious agents may seem less likely than reduced mobility , there do exist potential scenarios where this may be the case in both human and animal systems . For example , humans may travel to access health care in the case of severe symptom onset as has been the case anecdotally during the 2003/4 SARS outbreak and the 2013/14 Ebola outbreak . Also infectious opiate users in the USA may be more mobile than less infectious opiate users [30] . I-mobility may in fact be more relevant in the epidemiology of non-human infections , for example increased mobility in rabid dogs [31] and Gypsy moth caterpillars infected with baculovirus forfeit [32] . Our study has a number of limitations . We have not considered spatial variation in the age distribution of people , because these data were not available for our study population . Variability in local attack rates will very likely also be driven in non-trivial ways by spatial correlation in the proportion of the population in different age classes . This may be of particular significance in larger Chinese cities such as Guangzhou , in which urban areas are home to relatively few children and many rural locations have few working-age adults . There is also scope for the inclusion of an urban/rural distinction in the parametrization of the travel kernel [25] , and the simulation of multiple years of transmission , which would extend the applicability of our results beyond pandemic scenarios for influenza and other emergent pathogens . The refinement of this framework to include the above phenomena is a priority for future work and we would expect differential movement patterns with age and population to impact our findings . Though this study was limited to a standard SIR model , we would not expect the inclusion of a latent period , waning , or natural births and deaths to show make substantial differences to these findings . The primary results can be obtained using renewal equations which are only dependent on the probability of one individual escaping infection . Our sensitivity analysis with respect to kernel population power α provides some insight into the underlying mechanisms that give rise to the observed correlations between attack rate and population density under different mobility assumptions . For example , consider the special case where only infectious people are mobile and α tends to large values , making mobility dependent only on population density of location , and not on geographical distance . Under this scenario , high density pixels will draw in more and more infectious people and therefore generate higher attack rates . Conversely , if α = 0 , then mobility is dependent only on distance . Under this scenario , we can think of the infectious populations spilling out of their home locations into neighboring ones . Thus , any sparsely populated location that is adjacent to a densely populated location will see an influx of infectious individuals resulting in a greater proportion infectious in that location , and therefore a stronger FOI and subsequent attack rate . A schematic for the latter case is given in S7 Fig . These results illustrate the potential knock-on effects of little or no dependence between transmissibility and population density: that infectious people from more densely populated areas go to nearby sparsely populated areas and in some sense “seek out” people in those areas to infect so they can reach their quota ( I-mobility ) . Within the realm of parameters that are supported by studies of human movement and infectious processes , the behaviors implied by the models we presented here seem valid . Individual-based models have a number of advantages over other approaches . They can be coded in a generic way and adapted rapidly to different pathogen systems and specific scientific or policy questions . Even though they are often more substantially computationally burdensome than comparable meta-population approaches , they will likely be used with increasing frequency to address questions related to local attack rates . We have shown that mobility assumptions have implications for the interpretation of attack-rates derived from individual-based models , some of which assume implicitly that the spread of infection is driven by the movement of individuals . We have shown that , whichever mobility assumption is made in a given model , it is possible to modify this assumption by replacing isotropic K by a convoluted kernel L that accounts for the change in mobility assumption ( and so L may not be a stochastic matrix and hence functions as a non-isotropic kernel ) . In particular , the low-prevalence assumption makes this transformation achievable with minimal modification to existing computer programs . Therefore , developers of individual-based models may wish to consider alternate connectivity matrices for their simulations so as to explicitly reflect different spatial assumptions about the force of infection . We have also shown that the implications of typical assumptions that are made in spatially explicit FOI terms , including approximations to this crucial normalization , are non-trivial at small spatial scales . Such assumptions are , however , often not addressed explicitly and so may contribute unknowingly to results . We hope to offer clarity in the interpretation of FOI in spatial models , and to have provided a comprehensive framework from which we can gain a deeper understanding of the role of spatial mobility in disease transmission dynamics as infectious disease incidence data become available at higher and higher spatial resolution .
Data taken from populations we study here show that total contacts made per day , and contact durations , do correlate with population density ( p < 0 . 001 , [13] ) , but that the strength of the relationship is weak . This is in part due to working-age adults dominating the population of urban areas , but also to the phenomenon of urban isolation [33] . When investigating only the effect of mobility assumption in force of infection , our main results made the baseline assumption that total contact and duration of contact is independent of home location . The way in which these contacts are distributed in space does , however , depend on distance and population density , and is described via a spatial kernel K . In matrix notation , Kij is defined as the proportion of time spent by an agent from location i in location j . The assumption of uniformity of total contact therefore means that the rows of K sum to unity . Our model employs the offset gravity kernel , defined as follows: K i j ∝ N i N j α 1 + ( r i j / a ) p ( 1 ) with baseline parameters of a = 0 . 58 , p = 2 . 72 , α = 0 . 52 , where rij denotes the geodesic distance between the center-points of pixels i and j . Of the kernel structures studied in [25] , offset gravity is shown to best represent contact data . Imposing the constraint that K is stochastic renders redundant the factor Ni in the numerator ( owing to row-normalization ) . We used rectangular excerpts from the Landscan dataset [34] with the lower left corner of the rectangle located on the center of the city of Guangzhou , China . The rectangle is 55km from east to west and 33km from north to south , and a 4km boundary area was excluded after simulation . The boundary area was chosen according to the following rationale: when population density data for large suburban areas is truncated for the purpose of simulation , it is equivalent to imposing empty space outside of the boundary , and this modification may effect the attack rates calculated in pixels close to that boundary . We ran simulations on a large area of 1km by 1km pixels , and on smaller areas contained within this larger area . We found that attack rates agree on all pixels on the interior of the smaller area once a 4km perimeter is removed . Let A denote the S-mobility kernel and B the I-mobility kernel . Then the age-independent generalized FOI equation is given by: λ i = β ∑ j A i j ∑ k B j k T I k ∑ l [ A j l T ( N l - I l ) + B j l T I l ] . ( 2 ) For reduced mobility , movement of the non-infectious population is governed by a parameter δ such that A = ( 1 − δ ) E + δK , where E is an identity matrix representing absence of spatial mobility . Similarly , we describe mobility of infective individuals by ϵ such that B = ( 1 − ϵ ) E + ϵK . S-mobility thus corresponds to δ = 1 , ϵ = 0 and I-mobility to δ = 0 , ϵ = 1 . If K is the n × n spatial kernel , indexed by i , j , k , l , and C the 4 × 4 age-mixing matrix , indexed by a , b , c , d , then the age-explicit dual-mobility equation is given by: λ ( a , i ) D = β ∑ b , j K i j δ a b ∑ c , k K j k T C b c I ( c , k ) ∑ d , l K j l T N ( d , l ) ( 3 ) This can be combined with Eq ( 2 ) to give the age-dependent system with reduced mobility . In all simulations presented in this study , we use the pointwise product of the matrices defining number of contacts and duration of contact between age groups 0–4 , 5–19 , 20–64 and 65+ derived in [13] . These age-mixing matrices were constructed from contact surveys conducted in the region of Guangzhou used in our results . We define the gridded transmission model as ordinary differential equations . However , we also implement a stochastic compartmental version of the model and we calculate attack rates using recursive equations . We used a standard SIR model with S ˙ i = - S i λ i , I ˙ i = S i λ i - γ I i , R ˙ i = γ I i . ODE models were seeded proportional to population density ( σ = 10−8 × N/∑iNi ) , and agreed with final size calculations ( which assume infinitesimal seeding ) . Integration of ODEs with full FOI in the S- and I-mobility case , i . e . with Il ( t ) in denominators , showed low-prevalence approximations to be good . For example , in the main S-mobility result , the mean difference in pixel attack rates between the full FOI and low prevalence approximation was 6 . 22 × 10−4 with maximum difference 3 . 3 × 10−3 occurring in a pixel with population 726 . Therefore , numerical solutions for all figures were obtained using the low prevalence approximation ( c . f . S1 Protocol ) . A selection of smaller examples agreed when checked using the full FOI . The stochastic compartmental variant of our model selected the number of agents to infect from binomial distribution with parameters S ( a , i ) and 1 − exp ( −λ ( a , i ) ) . This method requires specification of a time-step , and we found Δt = 1/6 days to be sufficiently small ( results did not change when Δt was doubled , and results were consistent with the corresponding deterministic model ) .
|
We know that some places have higher rates of infectious disease than others . At the moment , we usually only measure these differences for large towns and cities , though modern data allows us to track movement at much higher resolution . In this paper , we used a computer simulation of an epidemic to propose ways that rates of incidence in small local areas might be related to population density . We found that if infectious people are better connected than non-infectious people , perhaps because they receive visitors , then , on average , higher density areas would have lower rates of infection . If infectious people were less connected than non-infectious people then higher density areas would have higher rates of infection . As data get more accurate , this type of analysis will allow us to propose and test ways to optimize interventions such as the delivery of vaccines and antivirals during a pandemic .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"infectious",
"disease",
"epidemiology",
"influenza",
"pathogens",
"social",
"sciences",
"human",
"mobility",
"population",
"biology",
"human",
"geography",
"infectious",
"diseases",
"geography",
"epidemiology",
"disease",
"vectors",
"population",
"metrics",
"earth",
"sciences",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"population",
"density",
"species",
"interactions"
] |
2019
|
Differential mobility and local variation in infection attack rate
|
Visceral leishmaniasis ( VL ) caused by Leishmania donovani remains of public health concern in rural India . Those at risk of VL are also at risk of other neglected tropical diseases ( NTDs ) including soil transmitted helminths . Intestinal helminths are potent regulators of host immune responses sometimes mediated through cross-talk with gut microbiota . We evaluate a meta-taxonomic approach to determine the composition of prokaryotic and eukaryotic gut microflora using amplicon-based sequencing of 16S ribosomal RNA ( 16S rRNA ) and 18S rRNA gene regions . The most abundant bacterial taxa identified in faecal samples from Bihar State India were Prevotella ( 37 . 1% ) , Faecalibacterium ( 11 . 3% ) , Escherichia-Shigella ( 9 . 1% ) , Alloprevotella ( 4 . 5% ) , Bacteroides ( 4 . 1% ) , Ruminococcaceae UCG-002 ( 1 . 6% ) , and Bifidobacterium ( 1 . 5% ) . Eukaryotic taxa identified ( excluding plant genera ) included Blastocystis ( 57 . 9%; Order: Stramenopiles ) , Dientamoeba ( 12 . 1%; Family: Tritrichomonadea ) , Pentatrichomonas ( 10 . 1%; Family: Trichomonodea ) , Entamoeba ( 3 . 5%; Family: Entamoebida ) , Ascaridida ( 0 . 8%; Family: Chromodorea; concordant with Ascaris by microscopy ) , Rhabditida ( 0 . 8%; Family: Chromodorea; concordant with Strongyloides ) , and Cyclophyllidea ( 0 . 2%; Order: Eucestoda; concordant with Hymenolepis ) . Overall alpha ( Shannon’s , Faith’s and Pielou’s indices ) and beta ( Bray-Curtis dissimilarity statistic; weighted UniFrac distances ) diversity of taxa did not differ significantly by age , sex , geographic subdistrict , or VL case ( N = 23 ) versus endemic control ( EC; N = 23 ) status . However , taxon-specific associations occurred: ( i ) Ruminococcaceae UCG- 014 and Gastranaerophilales_uncultured bacterium were enriched in EC compared to VL cases; ( ii ) Pentatrichomonas was more abundant in VL cases than in EC , whereas the reverse occurred for Entamoeba . Across the cohort , high Escherichia-Shigella was associated with reduced bacterial diversity , while high Blastocystis was associated with high bacterial diversity and low Escherichia-Shigella . Individuals with high Blastocystis had low Bacteroidaceae and high Clostridiales vadin BB60 whereas the reverse held true for low Blastocystis . This scoping study provides useful baseline data upon which to develop a broader analysis of pathogenic enteric microflora and their influence on gut microbial health and NTDs generally .
Visceral leishmaniasis ( VL ) , also known as kala-azar , is a potentially fatal disease caused by obligate intracellular parasites of the Leishmania donovani species complex . VL is a serious public health problem in indigenous and rural populations in India , causing high morbidity and mortality , as well as major costs to both local and national health budgets . The estimated annual global incidence of VL is 200 , 000 to 400 , 000 , with up to 50 , 000 deaths annually occurring principally in India , Bangladesh , Sudan , South Sudan , Ethiopia and Brazil [1] . The outcome of infection with L . donovani depends on the host immune response [2] , that in turn could be influenced by other factors , one of which could be co-infection with other pathogens . Recent interest has focused in particular on the influence that gut microbial composition has on health and disease , in particular bacteria such as Enterobacteriaceae which have been associated with systemic inflammation ( reviewed [3] ) . In the case of VL , in particular , it has been proposed that systemic inflammation due to microbial translocation from the gut is part of disease pathogenesis [4 , 5] . Often those at risk of VL are also at risk for other neglected tropical diseases ( NTDs ) such as infestation with soil transmitted helminths or infection with lymphatic filariasis or Schistosoma mansoni . Intestinal helminths are particularly potent regulators of their host’s immune response and can ameliorate inflammatory diseases [6] . Few published studies are available on interaction between L . donovani infection and other NTDs . Hassan and coworkers [7] showed that , despite the development of a functional anti-L . donovani T helper 1 response , mice with established S . mansoni infections fail to control L . donovani growth in the liver and spleen . Similarly , we recently demonstrated [8] that chronic infection with intestinal nematode Heligmosomoides polygyrus exacerbated secondary L . donovani infection in mice , resulting in higher parasite burdens in liver and spleen compared to worm free mice . This increased parasite load was accompanied by increased interleukin-4 and interleukin-10 transcription in spleens . In contrast , the presence of the filarial parasite Brugia malayi L3/adult worms inhibited progression of L . donovani infection associated with a T helper 1 response in a hamster model [9] . However , much may depend on the timing of infection with the other NTD and with L . donovani . A chronic helminth infection such as may be expected in areas endemic for VL may have a different effect compared to a newly established primary infection . In humans , the only co-infection studies reported to date are in relation to American tegumentary leishmaniasis [10–12] . In an initial study of 120 patients with cutaneous leishmaniasis caused by L . braziliensis [11] patients co-infected with helminths took longer to heal than helminth-free patients . However , in a follow up study of 90 patients there was no statistically significant benefit to treating patients with anti-helminthics [12] . In a more recent retrospective cohort of 109 L . braziliensis patients from Brazil [10] it was shown that patients with positive parasitological stool examinations had more mucosal lesions and more often had a poor response to therapy . Interestingly , these effects were observed for total intestinal helminths ( based cumulatively on Ancylostomidae , Ascaris lumbricoides , Strongyloides stercoralis , Trichuris trichuria , Enterobius vermicularis , S . mansoni ) , and for A . lumbricoides alone or nematodes cumulatively ( A . lumbricoides , Str . stercoralis , T . trichuria , E . vermicularis ) , but not for protozoan ( based cumulatively on Blastocystis hominis , Entamoeba coli , E . histolytica , Giardia lamblia , Endolimas nana ) infections . One limiting factor for wide-ranging analysis of gut pathogen load has been the specialized and time-consuming processes involved in microscopic identification of parasites . While multiplex PCR for fecal DNA samples is facilitating larger scale studies for specific helminth and protozoan gut pathogens [13] this does not provide data on the potential influence of pathogenic species on gut microbes more broadly . For example , it was recently reported that the gut microbiome contributes to impaired immunity in pulmonary tuberculosis patients in India [14] . The presence of gut helminths is known to influence the gut microbiome [15 , 16] , and there is evidence [6] of important cross-talk between the gut microbiome and helminths in mediating host immunological effects . A meta-taxonomic approach to look at both commensal and pathogenic species in the gut would therefore be of great value . Numerous studies have used high-throughput , massively parallel amplicon-based sequencing of 16S ribosomal RNA ( 16S rRNA ) gene regions to study the prokaryotic composition of the gut microflora [17–19] , including studies in India [14 , 20–24] . Recent studies in mammalian hosts [25] have also assessed parasite diversity using amplicon-based high-throughput sequencing of 18S ribosomal RNA ( 18S rRNA ) gene regions and classified sequence reads into multiple parasite groups . Comparison of results with standard methods including microscopic observation of helminths in the intestines , and PCR amplification/ sequencing of 18S rDNA from isolated single worms , suggests that this new technique is reliable and useful to investigate eukaryotic parasite diversity in fecal samples . However , this approach has not to our knowledge been used in humans . Here we present a scoping study that uses 16S rRNA and 18S rRNA meta-taxonomic analysis of fecal samples with the primary aims of determining the composition of gut prokaryotic and eukaryotic microflora in an area of India endemic for VL and whether this differs between VL cases and non-VL endemic controls . We compare microscopic determination of helminth infections with 18S rRNA data , and report on secondary aims of determining the potential influence of prokaryotic or eukaryotic taxa known to contain pathogenic species on gut microbial diversity per se .
The enrolment of human subjects complies with the principles laid down in the Helsinki declaration . Institutional ethical approval ( reference numbers: Dean/2017/EC . 148 ) was obtained from the ethical review board of Banaras Hindu University ( BHU ) , Varanasi , India . Informed written consent was obtained from each participant at the time of enrolment , or from their legal guardian if they were under 18 years old . Only patients who had not previously received treatment and who agreed to participate in the study were enrolled . All clinical treatment and follow-up records were maintained using standardised case report forms on an electronic server . Samples were collected between May and June 2017 at the Kala-azar Medical Research Center ( KAMRC ) , Muzaffarpur , Bihar , India . Active VL cases ( N = 23 ) were diagnosed by experienced clinicians based on clinical signs , including fever ( >2 weeks ) , splenomegaly , positive serology for recombinant antigen ( r ) -K39 and/or by microscopic demonstration of Leishmania amastigotes in splenic aspirate smears . Fecal samples ( cf . below ) were collected prior to treatment . Endemic controls ( EC; N = 23 ) that did not have VL were attendants accompanying VL patients at KAMRC ( N = 16 ) or from field locations nearby ( N = 7 ) . Basic demographic details ( age and sex ) on participants are provided in Table 1 . Metadata by individual are provided in S1 Table . Stool collection containers and toilet accessories ( to avoid urine contamination; OM-AC1; DNA Genotek ) were distributed to the study participants and the containers collected immediately after defecation , either that evening or when participants defecated the following morning and stored at 4°C until processing . Most subjects defecated in the morning and the containers were then returned and processed within 3–4 hours of storage at 4°C . In rare instances defecation may have occurred at time when the nursing staff was not immediately available . Importantly , there was no difference in the collection and processing of samples between VL subjects and controls . The sample was divided into two parts: 200 μg to OMNIgene GUT tubes ( OM-200; DNA Genotek ) , and approximately 1 gram was concentrated by the formalin-ethyl acetate sedimentation method and used for diagnostic microscopy . For microscopy , formalin-ethyl-acetate sedimentation was used to concentrate parasites following standard methodology [26] . Briefly , faeces were homogenized in 10 ml of 10% formalin containing 0 . 1% Triton-X-100 and kept at room temperature for at least 30 minutes for fixation . Samples were then filtered through two layers of gauge mesh , centrifuged 1500 g for 3 minutes , and the supernatant discarded . Seven ml of 10% formalin was added followed by 3 ml ethyl acetate , and samples shaken for 30 seconds and centrifuged 1500 g for 3 minutes . Supernatants were discarded and saline added to the sediment , which was then analysed under a light microscope at 10x magnification for the presence of helminth eggs . A portion ( 200μg ) of each stool sample ( as above ) was transferred to OMNIgene GUT tubes ( DNA Genotek ) following the OMNIgene GUT protocol for transferring specimens from the collection tubes to this kit ( protocol PD-PR-00434 ) . The specimen was homogenised and stabilised within these tubes prior to DNA extraction . Genomic DNA was extracted with the QIAamp PowerFecal DNA Kit ( QIAGEN ) following the manufacturer’s protocol with the exception that proteinase K was added after the bead beating step and the sample incubated at 65°C for 30 minutes . This was to ensure removal of histones from eukaryotic DNA which could inhibit downstream amplification . The remainder of the protocol was carried out as per manufacturer’s instructions . Purified genomic DNA was eluted in 100μl of PCR-grade water . Genomic DNA samples were shipped to the Telethon Kids Institute in Perth , Australia , at ambient temperature for further analysis and stored at -20°C upon arrival . As a positive control for the 16S rRNA and 18S rRNA sequencing we used the ZymoBIOMICS Microbial Community DNA Standard ( Integrated Sciences , Chatwood , NSW , Australia ) which comprised DNA for 8 bacterial ( Listeria monocytogenes—12% , Pseudomonas aeruginosa—12% , Bacillus subtilis—12% , Escherichia coli—12% , Salmonella enterica—12% , Lactobacillus fermentum—12% , Enterococcus faecalis—12% , Staphylococcus aureus—12% ) and 2 eukaryotic ( Saccharomyces cerevisiae—2% , and Cryptococcus neoformans—2% ) species . To ensure that amplification of 16S rRNA and 18S rRNA amplicons worked well in the presence of human DNA , we included replicates of the positive control with human DNA spiked in at 0% , 0 . 1% , 0 . 5% , 1% , and 10% of the total 20 ng reaction , as indicated . For the 18S rRNA positive control we also spiked in DNA from two protozoan parasites Leishmania major ( 1 . 2 ng/20 ng reaction ) and Toxoplasma gondii ( 1 . 2 ng/20 ng reaction ) to some mock controls as indicated . Negative controls included 4 unused OMNIgene GUT tubes processed for DNA in the Indian laboratory using the QIAamp PowerFecal DNA Kit ( QIAGEN ) kits as for preparation of experimental sample DNAs . These were included in the preparation of amplicons for sequencing . Genomic DNA was quantified with the Qubit 2 . 0 dsDNA HS assay kit ( Invitrogen ) and diluted to 10 ng/μl for amplicon PCR . Aliquots for sequencing of 16S rRNA gene were sent to the Australian Genome Research Facility ( AGRF ) for amplification , Nextera XT dual-indexing , pooling and sequencing . The V3-V4 region of the gene was amplified with primers 341F ( 5’CCTAYGGGRBGCASCAG ) and 806R ( 5’GGACTACNNGGGTATCTAAT ) [27] . Aliquots for sequencing of the 18S rRNA gene were first amplified with the V9 region primers 1391f ( 5'TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGGTACACACCGCCCGTC ) and EukBr ( 5’GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGTGATCCTTCTGCAGGTTCACCTAC ) from the Earth Microbiome Project ( EMP ) standard protocol ( http://www . earthmicrobiome . org/protocols-and-standards/18s/ ) , with Illumina Nextera adapters attached ( adapters in bold ) . The mammalian blocking primer ( 5’GCCCGTCGCTACTACCGATTGG/ideoxyI//ideoxyI//ideoxyI//ideoxyI//ideoxyI/TTAGTGAGGCCCT/3SpC3/ ) was also included to reduce amplification of human 18S rRNA . Samples were amplified in duplicate with the PCR mixture and conditions specified in the EMP protocol , and duplicates were pooled after amplification . These amplicons were then sent to AGRF for purification , dual-indexing , pooling and sequencing . Demultiplexed sequence data was received from AGRF in FASTQ format and imported into the QIIME2 v2018 . 11 [28] version of the QIIME software [29] for analysis . Full documentation of all analysis code in this project can be found at https://rachaellappan . github . io/VL-QIIME2-analysis/ . In brief , for both the 16S rRNA and 18S rRNA data , amplicon primers were first removed with the cutadapt plugin [30] . Reads were denoised , filtered and trimmed ( where the median quality score fell below 30 ) with the DADA2 plugin [31] . Amplicon sequence variants ( ASVs ) were classified within QIIME2 using the SILVA v132 database [32] , with a classifier trained on the amplified region [33] . These ASVs equate to classifying operational taxonomic units ( OTUs ) based on 100% sequence identity . This differs from the traditional method of clustering sequences and calling OTUs based ( usually ) on 97% sequence identity [34] . Note that we use the terminology of ASVs throughout . The majority taxonomy with 7 levels was used for both the 16S rRNA and 18S rRNA datasets , i . e . this taxonomy is taken to species level , but since species-level identification was not complete we elected to use genus-level classifications for most analyses . De novo phylogenetic trees used in downstream measures of diversity that included phylogenetic distances ( cf . below ) were created using log10 read counts and the phylogeny align-to-tree-mafft-fastree ( MAFFT multiple sequence alignment program ) [35 , 36] plugin in QIIME2 . Stacked bar plots of percent relative abundance of different taxa across samples were generated in QIIME2 . Aggregate percent abundance of taxa at the population level was determined by summing read counts per taxon across all study participants and calculating these as a proportion of the total read count . Prior to calculating downstream measures of diversity 18S rRNA data was also filtered to remove a small number of mammalian 18S rRNA reads , and also to remove bacterial 16S rRNA reads due to non-specificity of primers for eukaryotic 18S rRNA as described in the results section . Alpha ( within-sample ) diversity was measured using Shannon’s Diversity [37] , Faith’s Phylogenetic Diversity ( PD ) [38] , and Pilou’s Evenness [39] indices , calculated per sample within QIIME2 using rarefied counts ( i . e . subsampled to the same sequencing depth across samples; rarefied to 21 , 383 for the 16S dataset , and 2 , 319 for the filtered 18S dataset ) . Boxplot figures for alpha diversity were created using GraphPad Prism v8 , with between group statistical differences determined using the Mann-Whitney or ANOVA nonparametric tests in Prism . Beta ( across-sample ) diversity carried out using rarefied counts ( as above ) was measured using the Bray-Curtis dissimilarity statistic [40] based on compositional dissimilarity between samples taking abundance into account , unweighted UniFrac distances that measures phylogenetic distances between taxa , or weighted UniFrac distances [41] that measures phylogenetic distances and also takes relative abundance into account , as indicated . Principal coordinates ( PCoA ) plots for these indices were either generated within QIIME2 using the EMPeror graphics tools [42] , or in QIIME1 version 1 . 9 . 1 [29] using the make_2d_plots . py script . Analysis of composition of microbiomes ( ANCOM ) [43] , previously reported [44] as a sensitive method ( for samples >20 ) with good control of false discovery rate , was also used within QIIME2 to determine between group differences in taxon abundance . The use of log ratios within this test takes account of differences in sequencing depth . Significance is reported as a W-value which provides a measure of how significant a specific taxon is relative to a number of other taxa used in the analysis .
A total of 12 . 5 million raw paired-end reads were generated across 46 faecal samples ( mean±SD 271 , 509±166 , 138 reads/sample ) for 16S rRNA amplicons , and a total of 1 . 6 million raw paired-end reads ( mean±SD 34 , 112±6 , 852 reads/sample ) for 18S rRNA amplicons . After read pre-processing of 16S rRNA sequence data there were 91 , 923±69 , 706 joined paired-end reads ( minimum read depth 21 , 383; 33±6% of raw paired-end reads retained ) per sample across VL and EC samples . Rarefaction plots ( Part ( a ) in S1 Fig ) based on Shannon’s diversity index confirmed equivalent alpha diversity across the range of read depths from 2 , 000 to >20 , 000 ( i . e . maximum alpha diversity is achieved at 2 , 000 reads ) . Using this read pre-processed data , 348 unique genera were identified from the 16S rRNA reads classified as ASVs within DADA2 . The full taxonomy of these prokaryote 100% sequence identity ASVs is provided in S2 Table . For 18S rRNA the initial pre-processed read depth across experimental samples ranged from 21 , 460–49 , 822 joined paired-end reads ( 95±2% of raw paired-end reads retained ) . However , preliminary ASV classification demonstrated high abundance of unclassified “Eukaryota” across all samples ( S2 Fig ) . As can be seen from part ( a ) in S2 Fig VL cases and endemic controls were equally distributed across the range of abundances for this unclassified taxon . BLAST analysis of these unclassified taxa revealed alignment to bacterial 16S rRNA genes , demonstrating that the EMP primers for 18S rRNA amplification were not eukaryote-specific . These sequences were filtered out , leaving 77–27 , 315 joined paired-end reads per sample across the 46 samples . Using this filtered dataset , 62 unique genera were identified from the 18S rRNA reads classified as ASVs within DADA2 . The full taxonomy of these eukaryotic 100% sequence identity ASVs is provided in S3 Table . Shannon’s index rarefaction plots ( Part ( b ) in S1 Fig ) indicated equivalent alpha diversity across read depths ≥400 ( i . e . maximum alpha diversity that it was possible to determine in our samples was achieved at ≥400 reads ) . However , to provide robustness in downstream analyses of diversity a rarefied read depth of 2 , 319 was used . This resulted in loss of N = 8 VL and N = 7 EC samples for whom read depths were <1000 reads , leaving sample sizes of 15VL and 16EC in the downstream eukaryotic analyses . Fig 1 ( a ) provides a descriptive overview of bacterial gut microflora in our population in the form of a QIIME2-generated bar plot of 16S rRNA-identified bacterial taxa ordered by abundance across the 46 samples . Fig 1 ( b ) provides the bar plot for the mock positive controls ( see also Part ( c ) in S2 Fig for negative controls ) , demonstrating equivalent relative abundance of control bacterial taxa from the ZymoBIOMICS Microbial Community DNA Standard in the presence of increasing concentrations of human DNA . Table 2A shows the aggregated relative abundance at genus-level ( % total counts and family provided in parentheses ) for the 8 most prevalent taxa identified in the study sample . This includes Prevotella 9 ( 33 . 42%; Prevotellaceae ) , Faecalibacterium ( 11 . 32%; Ruminococcaceae ) , Escherichia-Shigella ( 9 . 14%; Enterobacteriaceae ) , Alloprevotella ( 4 . 46%; Prevotellaceae ) , Bacteroides ( 4 . 07%; Bacteroidaceae ) , Prevotella 2 ( 3 . 61%; Prevotellaceae ) , Ruminococcaceae UCG-002 ( 1 . 58%; Ruminococcaceae ) , and Bifidobacterium ( 1 . 54%; Bifidobacteriaceae ) . Fig 1 ( c ) provides a heatmap of relative abundances for the top genera ( Prevotella , Faecalibacterium , Bacteroides and Escherichia-Shigella ) that relate to previously described enterotypes discussed in more detail in the discussion section below [19 , 45] . Our study sample suggests a predominance of enterotype-2 driven by Prevotella , with only a small number of individuals conforming to enterotype-1 driven by Bacteroides and a lack of enterotype-3 driven by Ruminococcus . High Faecalibacterium abundance was generally associated with high Prevotella , whereas Escherichia-Shigella was associated with low Prevotella . Bacterial taxa ( e . g . Dialister , Megasphaera , Mitsuokella , Lactobacillus ) previously shown to be characteristic of Indian gut microbiomes [20 , 24 , 46] occurred at <1% in our study sample ( Table 2 ) . Fig 2 ( a ) provides a bar plot of 18S rRNA-identified eukaryotic taxa ordered by abundance across the 46 samples . Fig 2 ( b ) provides the bar plot for the mock positive controls , demonstrating equivalent relative abundance of the two control fungal eukaryotes in the ZymoBIOMICS Microbial Community DNA Standard , together with identification of Leishmania and Toxoplasma DNA spiked into some control samples in-house . Full details of positive and negative control samples are provided in S2 Fig . Table 2B shows the aggregated relative abundance at genus or lowest order level classified ( higher order classification also provided in parentheses ) for the eukaryotic taxa of interest ( i . e . excluding plant genera ) identified in the study sample . This includes Blastocystis sp . MJ99-568 ( 32 . 37%; Order: Stramenopiles ) , Blastocystis ambiguous taxon ( 25 . 49%; Order: Stramenopiles ) , Dientamoeba ( 12 . 12%; Family: Tritrichomonadea ) , Pentatrichomonas ( 10 . 13%; Family: Trichomonodea ) , Entamoeba ( 3 . 45%; Family: Entamoebida ) , Ascaridida ( 0 . 78%; Family: Chromodorea; likely Ascaris round worm cf . below ) , Rhabditida ( 0 . 77%; Family: Chromodorea; likely Strongyloides or hookworm cf . below ) , and Cyclophyllidea ( 0 . 19%; Order: Eucestoda; likely Hymenolepis tapeworm cf . below ) . Of these , unsupervised cluster analysis performed using the QIIME2 feature-table heatmap plugin in which both axes are clustered with the average linkage method using the Bray-Curtis distance metric ( Fig 2 ( c ) ) shows 3 clusters of individuals that are defined by high relative frequencies of: ( A ) N = 10 ( 3 VL; 7 EC ) Blastocystis Ambiguous taxon with/without Dientamoeba and/or Entamoeba; ( B ) N = 11 ( 9 VL; 2 EC ) with Pentatrichomonas and/or Blastocystis sp . MJ99-568 in the absence of taxa from cluster A; and ( C ) N = 10 ( 3VL; 7EC ) absence or low abundance of all 5 of these taxa but high frequency of Saccharomcyes and plant species . For easy reference , we tentatively refer to these below as “eukaryotic enterotypes” A , B and C with the caveat that further work is required to further define eukaryotic enterotypes . Species level identification carried out as part of the overall QIIME2 taxonomic analysis confirmed Pentatrichomonas as P . hominis and Dientamoeba as D . fragilis . Entamoeba was not identifiable to species level . Other studies in India [21 , 23] and elsewhere [18 , 45] have demonstrated changes in the composition of gut flora with age . In our study it was important to determine whether there were any global effects of age and/or gender on microbial diversity that needed to be adjusted for in analysis of differences between VL cases and EC . Parts ( a-b ) in S3 Fig provides bar plots of the 16S rRNA data reordered by age or gender ( and secondarily on relative abundances of taxa; key as per Fig 1 ( a ) ) , respectively . Visual inspection suggests no overt effects of either age or sex on composition of microbial flora . This is confirmed comparing species richness ( Number of ASVs ) and alpha diversity measures ( Faith’s PD , Pielou Evenness , Shannon’s diversity index ) between genders ( Parts ( c-f ) in S3 Fig ) and beta diversity PCoA plots for weighted UniFrac distances by gender ( Part ( g ) in S3 Fig ) and for age as a continuous variable ( Part ( h ) in S3 Fig ) . Note all comparisons have non-significant adjusted p-values . Similarly , parts ( a-b ) in S4 Fig provide bar plots of the 18S rRNA data reordered by age and gender , respectively . Again , visual inspection suggests no overt effects of either age or sex on composition of microbial flora . This is supported by comparisons of species richness ( Number of ASVs ) and alpha diversity measures ( Faith’s PD , Pielou Evenness , Shannon’s diversity index ) between sexes ( Parts ( c-f ) in S4 Fig ) and PCoA plots for weighted UniFrac distances by gender ( Part ( g ) in S4 Fig ) and for age as a continuous variable ( Part ( h ) in S4 Fig ) . In this case there was borderline significance ( p = 0 . 048 ) for difference between genders for Faith’s phylogenetic diversity ( all other p-values were not significant ) . On balance , we conclude there are no overt independent effects of age or sex on either 16S rRNA-determined bacterial diversity or 18S rRNA-determined eukaryotic diversity in gut flora that would affect analysis of microbial profiles comparing VL cases with EC . Substructure in microbial diversity due to microgeographic location could also confound downstream analysis comparing VL cases and EC . Generally , EC were matched to VL for location as they were attendants of patients at KAMRC . KAMRC is a centre that draws patients from across Bihar State in India , but principally ( in this study ) from within the district of Muzzaffapur . Part ( a ) in S5 Fig provides a map of subdistricts or blocks within Muzzaffapur District showing participant locations . Part ( b ) in S5 Fig shows the 16S rRNA relative abundance bar plot reordered by subdistrict ( key to taxa as per Fig 1 ( a ) ) . Box plots ( Parts ( c-f ) in S5 Fig ) comparing species richness ( Number of ASVs ) and alpha diversity measures ( Faith’s PD , Pielou Evenness , Shannon’s diversity index ) across blocks where N≥4 indicate no statistically significant differences by subdistrict ( Anova not significant ) . Principle coordinate plots based on weighted UniFrac distances ( Part ( g ) in S5 Fig ) also show no overt clustering by subdistrict . Similar results were obtained for the 18S rRNA data ( S6 Fig ) . We conclude there are no overt differences in gut microflora by subdistrict in this region of Bihar State . Fig 3 ( a ) and 3 ( b ) provide bar plots of the relative abundances of taxa reordered by VL case or EC status for 16S rRNA data and 18S rRNA data , respectively . Species richness ( Number of ASVs ) and alpha diversity box plots ( Faith’s PD , Pielou Evenness , Shannon’s diversity index ) comparing VL cases with EC ( Fig 3 ( c ) –3 ( f ) for 16S rRNA; Fig 3 ( g ) –3 ( j ) for 18S rRNA ) show no statistically significant differences in alpha diversity between these two groups . Beta-diversity PCoA plots based on Bray-Curtis distances ( Fig 4 ( a ) and 4 ( b ) ) that take relative abundance into account or using weighted UniFrac distances ( Fig 4 ( c ) and 4 ( d ) ) that take abundance and phylogenetic distances between taxa into count , also show no clear evidence for clustering based on VL cases versus EC status . Although there were no global differences in gut prokaryotic or eukaryotic microbial composition between VL cases and EC as determined by alpha and beta diversity measures , analysis of 16S rRNA data using ANCOM identified two bacterial genera , Ruminococcaceae UCG- 014 ( Family: Ruminococcaceae ) and Gastranaerophilales_uncultured bacterium ( Phylum: Cyanobacteria; Class: Melainabacteria; Order Gastranaerophilales ) , that were enriched in EC compared to VL cases ( Fig 5 ( a ) and 5 ( b ) ; W-values of 61 and 41 , respectively ) . These differences are also apparent in the aggregated relative abundance data presented in Table 2 , with Ruminococcaceae UCG-014 at 0 . 28% in EC compared to 0 . 05% in VL cases and Gastranaeophilales_uncultured at 0 . 24% in EC compared to 0 . 002% in VL cases . Analysis of 18S rRNA genus-level data using ANCOM did not show differences between VL cases compared to EC . However , visual inspection of the 18S rRNA bar plots ( Fig 2 ( a ) ) suggested a trend for higher abundance of P . hominis in VL cases compared to EC . In contrast , there were more individuals with higher relative abundance of Entamoeba in EC than in VL cases . This is also reflected in the aggregated relative abundance data for eukaryotic genera presented in Table 2B , and further highlighted on a heatmap ( Fig 6 ( a ) ) for eukaryotic protozoa log10 read counts ordered by P . hominis counts in EC and VL cases . This aligns with the “eukaryotic enterotypes” defined above , where 7/9 individuals classified as “eukaryotic enterotype B” defined by relative frequencies of read counts for P . hominis and Blastocystis sp . MJ99-568 were VL patients , and conversely 7/10 individuals classified as “eukaryotic enterotype A” defined by Blastocystis ambiguous taxon , with/without Entamoeba or D . fragilis , were EC . Comparisons of log10 read counts ( Fig 6 ( b ) and 6 ( c ) ; p = 0 . 05 ) and relative abundance ( Fig 6 ( d ) and 6 ( e ) ; p = 0 . 01 ) show that VL cases positive for P . hominis had significantly higher P . hominis burdens compared to EC , whereas this was not the case for the between group comparison of individuals positive for Entamoeba ( p-values not significant ) This suggests that when VL patients acquire P . hominis they have a more intense colonisation whereas when they acquire Entamoeba the relative abundance is the same as for EC individuals . Entero-pathogens have been observed to perturb bacterial diversity in gut flora , for example in the context of a bacterial enterotype defined by Escherichia-Shigella abundance [45] . We therefore looked across our sample to see if such effects could be observed in our study population . Looking across relative abundance plots for 16S rRNA data ( see Figs 1 and 3 ) for VL cases and EC there were no overt group-specific differences in relative abundance of bacterial taxa that could contain pathogenic species , e . g . Escherichia-Shigella . However , we did observe that some individuals positive for Escherichia-Shigella appeared to have reduced diversity in terms of other bacterial species in their microflora . Another eukaryotic genus that contains potentially pathogenic species is Blastocystis . In looking for effects on the bacterial microflora of eukaryotic taxa that might contain putative pathogens , we noted clustering according to Blastocystis abundance on Jaccard index beta diversity PCoA plots ( Fig 7 ( a ) –7 ( c ) ) and unweighted Unifrac PCoA plots ( Fig 7 ( d ) ) . This was true for both Blastocystis sp . MJ99-568 ( Fig 7 ( a ) ) and Blastocystis_ambiguous taxon ( Fig 7 ( b ) ) . Specifically , individuals with high Blastocystis abundance ( 46–98% ) were almost exclusively clustered to the right of the zero co-ordinate on PCoA axis 1 , whereas individuals with low Blastocystis abundance ( <6% ) were clustered to the left . To determine what bacterial taxa might be driving this variation we carried out biplot analyses , results of which are superimposed onto PCoA plots of Jaccard indices based on presence/absence of bacterial taxa ( Fig 7 ( a ) –7 ( c ) ) and onto PCoA plots of unweighted UniFrac distances that take phylogenetic distances into account ( Fig 7 ( d ) ) . The four most significant bacterial taxa driven by eukaryotic Blastocystis abundance were Escherichia-Shigella , Prevotella 9 , Succinivibrio , and Asteroleplasma for the Jaccard index analysis , and Escherichia-Shigella , Bacteroides thetaiotaomicron , Faecalibacterium , Prevotella 9 based on unweighted UniFrac distances . In both cases Escherichia-Shigella had the strongest effect ( as indicated by the length of arrow on the biplots ) , which manifests as a negative correlation ( slope = -0 . 96; r2 = 0 . 18; p = 0 . 003 ) between log10 counts for Escherichia-Shigella and Blastocystis ( Fig 7 ( e ) ) . We therefore conclude that , as previously observed [45] , high abundance of Escherichia-Shigella may be influencing dysbiosis of gut bacterial microflora in this population . Blastocystis relative abundance also had a more general influence on alpha-diversity measures ( Fig 8 ( a ) –8 ( c ) for Blastocystis MJ99-568 , p-values <0 . 001 and 0 . 004 for Faith’s PD and Shannon diversity index , respectively; Fig 8 ( e ) –8 ( g ) for Blastocystis Ambiguous taxon , p-values 0 . 007 , 0 . 013 and 0 . 003 for Pielou’s evenness index , Faith’s PD , and Shannon diversity index , respectively ) and on species richness ( Number of ASVs ) ( Fig 8 ( d ) for Blastocystis MJ99-568 , p<0 . 001; Fig 8 ( h ) for Blastocystis Ambiguous taxon , p = 0 . 008 ) . Notably , high abundance of both Blastocystis taxa was associated with higher species richness and higher measures of alpha diversity . To further identify specific bacterial taxa that might be related to the effects of Blastocystis on bacterial diversity we employed ANCOM to look at the family-level for differential 16S rRNA-determined bacterial abundance between high versus low Blastocystis carriers . This identified two bacterial taxa associated with high versus low Blastocystis count ( Fig 5 ( c ) and 5 ( d ) ) . Bacteroidaceae ( aggregate abundance 4 . 07% , see Table 2 ) were less abundant with high Blastocystis and more abundant with low Blastocystis ( Fig 5 ( c ) ; W-value = 31 ) . Conversely , Clostridiales vadinBB60 group ( aggregate abundance 0 . 89% , see Table 2 ) were more abundant with high Blastocystis and less abundant with low Blastocystis ( Fig 5 ( d ) ; W-value = 30 ) . One important reason for carrying out analysis of taxa based on 18S rRNA sequencing was to determine if it would be sufficiently sensitive to detect helminth species in faecal samples . Notwithstanding loss of read depth and sample numbers due to bacterial contamination of reads , we were able to detect 4 taxa containing pathogenic helminths , namely Ascaridida ( Ascaris ) , Trichocephalida ( Trichuris ) , Rhabditida ( Hookworm–Ancylostoma ) and Cyclophylidea ( Hymenolepis ) . Table 3 compares data based on microscopy and 18S rRNA analysis for those individuals ( N = 36 ) for whom both datasets were available . Analysis using 18S rRNA identified 3/3 positive for Ascaris by microscopy , 0/8 positive for Trichuris by microscopy ( with one that was positive for 18S rRNA but negative for microscopy ) , 7/7 positive for hookworm by microscopy ( plus one positive by 18S rRNA and negative by microscopy ) , and 4/4 positive for Hymenolepis by microscopy .
In this study we have used a meta-taxonomic approach to determine the composition of prokaryotic and eukaryotic gut microflora in a region of Bihar State in North-East India endemic for VL . Amplicon-based sequencing of the V3-V4 region of the 16S rRNA gene showed that the most abundant bacterial taxa identified in faecal samples from this region of India were Prevotella ( 37 . 1% ) , Faecalibacterium ( 11 . 3% ) , Escherichia-Shigella ( 9 . 1% ) , Alloprevotella ( 4 . 5% ) , Bacteroides ( 4 . 1% ) , Ruminococcaceae UCG-002 ( 1 . 6% ) , and Bifidobacterium ( 1 . 5% ) . Our study was also unique in using alternative primers for amplicon-based sequencing of the V9 region of the 18s rRNA gene to identify eukaryotic microflora . Although we encountered one issue with cross-amplification of 16S rRNA sequences due to sequence similarity of primers , we were successful in ( a ) blocking amplification of human 18S rRNA , and ( b ) in amplifying protozoan and metazoan sequences of pathogenic interest . In so doing we determined that the most prevalent protozoan taxa present in this region of India were Blastocystis sp . MJ99-568 ( 32 . 37% ) , Blastocystis ambiguous taxon ( 25 . 49% ) , D . fragilis ( 12 . 12% ) , P . hominis ( 10 . 13% ) , and Entamoeba ( 3 . 45% ) . Others have also used sequencing of 16S rRNA amplicon libraries to study the bacterial gut microbiome of people living in India [20–24 , 46] . A study [20] of the gut microbiomes of 43 healthy individuals from urban and semi-urban districts in Maharashtra state ( in the West of India , including Mumbai ) also found that the microbial community at genus level was dominated by Prevotella ( 34 . 7% ) , Bacteroides ( 15 . 2% ) , and Faecalibacterium ( 5 . 6% ) , with additional prominant species Megasphaera ( 4 . 7% ) and Dialister ( 3 . 9% ) . Similarly , the genera Prevotella ( 4 . 5% ) and Megasphaera ( 8 . 5% ) predominated in the gut microflora of 34 healthy Indian subjects from two urban cities ( Delhi from the North; Pune from the West , also near Mumbai ) [46] and were reported to be a distinctive feature of Indian gut microbiota compared with western cultures and to be a component of sharing OTUs with omnivorous mammals [46] . Elsewhere in India [24] , the gut microbiome of a cohort ( N = 110 ) of healthy subjects from North-Central India ( Bhopal City , Madhya Pradesh ) consuming a primarily plant-based diet was dominated by Prevotella , whereas that of omnivorous individuals from Southern India ( Kerala ) showed more prominent associations with Bacteroides , Ruminococcus and Faecalibacterium . This aligned with previously described human gut microbiome enterotypes [19] , with a high proportion of the North-Central population associated with enterotype-2 ( 73 . 5% ) driven by Prevotella compared to the South India population sample ( 54% enterotype-2 ) where enterotype-1 ( 30 . 3% ) driven by Bacteroides and enterotype-3 ( 16 . 1% ) driven by Ruminococcus were also prevalent . This study also showed that species belonging to the genera Bacteroides , Alistipes , Clostridium , and Ruminococcus were relatively depleted in the Indian population compared to China , Denmark and the USA , whereas Prevotella , Mitsuokella , Dialister , Megasphaera , and Lactobacillus were the major drivers in separating Indian samples from other populations . In our study we found a predominance of enterotype-2 driven by Prevotella , with only a small number of individuals conforming to enterotype-1 driven by Bacteroides and a lack of enterotype-3 driven by Ruminococcus . High Faecalibacterium abundance was generally associated with high Prevotella . Bacterial taxa ( e . g . Dialister , Megasphaera , Mitsuokella , Lactobacillus ) previously shown to be characteristic of Indian gut microbiomes [20 , 24 , 46] occurred at <1% in our study population . Hence this population from North-East India does not fall within the criteria identified by these other studies to be characteristic of Indian populations . One reason for differences between our study and other in India could be differences in collection and processing of samples . Importantly here , there were no differences in the collection and processing of samples from VL cases and endemic control subjects that would influence our search for differences due to VL status . One feature of the bacterial microbiota identified in our study was a high abundance of Escherichia-Shigella ( 9 . 1% aggregate abundance ) in a subset of individuals that could define a different enterotype or sub-enterotype in our population compared to other Indian studies . This could reflect socio-economic status of residents of this region of India , together with higher prevalence of enteric diseases [47] . In a recent study of 475 patients with acute gastroenteritis Castaño-Rodriguez and colleagues [45] found that patient microbiomes clustered into three putative enterotypes , dominated by Bacteroides , Escherichia-Shigella , or Faecalbacterium . The novel Escherichia-Shigella-dominated enterotype found in a subset of patients was predicted to be more pro-inflammatory , and included the presence of Enterobacteriaceae , Streptococcaceae , and Veillonellaceae . In our study high Escherichia-Shigella abundance appeared to be associated with an overall dysbiosis of the gut bacterial flora . Our study was also uniquely able to look at the interaction between prokaryotic and eukaryotic microflora in the gut . Interestingly , Escherichia-Shigella was the major bacterial taxon driving the effect of eukaryotic Blastocystis abundance on bacterial beta diversity measures , with a negative correlation between Blastocystis and Escherichia-Shigella log10 counts . This might suggest that each out-competes the other and/or may reflect other influences of Blastocystis on gut microbial homeostasis ( cf . below ) . Notably , however , there was no evidence for a difference in Escherichia-Shigella or other Enterobacteriaceae between VL cases and EC that might support a role for translocation across the gut of LPS from these bacteria in pathogenesis of VL as has previously been proposed [4 , 5] . As alluded to above , our study was the first to employ 18S rRNA analysis of eukaryotic gut microflora in parallel with analysis of prokaryotic 16S rRNA sequences . Although overall measures of alpha and beta diversity for prokaryotic or eukaryotic microflora did not differ significantly between VL cases and EC , differences were observed at the level of individual bacterial or protozoan taxa . Specifically , using ANCOM we determined that EC had significantly higher levels of Ruminococcaceae UCG- 014 ( Family: Ruminococcaceae ) and Gastranaerophilales_uncultured bacterium ( Phylum: Cyanobacteria; Class: Melainabacteria; Order Gastranaerophilales ) than VL cases . Both of these species are at relatively low aggregate relative abundance in our study population . However , previous studies in mice have shown that Ruminococcaceae UCG-014 is significantly reduced along with weight loss resulting from an intermittent fasting-mimicking diet [48] . This could reflect the effect of VL on cachexia and concomitant weight loss . Other research has noted perturbations of Ruminococcaceae UCG- 014 in a murine model of inflammatory bowel disease [49] and in mice fed on low versus high salt diets [50] . The only published data on uncultured Gastranaerophilales spp . found an increase with aging in mice , which the authors suggested pointed to diminished antimicrobial defence with age [51] . Larger sample sizes and further research will be required to determine the functional significance , if any , of the reduced abundance of Ruminococcaceae UCG- 014 and uncultured Gastranaerophilales in VL cases compared to EC . A more prominent effect was the higher abundance of P . hominis in VL cases compared to EC . P . hominis is an anaerobic flagellated protist that colonizes the large intestine of a variety of mammals , including dogs , water buffalo , cattle , nonhuman primates , and humans [52 , 53] . A recent epidemiological survey of faecal samples from northern China showed that 27% of dogs , 4% of human , and 47% of nonhuman primates carried P . hominis [52] , with multiple authors suggesting that zoonotic transmission poses significant risk [52 , 54 , 55] . P . hominis has been associated with chronic diarrhoea and/or inflammatory bowel disease in animals [56] , nonhuman primates [57] and humans [58] , and is found in 14% of pregnant women in Papua New Guinea . In our study affected VL cases had higher abundance of P . hominis than EC . Due to their immuonological status VL cases may be more vulnerable to zoonotic transmission in the context of close association between humans , dogs and cattle/buffalo in this poor rural region of North-East India . Independently of VL versus EC status , we observed high prevalence of Blastocystis spp . in our study population . Studies in a rural region of Brazil have also observed high prevalence of Blastocystis , with some evidence for genetic heterogeneity of Blastocystis subtypes within the population [59] . The two main genus-level taxa observed in our study , Blastocystis sp . MJ99-568 and Blastocystis ambiguous taxa , had similar effects on bacterial alpha and beta diversity measures . This included the effect of Escherichia-Shigella in driving the influence of Blastocystis on low bacterial beta-diversity , and the negative correlation between Blastocystis and Escherichia-Shigella counts as noted above . Using ANCOM we also determined that high Blastocystis was associated with low Bacteroides and with high Clostridiales vadinBB60 . Given the former , the high prevalence of Blastocystis in our population could account for the low representation of bacterial enterotype-1 normally associated with Bacteroides . Within the gut Bacteroides are generally considered as friendly commensals [60] . Carbohydrate fermentation by Bacteroides produces volatile fatty acids that are absorbed through the large intestine and contribute to the host’s daily energy requirement [61] . High Blastocystis appears to put Bacteroides at a disadvantage thus influencing gut flora homeostasis . Another aspect of the dysbiosis associated with high Blastocystis is its association with high Clostridiales vadinBB60 . Studies in mice have shown that abundance of Clostridiales vadinBB60 is influenced by manipulation of meat protein in the diet and plays a part along with other microbiota in bidirectional signalling between gut and brain to regulate body metabolism [62] . However , nothing is known concerning pathogenicity of this species . In summary , whilst high Blastocystis was negatively correlated with Escherichia-Shigella , it likely contributes to overall dysbiosis of gut microflora in this North-East Indian population particularly to the low prevalence of enterotype-1 normally characterised by high Bacteroides . Finally , an important aim of our study was to determine whether helminths could be detected using 18S rRNA analysis . This analysis yielded aggregate abundances of 0 . 78% for Ascaridida correlating with Ascaris identified microscopically , 0 . 77% for Rhabditida correlating with Strongyloides ( hookworm ) by microscopy , and 0 . 19% Cyclophyllidea correlating with Hymenolepis tapeworm by microscopy . The only apparent failure was detection of Trichuris using 18S rRNA in participants for whom small numbers of eggs had been identified in the stool samples , although we were able to identify Trichocephalida amplicons in one study participant negative by microscopy . Our inability to identify all Trichuris-infected individuals could also reflects loss of power in the 18S rRNA sample due to the presence of a large degree of contaminating bacterial 16S rRNA amplification caused by our demonstration that the EMP primers were not eukaryote-specific as originally reported . This has now been corrected on the EMP web-site . The use of eukaryote-specific 18S rRNA primers in future studies should allow more efficient detection of eukaryotic taxa in human faecal DNA , including the full range of soil transmitted helminth species . Thus , while the sample size employed in this scoping study was underpowered to determine any association between soil transmitted gut helminths and VL , our approach combining analysis of 16S rRNA and 18S rRNA analysis should work well in future studies designed to determine the interplay between gut microbiota and susceptibility to infectious diseases such as VL . Such studies may benefit initially by combining amplicon sequencing with a qPCR panel for eukaryotes which would also provide information on parasite load . In conclusion , this scoping study has provided novel data on prokaryotic and eukaryotic gut microbiota for individuals living in this region of North-East India endemic for VL . Interesting dysbioses have been observed when considering the influence of putative pathogenic prokaryotic and eukaryotic species on global measures of bacterial diversity in the gut , and some important associations seen between VL and the eukaryotic pathogen P . hominis . Our study provides useful baseline data upon which to develop a broader analysis of pathogenic enteric microflora and their influence on gut microbial health and NTDs generally .
|
Visceral leishmaniasis ( VL ) , also known as kala-azar , is a potentially fatal disease caused by intracellular parasites of the Leishmania donovani complex . VL is a serious public health problem in rural India , causing high morbidity and mortality , as well as major costs to local and national health budgets . People at risk of VL are also at risk of other neglected tropical diseases ( NTDs ) including soil transmitted helminths ( worms ) . Intestinal worms are potent regulators of host immune responses often mediated through cross-talk with gut bacteria . Here we have used a modern DNA sequencing approach to determine the composition of microbiota in stool samples from VL cases and endemic controls . This allows us to determine all bacteria , as well as all single-celled and multicellular organisms , that comprise the microorganisms in the gut in a single sequencing experiment from a single stool sample . In addition to providing valuable information concerning commensal and pathogenic gut micro-organisms prevalent in this region of India , we find some specific associations between single-celled gut pathogens and VL case status .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"gut",
"bacteria",
"helminths",
"tropical",
"diseases",
"geographical",
"locations",
"india",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"animals",
"protozoans",
"enterobacteriaceae",
"escherichia",
"neglected",
"tropical",
"diseases",
"cellular",
"structures",
"and",
"organelles",
"bacteria",
"infectious",
"diseases",
"blastocystis",
"zoonoses",
"protozoan",
"infections",
"ribosomes",
"people",
"and",
"places",
"biochemistry",
"rna",
"eukaryota",
"ribosomal",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"asia",
"leishmaniasis",
"biology",
"and",
"life",
"sciences",
"ruminococcus",
"non-coding",
"rna",
"organisms"
] |
2019
|
Meta-taxonomic analysis of prokaryotic and eukaryotic gut flora in stool samples from visceral leishmaniasis cases and endemic controls in Bihar State India
|
Surgical technique , including suture placement and tension , is believed to contribute to the outcome of bilamellar tarsal rotation surgery for trachomatous trichiasis . However , the immediate post-operative appearance that minimizes the chance of recurrence and other adverse outcomes has not been investigated . To explore whether the degree of correction immediately after surgery is predictive of surgical outcome at the 6-week post-operative visit , photographs taken immediately after surgery were used to predict surgical outcomes , including the severity of eyelid contour abnormality and trichiasis recurrence . Both eyelid contour abnormalities and recurrence were accurately predicted from the immediate post-operative photographs by an experienced oculoplastic surgeon 85% and 70% of the time , respectively . Participants with a “slight over-correction” that resulted in no eyelid contour abnormality and no recurrence were used to identify immediate post-operative contours that lead to a successful surgical outcome . The immediate post-operative eyelid contour is an important indicator of post-operative success of BLTR surgery . Based upon our findings , we developed a Surgery Photocard . This card illustrates some examples of immediate post-surgical appearances , which led to a successful outcome , as well as sub-optimal appearances , which led to poor surgical outcomes . The card also provides suggestions for improving the appearance by adjusting the suture placement or tension based upon standard oculoplastic principles .
Despite a World Health Organization ( WHO ) initiative in 1998 to eliminate trachoma by 2020 , this disease continues to plague poor regions of many developing countries . Even if active trachoma were to be eradicated today , millions of individuals would be at risk of blindness due to the trichiasis and corneal scarring caused by previous repeated ocular infection with Chlamydia trachomatis . [1] For those individuals who have already developed conjunctival scarring and trichiasis , the principle option to prevent unnecessary blindness is surgery . The WHO has endorsed the Bilamelar Tarsal Rotation ( BLTR ) procedure for the correction of trachomatous trichaisis ( TT ) , [2] , [3] and in many countries , this surgery is performed by non-physicians , hereafter referred to as surgical technicians . Recurrence of trichiasis after surgery is a key concern . Several studies have suggested that in addition to concurrent infection and inflammation , [4] surgical technique may contribute significantly to trichiasis recurrence . [4]–[7] Additionally , other adverse outcomes , such as eyelid contour abnormalities ( ECAs ) and eyelid closure defects , can occur after BLTR . The exact surgical factors that contribute to recurrence and other adverse outcomes have not been fully elucidated . The BLTR procedure can be divided into 2 parts: the full-thickness incision creating a distal eyelid fragment containing the eyelid margin and lash line and the external rotation of this distal eyelid fragment by suture placement . The TT clamp , currently under investigation in a prospective , randomized trial , was designed to modify how the incision is created and the subsequent effect on BLTR outcome . [8] However , the importance of suture placement and tension on the BLTR outcome has yet to be investigated . The current study aims to investigate how the immediate post-operative eyelid contour relates the BLTR outcome at 6 weeks . During pre-study training to test the TT clamp against standard BLTR instrumentation , and for several weeks during the prospective , randomized clinical trial , surgical technicians performing BLTR were observed placing the rotating sutures . We noted that even when the incision was made correctly , the placement and tying of the sutures had a tremendous impact on the immediate post-operative eyelid contour and correction of the trichiasis . We , therefore , worked with the surgical technicians during the training to achieve the immediate post-operative “slight over-correction , ” recommended in the WHO surgical manual . [3] When we observed significant over- or under-correction , we asked the surgical technicians to remove and replace the sutures . Typically , this correction resulted in a better eyelid contour and degree of eversion ( Figures 1 and 2 ) . Given these observations , we investigated whether the degree of correction immediately after BLTR was predictive of surgical outcome at the 6-week post-operative visit .
The ongoing Partnership for the Rapid Elimination of Trachoma ( PRET ) Surgery trial is designed to compare the TT clamp against the standard BLTR instrumentation . In rural southern Tanzania , 1917 participants with trichiasis and no previous trichiasis surgery in at least one eye were randomized to BLTR with either the TT clamp or standard instrumentation . A total of 3345 eyes underwent surgery as part of the study . For each participant , photographs were taken directly before and immediately after surgery , as well as at the 2- and 6-week follow up visits . At the 6-week post-operative visit , participants were examined for trichiasis recurrence , pyogenic granuloma formation and ECAs . Using a standardized grading system , photographs of each participant were graded for ECAs defined as none , mild , moderate , or severe . [9] Table 1 summarizes the grading system . Photographs were taken with the participant looking up , in order to completely visualize the eyelid margin and contour . When recurrence was present , the field grader recorded the number of lashes touching the globe as well as the location of trichiatic lashes . The location was defined by dividing the eyelid into three equal segments and describing the location of each lash base as nasal , central , or temporal . The current analysis utilizes immediate post-operative photographs and 6-week outcomes data from this ongoing study . At the 6-week PRET study visit , 3341 eyelids were evaluated for postoperative outcomes . Using the PRET-study outcome data for recurrence and ECAs , 200 eyelids were randomly selected from those that were normal or had only one outcome . An equal distribution of normal , recurrence , or ECA ( moderate or severe ) was chosen . Therefore , the 6-week photograph set consisted of 50 normal eyelids , 50 with recurrence , 50 with moderate ECA , and 50 with severe ECA . In order for an eyelid to be selected , both immediate post-operative and 6-week follow up pictures of gradable quality were required . 16 of 3258 ( 0 . 5% ) eyelids with gradable photographs were excluded because both recurrence and ECA were noted at 6 weeks . Approximately 4% of eyelids were excluded due to ungradable photographs ( blurry quality , patient not looking up ) . Review of the pre-operative photos confirmed that in no case was an ECA present prior to BLTR surgery . The immediate post-operative photographs were evaluated in a masked fashion by a fellowship-trained oculoplastic surgeon ( SLM ) , who was unaware of the distribution of outcomes . The evaluator was asked to record a predicted single 6-week outcome based on the appearance of the eyelid immediately after surgery - normal , recurrence ( with location noted as nasal , central , and/or temporal ) , or ECA ( moderate or severe ) . The primary aspect evaluated was the degree of rotation of the eyelid . When apparent under-rotation of the eyelid margin was observed in the photograph , recurrence was predicted . Conversely , when over-rotation of the eyelid margin was perceived , an ECA was predicted . This study was approved by the Johns Hopkins Medicine IRB and the National Institute of Medical Research , Tanzania . Each participant provided written informed consent for participation . The study conforms to the Tenets of Helsinki . Eyelids with true 6-week recurrence ( PRET grade ) were excluded from the ECA calculations; similarly , those with ECA at 6 weeks were excluded in the recurrence calculations . The evaluator was instructed to select only one adverse outcome for each eye ( either recurrence or ECA ) . Consequentially , if she scored an eyelid as having recurrence , automatically the eye was classified as “no abnormality” for ECA . Similarly , eyelids graded as having ECA by the evaluator automatically were classified as no recurrence . Evaluator predictions were compared with the true outcome at 6 weeks; sensitivity and specificity of the predicted outcome and corresponding 95% confidence intervals are reported .
The evaluator accurately identified 35 of 50 eyelids with recurrence ( Figure 3 , Table 2 ) , with a sensitivity of 70% ( 95% confidence interval ( CI ) : 57%–83% ) and a specificity of 90% ( 95% CI: 82%–98% ) . Among the eyelids with recurrence , 31 had nasal recurrence , 19 had central , and 14 had temporal recurrence ( some eyelids had recurrence in more than one location ) . The comparison of predicted and actual location ( s ) for recurrence is shown in Table 3 . Specificity was higher than sensitivity for predicting recurrence at each location , and ranged from 78% for nasal , 90% for central , and 94% for temporal . Sensitivity for predicting the location was higher for nasal and/or central recurrence ( 58% each ) than it was for temporal recurrence ( 43% ) ( Table 3 ) . The evaluator accurately identified 85 of 100 eyelids with ECAs ( Figure 4 , Table 4 ) . The number of false positives and false negatives were similar . The sensitivity and specificity of predicting the presence of an ECA were 85% and 88% , respectively . In addition , the severity of the ECA at 6 weeks could also be predicted in more than half of the cases ( 58 of 100 ) based on immediate post-operative photographs ( Table 5 ) . The sensitivity of predicting either a moderate or severe ECA compared to a prediction of no ECA was 64% , and a similar number of eyes were classified as moderate when they were severe as were classified as severe when they were moderate .
Although previous research has looked at various aspects leading to trichiasis recurrence , our study is the first to systematically evaluate the role of immediate post-operative eyelid contour on unfavorable outcomes . We found that the degree of correction and shape of the eyelid immediately after surgery were predictive of surgical outcome at 6 weeks , suggesting a novel strategy that could be employed to lower recurrence rates and the incidence of ECAs . It should be noted that our prediction rate was higher than the expected 25% prediction rate one would expect by chance alone . While specificity was quite high for each outcome , sensitivity was somewhat lower . A re-review of the 15 eyelids with recurrence that was not predicted by the evaluator showed that in the majority of cases , nasal under-correction was obscured by either eyelid swelling or incomplete opening of the post-surgical eyelid . In 2 cases , the immediate post-operative eyelid position appeared ideal , but at 6 weeks the eyelid had central recurrence . These cases highlight the fact that surgical outcomes are multifaceted and the immediate post-operative eyelid contour is not the only predictor of a successful surgery . The sensitivity was lower for the exact location of recurrence than the presence of recurrence . One explanation for this finding is the somewhat subjective division of the eyelid into three segments , resulting in difficulty determining the exact location when the offending lashes are near the junction between 2 locations . By combining the central and nasal locations for analysis , the sensitivity increased to 72% . Similarly , combining central and temporal aspects also resulted in a higher sensitivity ( 75% ) . BLTR is considered an effective surgical procedure for the treatment of cicatricial entropion of the upper eyelid [10] and is one of the WHO's procedures of choice for trachomatous entropion and trichiasis . [3] The procedure was first described by Wies for the treatment of spastic entropion of the lower eyelid . [11] A tendency towards over-correction with the Wies procedure led Ballen , in 1964 , to propose the technique for difficult upper eyelid problems such as cicatricial entropion in the setting of trachoma . [2] In Ballen's manuscript , the tension of the rotating sutures or desired degree of rotation of the distal fragment was not addressed . Twelve years later , Baylis noted that over-correction with the Wies procedure generally resulted from excess rotation due to improperly placed incisions or sutures . [12] He noted that contraction of the orbicularis tended to turn the eyelid margin inward , and with time , over-correction tended to lessen , as did correction of the entropion . He suggested that , “A small overcorrection , therefore , is desirable with the Wies procedure . ” [12] The WHO manual states that the sutures “should be tied firmly enough to produce a slight overcorrection , ” [3] but the manual does not elaborate on exactly what represents a slight over-correction . Other manuscripts also have referred to the desired “slight over-correction” without further description , [6] and some have advocated intentional over-correction , particularly in cases of severe forms of TT . [13] However , to our knowledge , what represents a “slight over-correction” or the optimum post-operative contour has not been specifically described or evaluated in the literature . In our study , we were able to use the post-operative contour to predict the surgical outcome , predicting recurrence when apparent under-correction was observed and predicting ECA when over-rotation of the eyelid was observed . Our successful predictions suggest that a future study of how the degree of rotation affects the outcome is warranted . In the PRET Surgical trial , although we initially provided brief training to the surgical technicians to assess the immediate post-operative contour , we still saw unfavorable outcomes at 6 weeks . One explanation is that while an experienced oculoplastic surgeon can accurately assess the immediate post-operative outcome , non-physician surgical technicians may have more difficulty with this evaluation . However , in our experience , surgical technicians improve with additional training . Another possibility is that over time , the surgical technician reverts back to pre-training practices . Therefore , based on our data , photographs , and clinical experience , we developed a Surgical Photocard to aid the BLTR surgical technician in the field ( Figures 5 and 6 ) . We tested a prototype card during a retraining session for the PRET surgical technicians prior to commencing surgery for our participants with recurrence . This card shows examples of immediate post-operative contours that lead to successful surgical outcomes ( Figure 5 ) , as well as examples of immediate post-operative contours that lead to recurrence or ECAs ( Figure 6 ) . In the cases of sub-optimal appearance , the card also describes how to adjust suture tension or placement to improve the immediate post-operative contour based upon basic oculoplastic principles and the experience of an oculoplastic surgeon ( SLM ) . This card can be kept with the other surgical supplies so that it is consistently brought to the field during surgical campaigns , providing a ready reference for the surgical technician . Several barriers to improving surgical outcomes still exist . In countries where funds and supplies are limited and the need for surgery is great , convincing surgical technicians to use additional suture material and time to replace them may be difficult . Another barrier to improving surgical outcomes is the common practice of someone other than the operating surgical technician removing the sutures . When this happens , the surgical technician misses the opportunity for valuable feedback . Recent data suggests that absorbable sutures are beneficial to the ultimate outcome;[14] however , they also reduce the likelihood that the surgical technician will have an opportunity for post-operative feedback . One limitation of this study is the use of photographs to evaluate the immediate post-operative eyelid position rather than in-person evaluation . While we have demonstrated a good agreement between field and photograph grades of postoperative outcomes at 6 weeks , [9] we have not studied the same types of correlation for the evaluation of immediate post-operative eyelids . At completion of surgery , the eyelid is still anesthetized , which sometimes makes it difficult for the participant to cooperate with the photographer and fully open or close the eyelid . Additionally , the eyelid can be quite swollen , obscuring the eyelid margin contour in a photograph . Thus , sub-optimal photographs can make it more difficult to evaluate the post-operative eyelid position than the evaluation would be in person . Another limitation of the study is the relatively short follow-up . There is good evidence that the eyelid position continues to change and recurrence continues to occur after 6 weeks . It is possible that some of the false positive recurrences were actually recurrences at the 1- or 2-year follow-up visits and that ECAs may change over time . This study was designed to select patients with just one outcome ( normal , recurrence , or ECA ) . While some participants had more than one adverse outcome at 6 weeks , these participants comprised only 0 . 5% of the gradable photographs . At the 6 week visit , only 4% of eyes had a missing or poor quality photograph . In summary , we found that the degree of correction and shape of the eyelid immediately post-operatively were predictive of surgical outcome at 6 weeks . This finding suggests that a post-operative assessment combined with improved suturing technique could lead to better outcomes , including lower recurrence rates and a lower incidence of ECAs . Additionally , we identified immediate post-operative contours representing the “slight over-correction” that leads to a successful BLTR outcome , which if used in training and during practice has the potential to aid the TT surgical technician in their assessment .
|
Trichiasis is a potentially blinding consequence of trachoma . The World Health Organization has promoted the bilamellar tarsal rotation ( BLTR ) procedure as a treatment for trichiasis from trachoma . Even if trachoma were to be eradicated today , a great number of individuals would still develop trichiasis and lose vision unless they received surgical treatment . Unfortunately , the recurrence rate after BLTR can be quite high , and surgical factors are thought to contribute to unfavorable outcomes . In this study we evaluated the relationship between immediate post-operative contour and 6-week outcomes utilizing immediate post-operative photos of 200 trichiasis surgeries . We found good agreement between the predicted and actual outcome . The analysis allowed the identification of immediate post-operative eyelid contours that were most likely to lead to successful surgery at 6 weeks , as well as contours that resulted in adverse outcomes . The description of the target degree of rotation and how to adjust the rotating sutures to best achieve this ideal should help surgeons worldwide improve their surgical technique and outcomes .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"ophthalmology",
"neglected",
"tropical",
"diseases",
"trachoma"
] |
2012
|
Relationship between Immediate Post-Operative Appearance and 6-Week Operative Outcome in Trichiasis Surgery
|
Enteroviruses ( family of the Picornaviridae ) cover a large group of medically important human pathogens for which no antiviral treatment is approved . Although these viruses have been extensively studied , some aspects of the viral life cycle , in particular morphogenesis , are yet poorly understood . We report the discovery of TP219 as a novel inhibitor of the replication of several enteroviruses , including coxsackievirus and poliovirus . We show that TP219 binds directly glutathione ( GSH ) , thereby rapidly depleting intracellular GSH levels and that this interferes with virus morphogenesis without affecting viral RNA replication . The inhibitory effect on assembly was shown not to depend on an altered reducing environment . Using TP219 , we show that GSH is an essential stabilizing cofactor during the transition of protomeric particles into pentameric particles . Sequential passaging of coxsackievirus B3 in the presence of low GSH-levels selected for GSH-independent mutants that harbored a surface-exposed methionine in VP1 at the interface between two protomers . In line with this observation , enteroviruses that already contained this surface-exposed methionine , such as EV71 , did not rely on GSH for virus morphogenesis . Biochemical and microscopical analysis provided strong evidence for a direct interaction between GSH and wildtype VP1 and a role for this interaction in localizing assembly intermediates to replication sites . Consistently , the interaction between GSH and mutant VP1 was abolished resulting in a relocalization of the assembly intermediates to replication sites independent from GSH . This study thus reveals GSH as a novel stabilizing host factor essential for the production of infectious enterovirus progeny and provides new insights into the poorly understood process of morphogenesis .
Enteroviruses , belonging to the family of the Picornaviridae , are non-enveloped , icosahedral viruses with a positive , single-stranded genome . Enteroviruses comprise many important pathogens of humans and animals . Although most enterovirus infections subside asymptomatically or mildly , they can also result in severe morbidity and even mortality . Polioviruses cause paralytic poliomyelitis; rhinovirus infections are associated with exacerbations of asthma and chronic obstructive pulmonary disease and enterovirus 71 may cause life-threatening encephalitis , in particular in Asia . Also coxsackieviruses and echoviruses have been reported to cause acute clinical manifestations , including fulminant sepsis , aseptic meningitis and myocarditis [1] . Epidemiological studies strongly suggest a linkage between coxsackieviruses and the development of type 1 diabetes mellitus [2] . Apart from polio , no vaccines are available that can protect against enteroviral infections . No drugs have been approved so far for the treatment or prophylaxis of enterovirus infections [3] . The RNA genome of enteroviruses encodes four structural ( VP1 , VP2 , VP3 and VP4 ) and seven nonstructural proteins ( 2A , 2B , 2C , 3A , 3B , 3C and 3D ) . Following receptor-mediated binding to the cell surface , the capsid is destabilized and the viral genome is delivered into the cytoplasm [4] , [5] . Genomic RNA is then translated into a viral polyprotein which is processed into the structural capsid proteins and the non-structural proteins that are involved in the replication and production of new positive-strand RNAs via a negative-strand RNA intermediate . These newly synthesized positive-strand RNAs are produced on replication vesicles derived from the Golgi complex and trans-Golgi network ( TGN ) and either enter a new round of translation/replication or are packaged into capsid proteins to yield new infectious virus particles [6] , [7] . Despite the fact that morphogenesis represents an important stage at the end of the virus replication cycle , many aspects of the molecular mechanisms dictating the assembly of viral particles remain obscure . It is known that this process occurs in a multi-tiered manner . In a first step , the capsid precursor P1 , consisting of four structural components VP1 to VP4 , is released from the P2–P3 polyprotein by a cis-cleavage , carried out by 2Apro [8] . Studies with geldanamycin , a specific inhibitor of the cellular protein chaperone , heat shock protein 90 ( Hsp90 ) , have shown that by interacting with Hsp90 , the nascent P1 is maintained in a processing-competent conformation and is protected from proteasomal degradation [9] , [10] . This interaction with Hsp90 ( and possibly Hsp70 and several cofactors ) , allows the precursor polyprotein P1 to be processed by 3CDpro thereby yielding an immature structural unit , the protomer particle , containing one copy of VP0 ( consisting of VP4 and VP2 ) , VP3 and VP1 [11] , [12] . Five of these protomers will then self-oligomerize into a pentameric particle . This self-assembly has been shown to be dependent on the modification of an N-terminal glycine residue of VP0 with myristate , stabilizing interactions at the fivefold axis between protomers forming pentameric subunits [13]–[16] . Further assembly of twelve pentameric subunits can generate a next higher-order particle , the empty capsid . It has been long a subject of debate whether newly synthesized progeny RNA is inserted into these preformed empty capsids or whether pentameric particles associate around an active replicating viral genome . An essential insight into this mechanism was provided by several studies suggesting a role for the non-structural protein 2C during encapsidation [17] , [18] . Evidence was provided that viral encapsidation is coupled to genomic RNA synthesis resulting from a direct interaction between the RNA replication machinery and capsid proteins , possibly pentameric particles [19] . This hypothesis was extended by a recent study in which genetic and biochemical evidence was provided that newly synthesized positive-stranded genomes were encapsidated without an apparent involvement of an RNA packaging signal [20] . More specifically , a direct protein-protein interaction between 2C , as an essential component of the replication complex , and capsid protein VP3 , possibly as part of a pentameric particle , was demonstrated . In this way pentameric particles will only condense at the site of replication around newly synthesized RNA genomes [19]–[21] . Further evidence for an interaction of protein 2C with capsid proteins during encapsidation was recently provided by a charged-to-alanine-scanning mutagenesis approach within protein 2C resulting in encapsidation-defective poliovirus and the identification of compensatory mutations in VP1 and VP3 [22] . In a last step and concomitantly with the encapsidation of genomic RNA , VP0 is processed into VP2 and VP4 , probably by an RNA-dependent autocatalytic process , eventually yielding infectious virions [23] . Several studies provided evidence for a critical role of glutathione ( GSH ) during enterovirus morphogenesis [24] , [25] . GSH is the most prevalent non-protein thiol in the animal cell and is involved in a multitude of cellular processes , including maintaining the cellular redox potential and detoxification of xenobiotics [where GSH serves as a cofactor in conjugation reactions catalyzed by glutathione peroxidases ( GPx ) and glutathione-S-transferases ( GST ) ] [26] , [27] . In addition , GSH has been shown to play a role in signal transduction , gene expression , cell proliferation and apoptosis by either maintaining a favorable redox status of the cell or by directly interacting with cysteine residues in polypeptidic chains [27] . An imbalance in GSH has been observed in various pathologies , including cancer , neurodegenerative disorders , cystic fibrosis and aging [28] . Through drug inhibition studies using buthionine sulfoximine ( BSO ) , a known irreversible inhibitor of the γ-glutamylcysteine synthetase , ( a key enzyme in the glutathione biosynthesis pathway ) , it was shown that GSH depletion results in a complete inhibition of the formation of infectious enterovirus progeny without affecting viral RNA and protein synthesis [24] , [25] . The observed dependence on GSH during enterovirus morphogenesis was most likely not by affecting favorable cellular redox conditions since addition of other reducing agents could not substitute for GSH . The exact mechanism by which GSH is implicated in the formation of virus progeny remains obscure to date . In this study , we describe the identification of TP219 , a novel inhibitor of the assembly of several enteroviruses that acts by scavenging free GSH . By employing this inhibitor , we provide evidence that a direct interaction between VP1 and GSH is essential for the transition into or the stabilization of pentameric particles before assembling into higher-order particles . Furthermore , genetic analysis of coxsackievirus B3 ( CVB3 ) escape mutants from GSH requirement demonstrated the relevance of a surface-exposed methionine at the interface of two protomers that may serve as a surrogate for GSH . Structural analysis of several naturally GSH-independent enteroviruses provided further evidence for the importance of this single residue in assembly . A role for GSH as a host factor during enterovirus morphogenesis is discussed .
We recently described a series of 9-arylpurines that inhibit in vitro enterovirus replication [29] . From this series , we selected TP219 ( designated compound 26 in reference [29] ) for further mechanistic studies ( Fig . 1A ) . TP219 inhibits CVB3-induced CPE in BGM cells showing a 50% effective concentration of 17±0 . 65 µM with only little adverse effects on the host cell at high concentrations ( Fig . 1B ) . TP219 was shown to exert antiviral activity against some enteroviruses ( e . g . coxsackieviruses A16 , A21 and A24 , CVB3 , echovirus 1 and 9 and poliovirus Sabin 3 ) but proved inactive against others ( e . g . echovirus 11 , poliovirus Sabin 1 and enterovirus 71 ) ( Table 1 ) . Interestingly , a cell type-dependent antiviral effect was observed . TP219 proved to be active against CVB3 , echovirus 9 and CVA21 in Vero , BGM or MRC-5 cells , but not in HeLa or RD cells . EV71 and PV Sabin 1 remained insensitive when tested on Vero or BGM . Thus , TP219 inhibits the replication of a selection of enteroviruses in selected cell lines . To explore the mechanism by which TP219 inhibits CVB3 replication , we systematically examined its effect on the different steps in the virus replication cycle . Towards that end , we used a reporter CVB3 , expressing a Renilla Luciferase gene ( RLuc-CVB3 ) placed between the 5′ UTR and the P1-coding region followed by a 3CDpro cleavage site allowing for proteolytic processing of the polyprotein ( Fig . 1C ) . BGM cells were infected with RLuc-CVB3 in the presence or absence of TP219 . Geldanamycin and guanidine HCl ( GuaHCl ) were used as positive controls . GuaHCl is a known inhibitor of viral RNA replication; geldanamycin is a known inhibitor of Hsp90 and was previously reported to inhibit P1 maturation without affecting viral RNA replication [9] . CVB3 RNA replication was completely blocked in the presence of GuaHCl , but not in the presence of TP219 and geldanamycin ( Fig . 1C left panel ) . Lysates of the infected cell cultures were subjected to end point titration to determine virus yields ( Fig . 1C right panel ) . In the absence of compound , high virus titers were measured , indicating that the virus encapsidated the viral genome efficiently and was capable of infecting new cultures . In contrast , treatment with TP219 or geldanamycin resulted in a pronounced reduction of virus titers , indicating that , despite normal RNA replication levels , no infectious virus particles were formed . Thus , TP219 treatment did not affect early stages ( such as attachment , entry or uncoating ) or RNA replication . The fact that TP219 did not affect viral RNA replication indicates formation of intact and functional non-structural replication proteins . The defect in virus production might however be due to an adverse effect on 3C ( D ) pro-mediated proteolytic processing of structural capsid proteins . To test this possibility , CVB3-infected BGM cells were labeled with [35S]Met both in the absence or presence of TP219 between 5 . 5 and 6h p . i . ( Fig . 1D ) [9] . During this period CVB3 efficiently shuts off translation of cellular mRNA hence only viral proteins are radiolabeled . TP219 was shown not to directly affect 3C ( D ) pro-mediated proteolytic processing of the capsid-coding region as normal levels of VP0 and VP1 were observed ( Fig . 1D ) . However , it cannot be ruled out that TP219 might have an indirect effect on proteolytic processing by affecting host cell metabolism which might require longer incubation periods . Thus , these data demonstrate that TP219 does not inhibit entry , translation or RNA replication of CVB3 , but interferes with virus morphogenesis . The effect of TP219 on the formation of different assembly intermediates ( protomers [5S] , pentamers [14S] , empty capsids [75S] and mature virions [150S] ) was next studied by sedimentation through a sucrose density gradient and analysis of gradient fractions by trichloroacetic acid ( TCA ) precipitation and liquid scintillation counting . In parallel , buthionine sulfoximine ( BSO ) , a previously described enzymatic inhibitor of GSH synthesis and enterovirus morphogenesis , was included as a reference [25] . Mock ( Fig . 2A–D ) , TP219- ( Fig . 2A and 2B ) or BSO- ( Fig . 2C and 2D ) treated BGM cells were infected with CVB3 . 14S pentamers , 75S empty capsids and the 150S virions were readily detected in lysates of the CVB3-infected control cultures . Because of the high background near the top of the 6–25% sucrose gradient ( Fig . 2A and 2C ) , 5S protomeric particles could not be distinguished . In the presence of TP219 or BSO , 14S , 75S or 150S assembly intermediates were not observed in the lysates of the infected cultures . Thus , TP219 , akin to BSO , prevents the formation of 14S assembly intermediates and as a consequence the formation of consecutive higher-order particles . To allow a better dissection of the 5S and 14S viral particles and to reduce interference of labeled cellular and viral material , every two fractions of the 6–25% gradient were pooled and subjected to immunoprecipitation using polyclonal anti-CVB antibodies ( Fig . 3A and 3B ) . Both 5S protomeric particles and 14S pentameric particles were detected in CVB3-infected control cultures . In analogy to the TCA-precipitated fractions , 14S peaks were largely reduced in the presence of either TP219 ( Fig . 3A ) or BSO ( Fig . 3B ) . Also the levels of 5S protomeric particles were somewhat reduced in the presence of both compounds . Intriguingly , several peaks of unknown nature with a sedimentation coefficient varying from 5S to 14S appeared . To further corroborate the above , we next analyzed the effect of TP219 and BSO on the distribution of VP1 in 6–25% sucrose density gradient fractions ( Fig . 3C–E ) . We therefore performed an immunoblotting analysis , using a monoclonal anti-VP1 antibody , on every two consecutive TCA-precipitated fractions and quantified the immunblot by densitometric analysis . Considering that protein composition ( VP0 , VP1 and VP3 ) remains intact during the transition of 5S into 14S particles , the data presented in Figure 3C–E clearly show that , akin to the immunoprecipitation experiments , TP219- ( Fig . 3D ) and BSO- ( Fig . 3E ) treatment affects the assembly pattern - as represented by VP1 - and results in a reduction of the levels of 14S pentameric particles as compared to the DMSO control ( Fig . 3C ) . Taken together , these data suggest that both TP219 and BSO affect the transition of protomeric subunits into pentameric subunits , or alternatively that they may affect the stabilization of pentameric particles , thereby resulting in the formation of several new assembly intermediates . We next explored whether , TP219 affects like BSO intracellular GSH levels . Intracellular levels of GSH in TP219- or BSO-treated BGM cells were monitored over a period of 24 hours . BSO treatment resulted in a progressive depletion of endogenous intracellular GSH levels and almost complete depletion ( 89±0 . 47% ) was observed 24 hours post incubation ( Fig . 4A ) . In contrast , already following one hour of incubation in the presence of TP219 , the intracellular GSH content was reduced by more than 90% . GSH levels were completely depleted ( 99±0 . 048% ) following 3 hours of incubation ( Fig . 4A ) . Conversely , treatment of HeLa cell cultures with TP219 hardly resulted in any reduction in GSH levels ( Fig . S1 ) . In parallel , we also determined the total intracellular GSH concentration in BGM and HeLa cells , being 6 . 3±0 . 40 nmol/mg protein and 7 . 5±0 . 18 nmol/mg protein respectively . It is unlikely that this small difference accounts for the TP219 insensitivity of HeLa cells . However , it has been shown previously that tumor cell lines ( such as HeLa cells ) may have an increased GSH biosynthesis [30] . This could then result in a rapid replenishment of the depleted GSH pools following TP219-treatment . Within this context , it is likely that the differences in antiviral activity between HeLa ( and most likely also RD ) and BGM ( and most likely also Vero and MRC-5 ) ( see Table 1 ) cells are linked to GSH depletion or the lack thereof . Depletion of GSH can be the consequence of an ( i ) increased GSH efflux ( ii ) increased cellular GSH oxidation or ( iii ) inhibition of GSH synthesis . TP219 did not markedly increase levels of extracellular GSH or intracellular GSSG ( oxidized GSH ) ( data not shown ) . Moreover , the kinetic profile of GSH depletion in BSO treated cells differed from that in TP219-treated cells suggesting a mechanism of GSH depletion independent from inhibition of GSH synthesis . GSH is known to react as a nucleophile to form S-substituted conjugates even in the absence of catalyzing enzymes . The chlorine at position 6 of the chloropurine TP219 might be susceptible to a nucleophilic attack by the -SH group of GSH . The formation of such a conjugate can be determined by high performance liquid chromatography ( HPLC ) and mass spectrometry ( MS ) analysis . To explore this possibility , glutathione ethyl ester ( GEE ) ( which was used as a surrogate because of its higher hydrophobicity and MS signal intensity compared to GSH ) [31] was co-incubated with TP219 . At different time points after incubation the samples were analyzed by HPLC . The peak corresponding to TP219 ( retention time ( rt ) = 12 . 6 min ) was significantly reduced in function of time with the concomitant formation of a new peak ( rt = 10 . 7 min ) ( Fig . S2 ) , which was analyzed by LC/MS/MS . Accurate mass determination ( electrospray negative ion mode ) indicated a peak with m/z 570 . 71 ( corresponding to C25H28N7O7S ) that was assigned as the conjugate [M-H]− ( Fig . 4B ) . Collision induced dissociation ( CID ) of this anion by MS/MS afforded fragment ions at m/z 300 . 11 , 210 . 08 and 128 . 03 ( Figure 4B ) , which have been proposed as an unambiguous identification of GEE trapped metabolites [32] . Taken together , these data suggest that TP219 forms an adduct with GSH , even in the absence of enzymes [31] , [33] . To confirm the role of GSH in CVB3 morphogenesis , TP219-treated infected cells were supplied with exogenous GEE ( which has better cell permeability than GSH ) . This resulted in a dose-dependent rescue of virus production ( Fig . 4C ) . Since GSH is essential for maintaining a reduced cellular environment , GSH depletion might indirectly result in an antiviral status . Therefore , we studied the effect on virus infectivity of the reducing agents N-acetyl cysteine ( NAC ) and dithiothreitol ( DTT ) compensating for GSH depletion . In agreement with earlier observations made for BSO neither of them could serve as a surrogate for GSH ( data not shown ) [25] . To identify whether we could drive CVB3 to replicate in the absence of GSH , three independent pools of CVB3 were repeatedly propagated in the presence of increasing concentrations of TP219 , and thus decreasing concentrations of GSH . Following 17 sequential passages , several GSH-independent variants emerged that replicated ( in contrast to the wildtype ( wt ) virus ) effectively in the absence of GSH . To delineate the genetic basis for the observed resistance to GSH depletion , the complete nucleotide sequence of the three selected variants was determined and compared with the wt viral genome . In all three mutant pools a T77M mutation in VP1 in combination with one or two mutations in VP1 ( V150I and N212S ) or VP3 ( A180T and K135R ) ( Table 2 ) was observed . A number of other mutations ( in 2A and 3A ) were not further considered , since they did not occur in all 3 variants . To assess the precise contribution of the identified mutations to the GSH-independent phenotype , recombinant viruses were engineered . Seven constructs were generated containing the identified mutations , either alone or combined , and RNA transcripts were transfected into BGM cultures . Mutation T77M alone ( or combined ) , proved sufficient to bypass the need for GSH ( Table 3 ) . Constructs containing solely VP1 mutation V150I or N212S or VP3 mutation K115R still demonstrated a GSH-dependent phenotype . The T77M mutant CVB3 proved also cross-resistant with BSO ( Figure S3 ) . Virus titers and the plaque phenotype of the different resistant viruses were similar to that of the wt virus , which might be indicative for a comparable viral fitness , at least in vitro ( Fig . 5 ) . Only the T77M/N212S mutant produced smaller plaques than wt virus , which may suggest a role for the additional K115R mutation in VP3 that was not reintroduced into the genome of the T77M/N212S mutant . The capsid of a mature virion consists of 60 protomers , each protomer consisting of a single copy of structural proteins VP1 , VP2 , VP3 and VP4 , arranged on a pseudo T = 3 icosahedral surface [34] . Viral capsid proteins VP1 , VP2 and VP3 are partially located at the outer surface of the capsid and adopt a similar eight-stranded antiparallel β-barrel fold which is conserved among enteroviruses . The antiparallel β-sheet barrel of VP1 harboring Thr-77 ( the main mutation conferring GSH insensitivity that was identified in the revertant screen ( Table 2 ) , consists of βC , βH , βE and βF , and βB , βI , βD and βG which are packed face-to-face [35] . The Thr-77 is predicted to be a solvent exposed residue and is positioned in the βB , flanking the rim of the canyon closest to the fivefold axes [34] . Interestingly , the VP1 ( Val-150 and Asn-212 ) and VP3 residues ( Ala-180 ) are , akin to the Thr-77 residue , exposed on the surface of the CVB3 [M strain ( 1COV ) ] capsid and line the canyon surrounding the fivefold axis point ( Fig . 6A ) . Although dispersed throughout the viral genome in the linear amino acid sequence , the identified amino acids all reside in close proximity to each other on the surface wall and more specifically opposing each other at the interface between protomeric subunits . Next , we wondered whether the lack of activity of TP219 against several enteroviruses , including enterovirus 71 and echovirus 11 ( Table 1 ) , resulted from a natural resistance to GSH-depletion and whether this could be linked to a structural functionality . Within this context we analyzed the available crystallographic structures for the presence of a solvent exposed methionine , the main mutation conferring GSH-independence in CVB3 escape mutants . Rhinovirus 14 was also included in the analysis since this virus proved to be sensitive to GSH depletion in BSO-treated HeLa cells ( data not shown ) . The naturally resistant viruses were found to be already carrying a solvent exposed methionine at the interface between two protomers ( Fig . 6B ) . This methionine was not present in the GSH-dependent viruses ( including echo1 , CVA21 and RV14 ) ( Fig . 6C ) . The assumed GSH-independence of echo11 and EV71 could be mapped to respectively a methionine at location 76 ( positioned on the βB sheet ) and 229 ( positioned on the βH sheet ) in VP1 . The assembly defect following GSH depletion likely results from the requirement for a direct interaction between GSH and the capsid . It has been described previously that capsid-binding molecules ( e . g . pleconaril [36] ) can protect against thermal-mediated inactivation by direct binding to and stabilizing of the virus particle [37] . We therefore incubated wt , T77M and T77M/A180T CVB3 at 46°C , a temperature detrimental for wt CVB3 infectivity , in the absence or presence of increasing concentrations of GSH . Wildtype CVB3 displays a temperature-sensitive phenotype , which is reversed upon addition of GSH ( Fig . 7A ) . Interestingly , the infectivity of the T77M and T77M/A180T CVB3 remains almost unaffected at this temperature . These data strengthens the hypothesis that GSH interacts directly with the viral capsid thereby rescuing virus infectivity during heat inactivation . Thus , introducing a methionine at position 77 in VP1 generates a GSH-independent and temperature-insensitive phenotype . To provide biochemical evidence for the interaction between GSH and the capsid , GSH-pull down assays were carried out ( Fig . 7B ) . BGM cell cultures were infected with wt CVB3 after which the lysates were co-incubated with glutathione-sepharose beads and assayed for GSH affinity . Autoradiographic analysis clearly reveals that GSH is able to pull down VP0 , VP1 , VP2 and VP3 ( lane 2 and 2′ ) . Under these conditions of the gel , VP4 cannot be detected . Interestingly , the pull down mix contains both VP0 and VP2 which is indicative for the presence of both mature and precursor virus interacting with GSH . To further corroborate this observation and to assess the GSH-binding properties of a GSH-independent enterovirus , Vero cells were infected with PV Mahoney , being a GSH-dependent virus ( observed by Ma et al . in the accompanying paper ) and PV Sabin 1 , being a natural GSH-independent virus ( Table 1 ) . In addition , we assessed the effect of TP219-induced GSH depletion on this GSH-binding ( Fig . 7C ) . As expected , no maturation cleavage of VP0 into VP2 and VP4 could be observed in TP219-treated cells infected with PV Mahoney ( compare lanes 1 and 2 ) , indicating that no infectious particles were formed , whereas this cleavage was readily observed for the GSH-independent PV Sabin 1 ( compare lanes 3 and 4 ) . Akin to CVB3 , capsid proteins could be readily detected in the pull down of untreated cell lysates infected with both PV Mahoney ( line 5 ) and PV1 Sabin 1 ( line 7 ) . However , upon TP219-induced GSH depletion capsid proteins of PV Mahoney , but not those of PV Sabin 1 , lost their GSH-binding properties ( compare lanes 6 and 8 ) . Together , these observations suggest that GSH interacts with enterovirus capsid proteins and that following GSH depletion , GSH-dependent viruses , unlike GSH-independent viruses , lose their GSH-binding properties . It has been shown that ( i ) 5S and 14S particles associate with vesicular membranes containing the RNA replication complexes [38] and that ( ii ) encapsidation of progeny viral RNA is governed by a direct interaction between VP3 ( possibly as part of these 5S or 14S particles ) and replication protein 2C [20] , [21] . More specifically a role for residue Glu-180 in VP3 ( located at the interface of two protomers ) was demonstrated to be an essential contact point with protein 2C . Intriguingly , the genetic evidence described above also implicates this region in GSH-independence . Therefore the impact of GSH depletion on the subcellular ( co- ) localization of VP1 and protein 2C in cells infected with either wt , T77M and T77M/A180T CVB3 was studied . Both in the absence and the presence of GSH , protein 2C of both wt and mutant CVB3 showed a distinct reticulated pattern in the perinuclear region ( Fig . 8 ) , suggesting that GSH depletion had no effect on the localization of protein 2C as part of the replication complex . This is also in line with the above observation that depletion of GSH levels does not inhibit viral RNA replication . In the absence of TP219 ( i . e . in the presence of GSH ) , VP1 and 2C of both wt and the T77M CVB3 were observed to colocalize . Similar results were observed for T77M/A180T ( Fig . S4 ) . However , in the presence of TP219 ( i . e . in the absence of GSH ) , VP1 of wt CVB3 showed a more dispersed pattern and no longer co-localized with protein 2C . Hence , GSH might be essential for localizing VP1 , presumably as part of a 5S or 14S assembly intermediate , to the sites of replication allowing the encapsidation of newly synthesized viral RNA . Importantly , introducing a surface-exposed methionine in the capsid ( T77M ) resulted in a relocalization of VP1 to the replication sites even in the absence of GSH .
We here describe the discovery of TP219 as a novel inhibitor of the replication of several enteroviruses . We showed that TP219 forms a covalent adduct with GSH in an enzyme-independent way , thereby rapidly depleting intracellular GSH levels . As a consequence , the formation of infectious progeny virus is prevented without interfering with earlier events in the viral life cycle , such as entry , translation , proteolytic processing and RNA replication . By employing TP219 , we further studied the details of virus assembly and present several lines of evidence that GSH is required during the transition of protomeric particles into pentameric particles . Inhibition of virus assembly following GSH depletion was not rescued by treatment with other reducing agents , indicating that a reducing environment is not a requirement for morphogenesis . We provided biochemical and microscopical evidence for a direct interaction between VP1 and GSH and that this interaction is essential for the translocation of assembly intermediates to the sites of replication . Furthermore , genetic and structural analysis of several GSH-independent viruses point towards the interface between protomers as a region critical for GSH binding/independence . Capsid proteins VP1 , VP2 and VP3 have the conserved β-sandwich motif consisting of eight-stranded antiparallel β-sheets packed face-to-face [35] . It has been previously suggested that modulation of the interface between protomers might be a mechanism shared by enteroviruses for the control of structural transitions during morphogenesis [39] . Within this context , it may be plausible that intact β-sheets at the protomeric interface are a prerequisite for morphogenesis and that ( a ) residue ( s ) within the β-barrel region interact with GSH , thereby modulating or stabilizing energetically favorable interactions that promote virus assembly . Indeed , fractionation studies revealed that in the absence of GSH the formation of pentameric particles was prevented without affecting the levels of protomers . These data suggest that GSH is needed for the transition into or the stabilization of pentameric particles required for assembly into higher-order particles . Upon sequential passaging of ( GSH-dependent ) CVB3 in the presence of suboptimal levels of GSH , three GSH-independent variants were obtained . These virus isolates all contained a surface-exposed T77M mutation positioned in the βB-sheet of VP1 at the protomeric interface . Introducing this methionine into the wildtype CVB3 genome fully rescued the production of progeny virus , indicating that a single residue can mimic the presence of GSH . In addition , our antiviral data showed that some viruses , e . g . EV71 and echovirus 11 , proved to be naturally insensitive to GSH-depletion . Genetic analysis of the available crystallographic structures of these GSH-insensitive viruses revealed the presence of a similar solvent-exposed methionine at the interface between two protomers . This further supports the hypothesis that GSH ( or a surface-exposed methionine ) may promote stabilization of β-sheets at the interface between protomers allowing further assembly into pentamers . Taking into account that GSH is an essential component of the cellular metabolism and is present at very high concentrations in mammalian cells , the natural presence of a surface exposed methionine in the capsid of several enteroviruses suggests that these viruses may already have evolved to assemble in the absence of glutathione despite the lack of a selective pressure ( thus the absence of GSH ) . Whether the presence of a surface methionine represents a beneficial mutation conferring fitness advantage as also the question why most viruses do not already carry this residue remains to be answered . Since addition of other sulfhydryl reducing agents such as DTT and NAC could not substitute for GSH , it seems unlikely that the apparent requirement for GSH during virus assembly results from an imbalance of the intracellular redox status or from an indirect effect on local pH which in turn has been reported to be important during assembly of bovine enterovirus in vitro [40] . This could indicate that GSH is directly rather than indirectly involved in the formation of pentameric particles . This was confirmed by employing a biochemical assay in which we demonstrated that GSH directly interacted with capsid proteins VP1 , VP2 , VP3 and VP0 of CVB3 , PV ( M ) and PV Sabin 1 . Further evidence for this direct interaction between the capsid and GSH was provided by heat inactivation experiments . The temperature sensitive ( ts ) phenotype of wt CVB3 is suppressed upon addition of GSH or changed to a temperature insensitive ( non-ts ) phenotype upon introducing a methionine at Thr-77 . The fact that GSH ( and methionine ) is ( are ) capable of stabilizing virus infectivity following heat inactivation is in line with previous observations for capsid binding agents that , by direct binding to the virion , are also capable of rescuing infectivity following thermally induced conformational rearrangements [37] , [41] . We therefore propose a critical role for GSH as a direct binding and stabilizing cofactor in the interaction between protomeric units assembling into pentameric particles . This hypothesis was further corroborated by biochemical assays showing a loss of GSH-binding properties of PV Mahoney capsid proteins following TP219 treatment and suggests that GSH depletion results in malformed protomeric particles that are unable to interact with GSH and also unable to fully assemble into pentameric particles . Remarkably , GSH-independent enteroviruses retain their GSH-affinity albeit this affinity is no longer required for assembly . Likewise , it has been shown that the temperature sensitivity of PV3/Sabin could be attributed to a S91F substitution in VP3 at the interface between protomers which resulted in the impaired formation of 14S pentameric particles at supraoptimal temperatures [39] , [42] . Since non-ts revertants contained a number of second-site mutations , GSH or a surface-exposed methionine could in a similar way suppress conformational changes induced by heat inactivation . It was recently shown that encapsidation of progeny viral RNA is directly linked to an interaction between VP3 , ( possibly as part of a pentameric particle ) , and non-structural protein 2C [20] . More specifically , Glu-180 in VP3 at the interface of 2 protomers was suggested to be an essential determinant for interaction with protein 2C . Further evidence for a direct interaction of 2C with capsid proteins was provided by a study in which several mutations in VP1 and VP3 were identified to compensate for a defect in encapsidation [22] . Despite the fact that 2C and capsid precursors ( protomeric and pentameric particles ) have been shown to co-localize at the surface of replication vesicles and can be cross-linked to viral RNA , it is still unknown whether 2C interacts directly with VP1 or VP3 or with a higher-order assembly intermediate [38] , [43] . Our fractionation studies revealed almost exclusively protomers in TP219-treated cells . Strikingly , we observed an altered subcellular co-localization pattern of 2C and capsid protein VP1 ( presumably as part of a protomeric particle ) in TP219-treated cells by confocal analysis . On the contrary , the presence of a surface-exposed methionine resulted in a relocalization of VP1 to the sites of replication independent from GSH . Thus , GSH ( or its “surrogate” T77M ) proves essential for localizing assembly intermediates to the sites of replication . It is tempting to speculate that GSH is required for transition of protomers into pentamers and that pentamers are a prerequisite for a functional interaction with protein 2C . However , it cannot be formally ruled out that GSH depletion might also result in misfolded protomeric particles , resulting in a loss in the capacity to interact with 2C . Intriguingly , GSH depletion following BSO-treatment in CVB3-infected HeLa cells was earlier shown not to affect the levels of protomeric and pentameric particles , but increased the levels of empty capsid particles [25] . In addition , Ma et al . observed an accumulation of pentameric particles and slightly faster sedimenting 150S particles in BSO-treated PV ( M ) -infected HeLa cell cultures [accompanying paper] . By contrast , our fractionation studies revealed the absence of pentameric and empty capsid particles in both BSO-treated and TP219-treated cells . It remains to be studied whether the use of strain CVB3/0 or PV1 ( M ) and HeLa cells instead of a Nancy CVB3 and BGM cells ( our study ) explains this difference . Our data suggest that formation of infectious progeny involves the recruitment of assembly intermediates to a membranous surface containing the RNA replication complex . It has been demonstrated that the N-terminal glycine of VP4 is cotranslationally modified with a myristic acid linking each protomer within the pentameric structure to the two adjacent protomers at the 5-fold axes thereby stabilizing the structural elements within a pentameric unit [15] . Despite the fact that non-myristoylated protomeric particles can be found at the site of replication [13] and that purified non-myristoylated capsid proteins can assemble in solution into virus like particles [44] , it has also been demonstrated that myristoylation might be a determinant for a correct membrane targeting of cytoplasmatically processed protomers [45] . However , since a N-myristoyl group confers only low hydrophobicity , additional ( hydrophobic ) factors may be required to improve membrane affinity [46] , [47] . Within this perspective , we hypothesize that the tethering of myristoylated-protomers to membranous replication vesicles is enhanced by GSH or the presence of a methionine . Indeed , it has been recently suggested that glutathionylation of activated S100A9 , a calcium-binding protein involved in fatty acid ( FA ) transport in neutrophils , might facilitate the interaction of S100A9 with the lipid environment during the translocation from the cytosol to the membrane , probably by increasing protein hydrophobicity [48] . Similarly , the presence of a surface exposed methionine , one of the most hydrophobic amino acids , has been described to be a key regulator for tight phospholipid membrane binding [49]–[51] . Albeit based on much speculation , methionine or GSH could then generate a hydrophobic surface region that increases membrane affinity following N-myristoyl-guided membrane targeting . Translocation of protomers to membranes may increase the change to encounter other protomers to form pentamers and , hence , the efficiency of assembling into higher order particles . It remains unclear how GSH interacts with capsid proteins . Our data suggest that GSH is capable of interacting with capsid proteins both during and after maturation into infectious virus particles . This idea is supported by fractionation studies on sucrose gradients ( showing that GSH is involved in the transition of 5S into 14S ) and by heat inactivation ( showing that GSH interacts with the infectious virion ) and pull down experiments ( demonstrating the presence of both VP0 and VP2 following GSH-pull down ) . In addition , in the accompanying paper Ma et al . demonstrate that capsid proteins can be readily detected following GSH-pull down of purified assembly intermediates . Despite the fact that these data suggest that GSH can interact with the maturing particle , GSH has not been observed in crystallographic structures for which several reasons could be hypothesized: the purification process might have resulted in the loss of GSH or GSH might be displaced by an unknown mechanism during/after morphogenesis . It might also be worth to explore in more detail whether unexplained electron densities exist in particles that may explain the presence of GSH . This might also be interpreted to mean that GSH interacts with capsid proteins probably in a non-covalent way which is corroborated by the fact that capsid proteins can be detected following pull down experiments using GSH-sepharose beads . In this system , the GSH ligand is attached to sepharose by a coupling via its sulfhydryl-group to the linker , implying that viral capsids show affinity for GSH without the apparent involvement of a SH-group . This also implies that ( i ) GSH does not form disulfide bonds with free cysteines in the viral capsid and ( ii ) opens the possibility that this free SH-group might be the determinant for a correct assembly process . Indeed , genetic analysis of GSH-independent viruses resulted in the identification of methionine – another sulfur containing amino acid - as the main determinant for GSH-independence . Further research is warranted to provide more details on this interaction and to clarify why GSH has not been observed thus far in the atomic structures of purified virus particles . Reducing agents , including GSH and NAC have been frequently described to act as an non-conventional antivirals inhibiting the replication of several viruses , including HSV-1 , HIV-1 , influenza , and Sendai virus via various mechanisms [28] . This is the first paper , to our knowledge , that describes in detail the dependency of virus replication on GSH . Glutathione is the most potent intracellular reductant in the cell and is implicated in various cellular processes , including signal transduction pathways , gene expression , cell proliferation or cell death [27] . Consequently , an imbalance in GSH has been associated with various pathologies , including neurodegenerative disorders and cystic fibrosis . In recent years a dysregulated GSH system has also been associated with the development of tumor chemoresistance , mostly by GST-mediated GSH conjugation to chemotherapeutic agents , resulting in less toxic GSH-drug complexes that are readily exported from the cell . It is expected that understanding these mechanisms will not only increase therapeutic response but will also decrease drug resistance [52] , [53] . Within this context , TP219 could be a useful tool to study the direct impact of GSH depletion ( by omitting long incubation periods as for BSO ) on chemosensitivity in resistant tumors as well as the role for GSH in other processes such as virus replication .
BGM , HeLa , Vero , RD and MRC-5 cells were maintained in MEM ( Gibco ) , supplemented with 10% fetal bovine serum ( FBS ) ( Integro ) , 1% bicarbonate ( Gibco ) and 1% L-glutamine ( Gibco ) . The synthesis of TP219 is described elsewhere [29] . Enviroxime was synthesized by Dr . G . Pürstinger ( Institut für Pharmazie , Universität Innsbruck , Austria ) . Geldanamycin was from Biovision ( Milpitas ) . L-Buthionine sulfoximine ( BSO ) , guanidine hydrochloride ( GuaHCl ) , N-acetyl cysteine ( NAC ) , glutathione ( GSH ) , glutathione ethyl ester ( GEE ) and dithiothreitol ( DTT ) were purchased from Sigma Aldrich . Coxsackievirus B3 ( CVB3 ) was obtained by transfecting in vitro-transcribed RNA derived from plasmid p53CB3/T7 as previously described [54] . This plasmid contains a full length cDNA of CBV3 strain Nancy behind a T7 RNA polymerase promotor [55] . CVB3 expressing Renilla luciferase ( RLuc-CVB3 ) was obtained by introducing the Renilla luciferase coding sequence between the 5′ untranslated region and the P1 coding region . The luciferase protein is followed by a 3CD cleavage site allowing posttranslational cleavage from the polyprotein by 3CDpro . Poliovirus Sabin 1 and 3 strains were from Prof . B . Rombaut ( Vrije Universiteit Brussels , Belgium ) . Enterovirus 71 ( BrCr ) , CVA16 ( G-10 ) and CVA21 ( Coe ) were obtained from the National Institute for Public Health and Environment ( RIVM , the Netherlands ) . CVA9 ( Bozek ) , echovirus 9 ( Hill ) and echovirus 11 ( Gregory ) kindly provided by Dr . K . Andries at Janssen Pharmaceutica ( Beerse , Belgium ) . For immunoprecipitation of sucrose gradients we used pooled rabbit polyclonal anti-enterovirus antibodies ( including CVB1 , 2 and 3 ) ( Accurate Chemical and Scientific Corporation ) . For immunofluorescence or immunoblotting , we used as primary antibodies a mouse monoclonal anti-CVB3 VP1 ( Dako ) and a rabbit polyclonal anti-CVB3 2C ( kindly provided by Prof . L . Whitton , Scripps Research Institute , USA ) . The secondary antibodies included Alexa Fluor 488-conjugated goat anti-mouse or 568-conjugated donkey anti-rabbit for immunofluorescence and goat-anti-mouse-IRDye680 for immunoblotting . For single cycle virus infections , virus was added to subconfluent cell layers and allowed to adsorb for 1 hour , after which virus was removed and fresh ( drug-containing ) medium was added to the cells . At 8 h p . i . , cells were subjected to three cycles of freeze-thawing after which virus titers were determined by endpoint titration . Alternatively , cells were lysed to determine the intracellular Renilla luciferase activity with the Renilla Luciferase Assay System ( Promega ) . For multicycle virus infections , cells grown to confluency in 96-well plates were subjected to serial dilutions of the compound and inoculated with the appropriate virus . After 3 days of incubation , cell viability was measured: after removal of the medium , 10% MTS/PMS ( Promega ) was added to each well and quantified spectrophotometrically at 498 nm in a microplate reader . The 50% effective concentration ( EC50 ) was defined as the concentration of compound that inhibited virus-induced cytopathic effect formation by 50% and was calculated using logarithmic interpolation . In vivo pulse-chase metabolic labeling studies were performed as described previously [56] . Briefly , BGM cells , grown to confluency in a 24-well plate were infected with CVB3 at a multiplicity of infection ( MOI ) of 50 . At 5 h p . i . , the cells were starved for methionine by replacing the medium with methionine-free medium for 30 min . Subsequently , the cultures were pulse-labeled in methionine-free medium containing 35S-labeled methionine for 30 min in the absence or presence of TP219 ( 50 µM final concentration ) . At 6 hours p . i . , the cells were washed with PBS , lysed , and translation products were analyzed on a 12 , 5% polyacrylamide gel containing SDS , fixed , and then exposed to Kodak XAR film . BGM cells grown to confluency in 75 cm2 flasks were infected with CVB3 at an MOI of 10 . Following a one hour adsorption period , TP219 ( 400 µM ) was added in complete medium . BSO-treated ( 2 mM ) cells were pre-incubated for 48 hours . At 3 . 5 h post-infection , cells were starved for methionine by replacing the medium with methionine-free medium . Following a 30 minute incubation period , the cultures were pulse-labeled in methionine-free medium containing 35S-labeled methionine ( 20 µCi/ml ) . At 6 h . p . i . , the cells were washed three times with 1% of ice-cold unlabeled methionine . Following three cycles of freeze-thawing , equal volumes of supernatants were loaded onto 6 to 25% sucrose gradients ( wt/vol ) for a separation of the 5S and 14S subunits or onto 15 to 30% ( wt/vol ) for a separation of 75S and 150S subunits . Gradients were centrifuged at 39 , 000 rpm in a Beckman SW40 rotor for 16 h ( 6 to 25% gradients ) or 2 h 15 min ( 15 to 30% gradients ) . Gradients were fractionated in 400 µl aliquots from the top and precipitated using trichloroacetic acid ( TCA ) . The radioactivity present in each fraction was quantified by liquid scintillation counting . Immunoprecipitation studies were performed as previously described [57] . Briefly , every two consecutive samples of a sucrose gradient were pooled and immunoprecipitated using rabbit polyclonal anti-CVB3 antibodies . Following a 1 hour incubation period at 4°C , a 10% suspension of fixed Staphylococcus aureus strain Cowan I was added . Following 30 min incubation at 4°C , the samples were centrifuged for 5 min at 4°C . The immunoprecipitates were resuspended in NET buffer ( 150 mM NaCl , 5 mM EDTA , 50 mM Tris , 0 . 05% Triton X-100 , 0 . 1% bovine serum albumin , 0 . 2% methionin and 0 . 1% cysteine ) and centrifuged for 5 min at 4°C . The immunoprecipitates were resuspended in 2% SDS and boiled for 5 min . Finally , following centrifugation , the resulting supernatants were assayed for radioactivity by liquid scintillation counting . TCA-precipitated fractions of a sucrose gradient were subjected to SDS-PAGE , transferred to a PVDF membrane , incubated with primary antibody and secondary antibodies and scanned using an Odyssey Imager ( Li-COR ) . Densitometry was performed on immunoblot images using the ImageJ gel analysis tool . Cells grown to confluency in 96-well plates were incubated in the presence or the absence of 50 µM TP219 . At the indicated times post treatment with reduced glutathione ( GSH ) and oxidized glutathione ( GSSG ) levels were measured using the GSH/GSSG-Glo Assay ( Promega ) according to the recommendations of the manufacturer . TP219 ( 10 µM ) and glutathione ethyl ester ( 2 mM; pH 7 . 4 ) were co-incubated at 37°C . Samples were taken at four different time points ( 0 , 3 , 6 and 24 h ) and analyzed by HPLC . HPLC spectra were recorded on a Agilent 1120 compact LC instrument using a diode array detector ( 230 to 400 nm ) and an analytical ACE 5 C18-300 column ( 4 . 6 mm×15 cm ) at a flow rate of 1 mL/min . Solvents used were acetonitrile ( solvent A ) , and H2O ( 0 . 05% TFA ) ( solvent B ) . The gradient was used as follows: a 2 to 10% in 7 min ( solvent A ) followed by a 10 to 100% in 8 min ( solvent A ) . Mass spectrometry analysis was performed using a HPLC-Waters 12695 connected to a Waters Micromass ZQ spectrometer ( column: Sunfire C18 , 4 . 6×50 mm , 3 . 5 µm particle size ) . Solvents used were acetonitrile ( solvent A ) and H2O ( 0 . 1% formic acid ) ( solvent B ) . A 15 to 95% gradient ( solvent A ) was carried out in 5 min ( 1 mL/min ) . Electrospray ionization , positive ion mode; capillary voltage 3 . 5 kV , cone voltage 30 V . TP219 resistant ( GSH-independent ) CVB3 was generated by culturing CVB3 in the presence of increasing concentrations of TP219 . After 3 days of incubation , lysates from those cultures that exhibited CPE in the presence of the highest concentration of compound were collected and were used to infect new cell monolayers for successive rounds until viral replication was observed at concentrations that do not allow replication of wt virus . Subsequently , viral RNA was isolated ( Macherey-Nagel ) and both DNA strands were sequenced [cycle-sequencing method ( ABI Prism Big Dye Terminator Cycle Sequencing Ready Reaction Kit ) ] using an ABI 373 Automated Sequence Analyzer ( Applied Biosystems ) Mutant CVB3 clones were constructed , containing either single or multiple amino acid replacements at diverse positions in the capsid region . The eight clones were designated CVB3[T77M] , CVB3[V150I] , CVB3[N212S] , CVB3[K115R] , CVB3[A180T] , CVB3[T77M/V150I] , CVB3[T77M/N212S] , CVB3[T77M/A180T] . The following synthetic oligonucleotides ( and their complementary reverse oligonucleotides ) were used for site-directed mutagenesis of the single mutants: ( CVB3[T77M] ) 5′- A TGT AGG TCA GCA TGC GTG TAC TTT ATG GAG TAT AAA AAC TC -3′ , ( CVB3[V150I] ) 5′- GTA CCA CCA GGT GGA CCT ATA CCA GAT AAA GTT GAT T -3′ , ( CVB3[N212S] ) 5′- GC ATC AAC ACG CTA AAC AGC ATG GGC ACG CTA TAT G -3′ , ( CVB3[K115R] ) 5′- CA CAT TGG TCA GGC AGC ATA AGG CTT ACG TTT ATG TTC T -3′; ( CVB3[A180T] ) 5′- ACA CAC TAC CGG TTT GTT ACT TCA GAT GAG TAT ACC G -3′; The mutated sequences are underlined . Site-directed mutagenesis was performed with plasmid p53CB3/T7 using the XL Blue Large Site-directed mutagenesis kit ( Stratagene ) , according to the manufacturer's instructions . After mutagenesis , the individual clones were verified by sequencing . The double mutants CVB3[T77M/V150I] and CVB3[T77M/A180T] were generated by cloning the single mutant CVB3[V150I] and CVB3[A180T] into CVB3[T77M] using respectively enzymes EcoRI/SpeI and BglII/SpeI . The double mutant CVB3[T77M/N212S] was generated by introducing the mutation at position N212S into CVB3[T77M] using mutagenesis . Next , enzymes BglII/SpeI were used to isolate the fragment containing the desired mutations , and reintroduce in an original , non-mutagenized clone of the same plasmid p53CB3/T7 . From these mutants , RNA transcripts and infectious viruses were generated as described previously [54] . For determination of viral plaques , BGM cells , grown to confluency in six-well plates , were infected with the appropriate virus at 37°C . After 1 h , the virus was removed and the growth medium was replaced with agarose . Giemsa solution was used to stain the cells . Wildtype , T77M and T77M/A180T CVB3 were incubated with GSH for 15 min at room temperature and subjected to heat treatment by incubation for 30 min at 46°C either in the absence or presence of various concentrations of GSH after which virus titers were determined by endpoint titration . TP219-treated BGM or Vero cells were infected with CVB3 , PV Mahoney or PV Sabin 1 ( MOI 10 ) , labeled with 35S-Translabel 4–6 hr post-infection and harvested in TNM buffer ( 10 mM Tris pH 7 . 5 , 10 mM NaCl , 1 . 5 mM MgCl2 , 0 . 1% Tween ) . After freeze-thawing , the viral proteins were analyzed by SDS-PAGE or incubated with Glutathione sepharose 4B beads ( GE Healthcare life science ) overnight at 4°C . The GSH-beads were washed five times with TNM buffer and boiled in 1× SDS-PAGE sample buffer and analyzed by SDS-PAGE . Cells were grown to subconfluency in a 8-well chamber slide ( Lab-tek , II , Nunc , Germany ) and infected with the appropriate virus at an MOI of 10 . After 1 h , virus was removed and cells were treated with TP219 ( 50 µM ) . After 5 h , cells were fixed with 4% paraformaldehyde and permeabilized with PBS containing 0 . 5% Saponin . Subsequently , cells were stained with primary and fluorescently-labeled secondary antibodies . Nuclei were stained with DAPI . Cells were visualized using a confocal laser scanning microscope ( LCSM , Leica Microsystems , Germany ) .
|
Enteroviruses contain many significant human pathogens , including poliovirus , enterovirus 71 , coxsackieviruses and rhinoviruses . Most enterovirus infections subside mild or asymptomatically , but may also result in severe morbidity and mortality . Here , we report on the mechanism of antiviral action of a small molecule , TP219 , as an inhibitor of enterovirus morphogenesis . Morphogenesis represents an important stage at the end of the virus replication cycle and requires multiple steps , of which some are only poorly understood . Better understanding of this process holds much potential to facilitate the development of new therapies to combat enterovirus infections . We demonstrate that TP219 rapidly depletes intracellular glutathione ( GSH ) by covalently binding free GSH resulting in the inhibition of virus morphogenesis without affecting viral RNA replication . We discovered that GSH directly interacts with viral capsid precursors and mature virions and that this interaction is required for the formation of an assembly intermediate ( pentameric particles ) and consequently infectious progeny . Remarkably , enteroviruses that were capable of replicating in the absence of GSH contained a surface-exposed methionine at the protomeric interface . We propose that GSH is an essential and stabilizing host factor during morphogenesis and that this stabilization is a prerequisite for a functional encapsidation of progeny viral RNA .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"antimicrobials",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"emerging",
"infectious",
"diseases",
"host-pathogen",
"interactions",
"virology",
"emerging",
"viral",
"diseases",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"viral",
"structure",
"microbiology",
"antivirals",
"pathogenesis",
"viral",
"replication",
"viral",
"packaging"
] |
2014
|
Binding of Glutathione to Enterovirus Capsids Is Essential for Virion Morphogenesis
|
Neutrophils ( PMN ) play a central role in host defense against the neglected fungal infection paracoccidioidomycosis ( PCM ) , which is caused by the dimorphic fungus Paracoccidioides brasiliensis ( Pb ) . PCM is of major importance , especially in Latin America , and its treatment relies on the use of antifungal drugs . However , the course of treatment is lengthy , leading to side effects and even development of fungal resistance . The goal of the study was to use low-level laser therapy ( LLLT ) to stimulate PMN to fight Pb in vivo . Swiss mice with subcutaneous air pouches were inoculated with a virulent strain of Pb or fungal cell wall components ( Zymosan ) , and then received LLLT ( 780 nm; 50 mW; 12 . 5 J/cm2; 30 seconds per point , giving a total energy of 0 . 5 J per point ) on alternate days at two points on each hind leg . The aim was to reach the bone marrow in the femur with light . Non-irradiated animals were used as controls . The number and viability of the PMN that migrated to the inoculation site was assessed , as well as their ability to synthesize proteins , produce reactive oxygen species ( ROS ) and their fungicidal activity . The highly pure PMN populations obtained after 10 days of infection were also subsequently cultured in the presence of Pb for trials of protein production , evaluation of mitochondrial activity , ROS production and quantification of viable fungi growth . PMN from mice that received LLLT were more active metabolically , had higher fungicidal activity against Pb in vivo and also in vitro . The kinetics of neutrophil protein production also correlated with a more activated state . LLLT may be a safe and non-invasive approach to deal with PCM infection .
Paracoccidioides brasiliensis ( Pb ) is a non-sexual thermodimorphic fungus that exists in either a mycelium or a yeast form; the latter being pathogenic to humans and can cause an important and neglected systemic infection called paracoccidioidomycosis ( PCM ) . The likelihood of infection and its severity depends on the amount of inhaled fungi as well as the immunological status of the individual [1] . Patients with immune suppression or defects in immune cell activation are more susceptible to PCM [2 , 3] . PCM presents as a primary acute infection that is later transformed to a chronic phase . However , regardless of the stage of the disease , inflammatory cells play a central role in fighting Pb , particularly the neutrophils or polymorphonuclear cells ( PMN ) [4] . Besides the production of several direct antimicrobial factors , PMN may also secrete cytokines , chemokines and growth factors [5] that promote the host response against the infection . PMN are not only critical for the innate immune response , but can also help the adaptive immune response by interacting with B lymphocytes [6] , T cells [7] and dendritic cells [8] . Previous studies have reported prominent neutrophilic infiltrates in paracoccidioidomycotic lesions in experimental animal models such as hamsters [9 , 10] , rats [11] and also in tissue samples from patients [12] . Along with macrophages and plasmocytes , PMN are conspicuous in PCM granulomatous lesions and lead to altered morphology of the nearby fungal cells [13] . The immunological defense against fungi relies on the interaction between specific components of the fungal cell ( pattern-associated molecular patterns or PAMPs ) and pattern recognition receptors ( PRRs ) on host phagocytes . Through the binding of PAMPs to PRRs , a signaling cascade is initiated leading to release of pro- and anti-inflammatory cytokines linked to phagocytosis and intracellular fungal cell killing [14] . PMN can also help to eradicate pathogens via phagocytosis and the generation of reactive oxygen species ( ROS ) during the respiratory burst [15] . Nevertheless , despite the crucial role of inflammatory cells , they are usually not sufficient to entirely eliminate the Pb on their own , and patients usually need additional antifungal drug therapy [16] . Itraconazole , for instance , is effective in treatment of PCM , although its use may allow the relapse of the disease several months after discontinuation of the drug therapy [17] . Antifungal medication can also lead to diverse side-effects including dizziness , headaches , epigastric pain [16] and , more importantly , to the development of drug-resistance in the targeted microorganisms [18] . Therefore we asked whether there could be a novel way of activating PMN through the safe and non-invasive technique of low-level laser therapy ( LLLT ) . LLLT uses non-thermal and non-ionizing light irradiation that has been successfully used for acceleration of healing as well as reduction of pain and inflammation [19–21] . Although LLLT may often work as an anti-inflammatory modality [22] , it can , depending on the parameters , also trigger the activation of immune cells [23 , 24] and the activation of pro-inflammatory pathways [25] . While the activation of PMN by LLLT is not a completely novel process and has been reported in vitro [26 , 27] , the use of LLLT to help the organism to combat PCM is a new idea; thus , we aimed to assess the fungicidal capacity of PMN after LLLT by characterizing these cells on secretory protein levels , mitochondrial activity and ROS discharge following a first and second exposure to Pb .
This research was carried out in accordance with the ethical principles required for animal experimentation and was approved by the Ethics Committee on Animal Research of the Federal University of Alfenas , under the protocol registration No . 477/2012 . The animal procedures were conducted in accordance with the guidelines with animal care and use committee at Brazil`s National Council for the Control of Animal Experimentation . Swiss outbred female mice were kept in controlled temperature rooms and fed with sterile food and distilled water ad libitum . The animals were kept under a 12 light/12 dark cycle , and it was ensured that personnel did not enter the mouse facilities during the dark cycle . Isolates of the highly virulent Paracococcidioides brasiliensis Pb18 strain [28] were grown in semi-solid culture of Fava Netto [29] , with the culture medium replaced routinely every 7 days . A polysaccharide preparation known as Zymosan , derived from cell walls of the yeast Saccharomyces cerevisiae and containing β-D-glucan was commercially obtained ( Sigma-Aldrich , St . Louis , MO , USA ) . Pb18 cells or Zymosan were washed with sterile 0 . 9% saline solution and centrifuged ( 5810R Centrifuge , Eppendorf , NY , USA ) 3X at 1300g . A fungal suspension containing 5x107 yeast cells/ml was measured using a cell viability count after staining by the vital dye Janus Green B [30] and a hemocytometer . Zymosan particles were directly counted by hemocytometer . At 6 weeks of age and weighing approximately 25g , the animals received an “air pouch” as described by Harmsen and Havell in 1990 [31] and modified by Meloni-Bruneri et al . in 1996 [32] . An air pouch was produced in the dorsal region of mice by a subcutaneous injection of 2 ml of air; then , 0 . 1 ml of either the fungal suspension , Zymosan or saline was subsequently injected in the same region . It was previously shown by our group that P . brasiliensis elicits a marked neutrophil recruitment in vivo after air-pouch inoculation of the virulent Pb18 in mice; the mechanism behind this cell recruitment is probably due to chemotactic factors produced by the fungi and injured tissue [32] . In order to show that the PMN recruitment was truly invoked by the fungal cells or its derivatives and not by the air-pouch procedure itself , two additional groups were created and consisted of saline solution inoculation either followed or not by LLLT . The animals were divided into four groups , namely , group 1: animals infected with Pb18 and light irradiated; group 2: animals infected with Pb18 but not irradiated; group 3: animals inoculated with Zymosan and light irradiated; and group 4: animals inoculated with Zymosan and not irradiated . LLLT was performed on two points on each hind leg; the laser device used was a Twin flex laser ( MMO , São Carlos , SP—Brazil ) with a spot size of 0 . 04 cm2 . The laser parameters were: continuous wave near-infrared light ( 780nm ) to deliver 12 . 5 J/cm2 with a 50 mW total power; the total energy was 0 . 5 J per point ( 30 seconds per point ) . Our goal was to reach the bone marrow of the femoral bones , where the process of blood cell formation , known as hematopoiesis , including neutrophils is originated [33] . LLLT was performed on alternate days , with the animals first irradiated immediately after infection and last just before the neutrophil collection . In that way , the animals were irradiated on day 0 ( infection or inoculation ) ; day 2; day 4; day 6; day 8; and day 10 ( collection of PMN ) ; thus , 6 irradiations were performed . PMN were collected 10 days after the infection or the inoculation of the mice . The animals were anesthetized with a lethal dose ( 0 . 5 ml of a 10% ketamine hydrochloride and 2% Xylazine solution ) ; after a skin flap procedure was performed , the cells were collected and placed in sterile tubes with the help of a sterile glass Pasteur pipette and were subsequently dissociated by pipetting . The cells were then transferred and stored in Falcon tubes containing RPMI ( Sigma-Aldrich , St . Louis , MO , USA ) with 10% Fetal Bovine Serum ( FBS—Sigma ) and were kept refrigerated ( 2–6°C ) to be used for the subsequent experiments described below . The cells were quantified using a hemocytometer and the cell viability was assessed with 0 . 2% Trypan blue ( Sigma ) . For the fungal co-culture experiment with PMN , the refrigerated cells were centrifuged at 1780g and washed once before suspension in 15 ml of RPMI; then , the cells were quantified in a hemocytometer and viability was assessed with Trypan blue . The final concentration was adjusted to 106 PMN/ml . Pb cells were 3X washed with sterile 0 . 9% saline and centrifuged at 1300g and re-suspended in RPMI with 10% FBS . The concentration of the suspensions was adjusted according to the concentration of the obtained phagocytic cells , so that the cultures remained in a proportion of 1 Pb to 25 PMN to be further utilized for the evaluation of PMN metabolic activity , ROS quantification and quantification of viable fungi . Cells were counted in an hemocytometer and the Pb viability was determined by the staining with Janus Green B vital dye [30] . After adjusting the PMN suspension ( 106 PMN/ml ) , and the Pb fungal suspension ( 4x104 cells/ml ) to provide the co-cultivation mixture ( 1ml of each suspension ) , which was added to 12 well plates ( Corning , New York , USA ) , the plates were incubated at 5% CO2 and 37°C for 2 , 6 and 18 hours . After incubation , the cells were centrifuged at 1780g and the PMN pellets had their viability assessed by 0 . 2% Trypan Blue staining . In a 96 well plate ( Corning ) we added 100 μl of a 106 Pb18/ml suspension and 100 μl of a 5x106 PMN/ml suspension maintaining a ratio of 1:5 ( Pb:PMN ) . The experiment was performed in triplicate . After 2 hours of incubation ( 5% CO2 and 37°C ) we added 20 μl of MTT ( Sigma ) to the wells . The plate was further incubated for 4 hours . The supernatant was removed , leaving only the pellet at the bottom of each well . Then , 200μL of DMSO ( Sigma ) was added to each well and the plate was read in a microplate reader at 540nm ( Anthos Zenyth 200 , Biochrom , Cambridge , UK ) . The BCA method ( Sigma ) allows colorimetric detection and quantification of the total level of protein in a solution . This method combines the reduction of Cu2+ to Cu+ by protein in an alkaline medium ( the Biuret reaction ) with highly sensitive and selective colorimetric detection of the Cu+ ion using a reagent containing bicinchoninic acid [34] . The assays were performed in triplicate and the optical densities were measured in a microplate reader ( Biochrom ) at a wavelength of 560 nm . The results were expressed in mg of protein/ml , comparing the optical density with a standard curve containing known concentrations of bovine serum albumin ( BSA—Sigma ) . The calibration curve was made with a BSA solution of 10 μg/ml at 6 different protein concentrations: 10; 5; 2 . 5; 1 . 25; 0 . 67 and 0 . 33 μg/ml . The total protein concentration of each sample was calculated by pipetting 50μl of previously disrupted cells ( ultrasonic method ) along with 200μl of BCA . All samples were pipetted in triplicate and the results corresponded to the mean of the values obtained after blank ( RPMI medium ) subtraction for PMN cultured in vitro or co-cultivated with Pb18 . The quantification of reactive oxygen species produced by the PMN oxidative “burst” was carried out by the luminol chemiluminescence assay . PMN were obtained from the experimental groups and adjusted to a suspension of 1x106 PMN cells/ml; for the co-cultivation experiments PMN cells were adjusted to the proportion of 1 Pb to 25 PMN ( Pb concentration 4x104 cells/ml , PMN 1x106 cells/ml ) . Luminol ( Sigma ) was used as the substrate for this assay; 135 μl of the PMN suspension was added into a cuvette along with 30 μl of luminol; followed , for the co-cultivation experimental groups , by 135μl of the Pb18 suspension . A luminometer ( Glomax 20/20 Luminometer , Promega , USA ) was used to measure the chemiluminescence signal over 30 minutes . Positive ( PMA—phorbol myristate acetate , Santa-Cruz , Brazil ) and negative ( DPI—diphenyleneiodonium , Sigma ) controls were employed . The material collected from the subcutaneous air-pouches was immediately centrifuged at 1780g ( 5810R Centrifuge , Eppendorf , NY , USA ) . The pellets were re-suspended in 100μl PBS , and spread on Petri dishes with the aid of a sterile Drigalski spreader . Similarly , after centrifugation at 1780g , 100μl of PMN/Pb mixed suspensions obtained after 2 hours of co-cultivation were spread on Petri dishes . The experiments were performed in triplicate . The fungal growth on plates was allowed to take place over a period of 12 days , when a paintbrush marker was used to highlight the colonies . The culture medium used in this procedure was BHI agar ( HiMedia Laboratories , India ) supplemented with 1% glucose , 30% growth factor mixture produced by the fungus itself and 10% FBS , as described by Singer-Vermes et al . in 1992 [35] . The results were analyzed using the Shapiro-Wilk normality test and were all considered to have a normal distribution . Groups were compared using a Student`s T test with the level of significance set at 5% . The software used for the analyses was Graph-Pad Prism 6 ( GraphPad Software , Inc; La Jolla , CA 92037 , USA ) .
The animals inoculated with saline showed no neutrophils at the site of infection even after 10 days ( S1 Fig . ) , which clearly showed that neither the air-pouch procedure alone nor the laser irradiation alone was responsible for the PMN recruitment . The PMN produced by the inflammatory stimuli ( either Pb18 infection or Zymosan inoculation ) were harvested from the subcutaneous air-pouches ( Fig . 1 ) , and whilst the total number of PMN recruited to these air pouches was significantly diminished ( p = 0 . 0001 ) when LLLT was used after the Pb infection , the number of PMN was significantly increased ( p = 0 . 0001 ) when LLLT was used after mice were inoculated with Zymosan ( Fig . 2 ) . Interestingly , the kinetic study of PMN cell viability showed that LLLT was able to sustain a more viable population of neutrophils for the 18-hour time course both after Pb infection ( Fig . 3A ) at 6 hours ( p = 0 . 0278 ) and also after Zymosan inoculation ( Fig . 3B ) at 2 hours ( p = 0 . 0274 ) . There was no statistical significant difference between the viability of the PMN from irradiated or non-irradiated mice after co-cultivation with Pb for up to 18 hours , though the viability of the irradiated cells was kept at high levels , similarly to the non-irradiated PMN ( Figs . 3C and 3D ) . After being co-cultivated with Pb18 the PMN recruited by either the Pb infection or Zymosan inoculation showed a significantly higher mitochondrial activity ( p = 0 . 0029 and p = 0 . 0004 , respectively ) if they had been previously light irradiated in vivo ( Fig . 4 ) . In addition , the Zymosan irradiated group had a significantly higher mitochondrial activity ( p = 0 . 0012 ) than the Pb irradiated group , while the non-irradiated Zymosan group also had a significantly higher mitochondrial activity ( p = 0 . 0001 ) than the non-irradiated Pb group ( Fig . 4 ) . Protein production was significantly enhanced with LLLT at earlier evaluation periods when compared to the non-irradiated groups ( p = 0 . 0001 and p = 0 . 009 for Pb and Zymosan recruited PMN , respectively ) . The kinetics of protein production illustrates an intriguing crescent behavior for non-irradiated/Pb stimulated PMN ( p = 0 . 002 ) and an opposite decaying curve for the LLLT/Pb neutrophils ( p = 0 . 001 ) ; this decaying curve was also obtained with the highly activated PMN from the irradiated Z groups ( Fig . 5 ) . Likewise , after Pb co-cultivation , the kinetic production of proteins by irradiated PMN ( Pb and Zymosan recruited ) underwent decreasing curves that were distinct from the growing curves produced by the non-irradiated groups; this led to very distant values between the non-irradiated and the irradiated groups at 18 hours of co-cultivation ( p = 0 . 002 ) ( Fig . 5 ) . In summary , after 2 hours the Pb or Zymosan recruited PMN were significantly more metabolically active than their non-irradiated counterpart ( p = 0 . 0001 and p = 0 . 009 , respectively ) ; in addition , after 18 hours of co-culture the Pb-recruited PMN that did not receive LLLT were significantly more active than the Pb irradiated group ( p = 0 . 002 ) ; the Zymosan-recruited group also showed an initial disparity between irradiated and non-irradiated groups ( p = 0 . 0043 ) when co-cultivated with the Pb; this disparity was neutralized after 18 hours of the co-culture ( Fig . 5 ) . As seen in Fig . 6A , a significantly higher amount of ROS production , as measured by chemiluminescence , was seen with PMN from LLLT treated mice for both Pb and Zymosan groups ( p = 0 . 0425 and p = 0 . 0325 , respectively ) . In the co-cultivated groups , the light irradiated PMN consistently produced a significantly higher amount of ROS than their non-irradiated counterparts ( p = 0 . 0356 and p = 0 . 0325 for the Pb and Zymosan recruited PMN , respectively ) ( Fig . 6B ) . The non-irradiated Pb PMN also produced more ROS than the Zymosan non-irradiated PMN after co-cultivation ( p = 0 . 0406 ) ( Fig . 6B ) . LLLT treatment of mice was able to induce a higher fungicidal capacity in PMN cells , which was indirectly shown by a significantly lower number of Pb colonies growing from material isolated from the air pouches when evaluated after a 12-day growth period ( p = 0 . 0002 ) ( Fig . 7A ) . Moreover , LLLT was able to induce a significantly higher fungicidal capacity in PMN recruited by either Pb or Zymosan after 7 ( p = 0 . 0369 and p = 0 . 0232 , respectively ) ( Fig . 7B ) or even after 12 days ( p = 0 . 0193 and p = 0 . 0492 , respectively ) ( Fig . 7C ) of co-cultivation with Pb . Nevertheless , none of the groups was able to totally inhibit the growth of the fungi .
PMN activation through LLLT to the bone marrow led to a higher cell activity that correlated with two main effects: enhancement of innate immunity , shown by the higher yield of ROS and inhibition of Pb CFU in the lesion; and possible stimulation of acquired immune response shown by the increased yield of proteins seen in the LLLT groups . Finally , it is worth mentioning that although LLLT could be an effective and totally safe technique to activate fungicidal neutrophils , it was still not enough to eradicate the PCM; as previously stated , the phagocytic activity of PMN is considered not sufficient to entirely kill Pb [4] . Further study is warranted to see if different LLLT parameters , different sites of mouse irradiation or even distinct Pb infection routes could produce even better results from this promising technique .
|
PCM triggers a typical granulomatous inflammatory reaction with PMN playing a major role; these inflammatory cells are crucial in the initial stages of PCM , participating in the innate immune reaction and also directing the acquired immune response in the later stages . In some PCM patients , these immune mechanisms are insufficient to eradicate the infection , and need to be boosted with antifungal drugs that have to be administered for long periods and can show serious side-effects . We aimed to develop a novel and safe way to activate PMN through low-level laser irradiation of the bone marrow in the mouse femoral medulla . LLLT increased PMN viability and activation , shown by a significantly greater production of protein and ROS , as well as a higher fungicidal capacity; PMN even retained their higher metabolic activity and fungicidal ability after a second exposure to the pathogenic fungus in vitro . This is the first time that LLLT has been shown to increase the immune response against a fungal infection , and could be a promising and safe technique to be used with antifungal drugs in PCM .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Low-level Laser Therapy to the Mouse Femur Enhances the Fungicidal Response of Neutrophils against Paracoccidioides brasiliensis
|
Fifty years of residual insecticide spraying to control Triatoma infestans in the Gran Chaco region of northern Argentina , Paraguay and Bolivia shows that vertically coordinated interventions aiming at full coverage have limited effects and are unsustainable . We quantified the spatial distribution of T . infestans domestic infestation at the district level , identified environmental factors associated with high infestation and then explored the usefulness of risk maps for the spatial stratification of interventions . We performed spatial analyses of house infestation data collected by the National Chagas Service in Moreno Department , northern Argentina ( 1999–2002 ) . Clusters of high domestic infestation occurred in the southwestern extreme of the district . A multi-model selection approach showed that domestic infestation clustered in areas of low elevation , with few farmlands , high density of rural houses , high mean maximum land surface temperature , large NDVI , and high percentage of degraded and deforested lands . The best model classified 98 . 4% of the communities in the training dataset ( sensitivity , 93 . 3%; specificity , 95 . 4% ) . The risk map evidenced that the high-risk area only encompassed 16% of the district . By building a network-based transportation model we assessed the operational costs of spatially contiguous and spatially targeted interventions . Targeting clusters of high infestation would have reached ∼80% of all communities slated for full-coverage insecticide spraying , reducing in half the total time and economic cost incurred by a spatially contiguous strategy . In disperse rural areas where control programs can accomplish limited coverage , consideration of infestation hot spots can contribute to the design and execution of cost-effective interventions against Chagas disease vectors . If field validated , targeted vertical control in high risk areas and horizontal control in medium to low risk areas may provide both a logistically and economically feasible alternative to blanket vertical insecticide spraying when resources are limited .
Over the past 15 years , the burden of Chagas disease has been significantly reduced ( from an estimated ∼30 million human cases in 1990 to ∼9–11 million in 2006 ) as a consequence of direct actions promoted by several multinational regional initiatives [1] , [2] . The key for such success has been the long-term implementation of residual insecticide applications to kill domestic triatomine bugs , the screening of blood donors for the presence of the flagellate parasite Trypanosoma cruzi , and the treatment of infected infants born to infected mothers [2] . In the Southern Cone of South America , T . cruzi transmission by the main vector , Triatoma infestans , was interrupted in Uruguay , Chile , Brazil and parts of Argentina and Paraguay [1] , [2] . However , only limited success was attained in the Gran Chaco region of northern Argentina , Bolivia and Paraguay ( the core of T . infestans distribution ) where Chagas disease is still highly prevalent [3] . Vector control activities in this impoverished and mostly rural area are severely hampered by economic , logistic and political constraints , resulting in sporadic spraying of households with pyrethroid insecticides , with the consequent high bug reinfestation rates of rural communities [3]–[5] . Furthermore , the lack of effectiveness of pyrethroid insecticides in peridomestic habitats [6]–[8] coupled with the recent finding of T . infestans sylvatic populations [9]–[11] and the emergence of pyrethroid resistance in multiple localities [12]–[15] render the elimination of T . infestans from the Gran Chaco an elusive challenge . New approaches , tools , and methods for Chagas disease vector control and disease prevention are urgently needed . The distribution of T . infestans infestations and T . cruzi transmission are highly heterogeneous , with a few premises , households or communities accounting for a significant fraction of bugs and/or parasite transmission [16]–[20] . Knowledge of the location of households or communities that are at highest risk can be used to target vector control interventions for maximum effect ( i . e . , spatial targeting ) in comparison to blanket or untargeted interventions that may fail to reach some of the high-risk households or communities in an effective manner [21] , [22] . Spatially targeted interventions tailored for local contexts of parasite transmission risk , vector biology and insecticide resistance can constitute the backbone of an integrated Chagas disease vector management strategy in the Gran Chaco . Most knowledge on the factors affecting the spatial distribution of T . infestans has emerged from village-wide surveys performed at the premise/household level [4] , [19] , [23]–[28] . Given the large amounts of data and detail needed for mapping transmission risk , this fine spatial scale is not appropriate for planning and implementing large-scale vector control interventions . Rather , district-wide ( e . g . , state or county equivalents ) entomologic and epidemiologic information collected at the community-level may represent the best compromise between the level of empirical detail needed and potential applicability for the design of vector control interventions . Unfortunately , to date few studies have quantified the spatial distribution of T . infestans and the factors associated with the distribution of bug infestation at the district level [7] , [18] , and none have developed risk maps of T . infestans distribution and T . cruzi transmission at this scale . Extending from previous research on the cost-effectiveness of Chagas disease vector control interventions [5] , and as part of a larger project on the eco-epidemiology and control of Chagas disease in the Argentinean Chaco , the present study aimed to: a ) quantify the spatial distribution of domestic infestation by T . infestans at the district level in the absence of recent vector control actions; b ) identify environmental factors responsible for the heterogeneous occurrence of bug-infested communities; c ) develop risk maps of domestic infestation and , based on such information , d ) evaluate the relative costs of reaching and spraying communities predicted to be at high risk of domestic infestation by T . infestans .
The Moreno Department ( centroid at 62°26′W , 27°15′S ) , located in the Province of Santiago del Estero , northwestern Argentina , covers an area of 16 , 788 km2 ( Figure S1 ) . Moreno is located in the dry Chaco region , a semiarid plain with hardwood forest under intensive exploitation characterized by an average total rainfall of 549 mm and a rainy season from October to May [29] . In 2001 Moreno had approximately 25 , 000 habitants and 5 , 439 houses of which 2 , 911 ( 54% ) were rural houses located in 275 communities [30]; most ( 75% ) of the rural communities consisted of 1–10 houses ( Figure S1 ) . Rural houses usually have adobe walls and thatched roofs , one or two bedrooms , and a 5–10 m wide veranda in the front . The peridomestic environment includes structures that do not share a roof with the bedrooms , such as storerooms , chicken coops and corrals . In 2001 , almost 37% of all houses in the department had unmet basic needs ( an index that combines lack of adequate housing , tap water , crowding , and income ) [30] . Exploitation of forest resources ( hardwood for charcoal and logs , hunting ) , raising goats ( and cattle ) , and subsistence agriculture are the main sources of income of rural villagers . Historically , soybean and cotton have been cultivated in eastern Moreno but recent increases in rainfall had favored the introduction of genetically modified soybean , expanding the agricultural frontier towards the center of the Department . A complete database with information on the number of domiciles per community infested by T . infestans was generated from records provided by the National Chagas Service ( NCS , Argentine Ministry of Health ) in the year 2006 [5] . Bug infestation data originated from householders' notifications to community leaders and vector control personnel as part of the horizontal control program implemented in Argentina since 1993 [5] . Data included only anonimized records and was provided by NCS via a data sharing agreement with the University of Buenos Aires . The prevalence of infestation ( i . e . , infestation index ) in 25 communities during 1999–2001 was positively correlated with the infestation prevalence assessed by timed manual collections performed by NCS staff in the same communities in 2002 in domiciles ( Pearson correlation , r = 0 . 45 , P = 0 . 02 ) , but not in peridomiciles ( r = 20 . 14 , P = 0 . 4 ) . Because householders' reports likely underestimated peridomestic infestation ( with which they have much less contact [31] ) , such information will not be analyzed in this study . We also excluded from all our analyses the three main cities of Moreno ( totaling ∼2 , 528 houses ) because vector-borne transmission of T . cruzi in this area is considered to be mostly rural . Only information for the years 1999–2002 ( the period for which almost all Moreno communities had infestation data ) was used , since we aimed to reach maximum spatial coverage of all rural communities in the Department within a short time-frame . After the year 2002 , vector control and surveillance activities in Moreno have been scarce , making of the analyzed dataset the best and most current in terms of spatial coverage and quality of entomologic data . Vector control activities in the Department from 1997 to 1999 were negligible because of lack of insecticides and NCS personnel [5] . A base digital map of the Moreno Department including the location of rural villages , main waterways and road infrastructure at a 1∶250 , 000 scale was obtained from Instituto Geográfico Militar of Argentina . A georeferenced Landsat 7 ETM+ ( NASA; http://landsat . gsfc . nasa . gov/ ) satellite image ( spatial resolution 28 . 5×28 . 5 m ) from October 2002 and cadastral paper maps from Santiago del Estero Province were used to digitize communities that were not present in the base digital map using ArcGIS 10 ( ESRI , Redlands , CA ) . The final map was then projected in Universal Transverse Mercator ( UTM ) , Zone 20S , WGS1984 datum , and the distance matrix from all the digitized communities associated with the NCS entomologic database . Several data sources were a priori selected to calculate environmental parameters deemed important in explaining the spatial distribution of T . infestans in Moreno . The scale of analysis ( i . e . , individual communities ) prevented the use of census-derived socio-demographic data , which are available at a much coarser scale ( Moreno is divided in 10 census units aggregating data from rural communities and urban centers ) . Instead , we considered environmental variables describing different attributes of the landscape rural populations depend on for subsistence , such as land-use type , elevation and NDVI . Additionally , other environmental factors such as temperature ( impacting T . infestans population dynamics and pyrethroid insecticide effectiveness ) and density of rural houses ( delineating areas of high population density ) were considered . The raw pixel values of the Landsat ETM+ image were first converted to surface reflectance [32] and then classified using an unsupervised method into five classes describing the degree of landscape modification in Moreno: bare soil; croplands; deforested lands for cattle raising ( a savanna-like landscape with grasses and very few trees ) ; degraded forest ( natural forest in which tall trees were extracted and only scrubs prevail ) , and undisturbed forest ( Figure S1 ) . The overall accuracy of data classification ( kappa statistic = 0 . 82 ) was assessed from 46 ground control points digitized from a high resolution Ikonos image ( Space Imaging , Atlanta , GA ) from October 2002 using Erdas Imagine 9 . 0 ( Erdas , Norcross , GA ) . The Landsat ETM+ image was later used to estimate the Normalized Difference Vegetation Index ( NDVI ) by calculating the spectral difference between the red and infrared bands . Elevation ( in meters ) of all communities was derived from the Shuttle Radar Topography Mission ( SRTM; http://seamless . usgs . gov ) Digital Terrain Elevation data ( ∼90 m horizontal resolution and 1 m vertical resolution ) . Land surface temperature ( LST ) for the year 2002 was derived from Moderate Resolution Imaging Spectroradiometer ( MODIS; http://modis . gsfc . nasa . gov/ ) images ( ∼1 km spatial resolution and 1°C accuracy ) . A total of 36 images with 8-day LST averages were used to calculate the mean maximum LST ( in degree Celsius ) throughout the Moreno Department during the year 2002 . The remaining 10 images ( making up the 46-image dataset for 2002 ) were excluded from the analysis due to their high ( >20% ) cloud cover . NDVI and mean maximum LST were two of the most important climatic variables explaining the spatial distribution of T . infestans at the continental and district-wide levels [7] , [19] . A 2 km circular buffer area was created around the centroid of each rural community ( in Moreno , most rural houses are found within 2 km of the community center ) to characterize its environmental attributes by calculating the percentage of pixels belonging to each land-use class and the average NDVI , LST and elevation using ArcGIS 10 . A local spatial statistic ( Getis Gi* ( d ) , [33] , [34] ) was applied to identify the precise location of clusters or “hot spots” of high prevalence of domestic infestation by T . infestans . The area nearby the centroid of each rural community was searched at increasing distances ( d ) for occurrence of communities with higher prevalence of infestation values than expected by 999 Monte Carlo randomizations [33] , [34] . Communities were identified as members of positive or negative clusters when the Gi* ( d ) value at distance d was higher or lower than the Monte Carlo expectation , respectively . Significance was assessed at an alpha of 0 . 05 . To identify the distance up to which clustering was maximized , we plotted the absolute value of the sum of Gi* ( d ) ( Σ|Gi* ( d ) | ) over 1 km increments and identified the distance at which such value was the highest . We performed an edge effect correction of the Gi* ( d ) statistic by including prevalence of infestation values from communities located in neighboring Departments ( Figueroa , Ibarra and Alberdi ) up to a distance of 25 km from the Moreno border ( buffer area edge effect correction , [33] ) . For all analyses , the distance radius up to which clustering was evaluated was 40 km ( one-third of the shortest dimension of the department ) . We consider the Gi* ( d ) test to be more suitable for our analysis in comparison to Kulldorff Spatial Scan test because the latter tends to overestimate the contribution of isolated households/small communities to the overall pattern of T . infestans distribution ( due to the estimation of high log-likelihood values , indicative of clustering , encompassing isolated infested households ) . A random selection of 80% of the 220 communities with full entomologic , human demographic and environmental data was used as a training dataset to identify the factors associated with the prevalence of domestic infestation ( multiple linear regression ) or the membership in a cluster of high T . infestans domestic infestation ( multiple logistic regression ) . Inference was based on a multi-model selection approach [35] . Under this analytic framework a set of candidate models are contrasted with each other and the best model ( or small set of good models ) selected given the support received from the data [35] . We evaluated model fit using the Akaike information criterion ( AIC ) where the best model had the lowest AIC value and differed from the next best model by at least 2 units [35] . Models with AIC values within 2 units were considered equally good in predicting the data [35] . We further estimated the Akaike weight ( ωi ) for each model , describing the probability that a particular model was the best given the candidate set of models [35] . For each independent factor ( j ) evaluated we estimated its sum of Akaike weights ( Σωi ( j ) ) as the sum of the ωi of the models in which variable j was included [35] . This metric ( bounded between 0 and 1 ) allows determination of the relative importance of each independent variable in predicting the data [35] . Model fit was evaluated by comparing the model predictions with data from the remaining 20% communities ( test dataset ) . Since the predicted values of the logistic regression were represented by probabilities , we used the Youden index [36] to identify the optimal probability cut-off point to classify a community as either member or non-member of a cluster of high T . infestans domestic infestation . Briefly , the Youden index is the probability value at which the sum of sensitivity and specificity are maximum [36] . The product of the regression coefficients from the best model and their respective rasterized GIS layers was then integrated into the logistic function to generate a predictive map showing the probability of membership to a high T . infestans domestic infestation cluster . The input GIS layers and the scale of the risk map were set to a spatial scale of 1 km2 ( the coarser scale of the input data , belonging to the MODIS LST estimates ) . We used the results from the developed risk maps to explore the utility of novel analytical approaches for the design of spatially targeted interventions . Our premise was that risk maps could be used as predictive tools to identify the areas of high bug infestation where interventions could be targeted . First , we developed a scenario under which a blanket spraying campaign reached every community ( i . e . , following the rule of contiguity in which brigades visit the nearest neighbor of a treated community ) . In a second scenario , we used the developed risk map to guide the selection of areas of predicted high infestation clustering where interventions were targeted ( i . e . , a spatially targeted approach ) . We used the Vehicle Routing Problem ( VRP ) solver within the Network Analyst extension of the ArcGIS 10 . 0 software ( ESRI , Redlands , CA ) to calculate the total distance and time it will take two teams of two technicians each to reach those communities slated for full insecticide spraying ( i . e . , with domestic infestations higher than 10% ) under each scenario . The VRP algorithm finds the best ( i . e . , lowest cost ) route for a fleet of vehicles moving over a road network by implementing a multidimensional version of the Traveling Salesman Problem ( TSP ) . Given a list of communities and their pair-wise distances , a TSP algorithm tries to find a shortest possible tour that visits each community exactly once [37] . The VRP is a modification of the TSP algorithm because it uses the pair-wise matrix to assign community-to-depot trips , one at a time , to the most appropriate route ( the algorithm requires 2 or more vehicles to generate the routes ) . The initial solution is then improved upon by re-sequencing the trips on each route , as well as by providing mobility constraints such as priority areas or adding/removing vehicles . In the VRP , therefore , depots ( the home base of the spraying teams ) and communities can be visited more than once to optimize the spatial allocation of vehicles . Levy et al . [21] had recently applied the TSP algorithm to identify the ordering of districts within a city for the rational spraying of insecticides against T . infestans . Our analysis expands on such work by applying the VRP algorithm , by considering multiple vehicles and the explicit road network ( rather than districts ) , and by comparing two spatially-explicit vector-control strategies suitable for rural communities . Our model was spatially and temporally explicit , and had the following attributes: a ) due to resource constraints , only two trucks with two technicians each were available for spraying; b ) each truck was stationed in one of the two main cities of Moreno ( Quimili , Moreno's capital and Tintina , the district's second largest city ) ; c ) each team worked 8 hours a day and spent an average of 2 hours spraying a single rural household ( such time accounted for spraying and breaks ) ; d ) calculations considered travel time from each city to every community and from community to community; e ) if a community had more houses than the ones a team could spray on a single day , the model accounted for the overnight stay of the team in the village until spraying of the community was completed; e ) each team was scheduled to spray half the communities in the Department; f ) a community can only be sprayed once ( by the team that reaches it the earliest ) ; g ) mobility using paved roads was faster ( and preferred ) than on dirt roads; and h ) for the spatially-targeted strategy , communities with the highest risk of T . infestans infestation were prioritized for control at the beginning of the campaign . The model outputs included a per-community summary of the distance traveled to reach it and the total time spent spraying it , and a cumulative summary of the total distance covered and the duration of the spraying campaign . The model did not account for days lost due to climatic constraints ( rainfall or windy conditions , not favorable for insecticide spraying ) or vehicle malfunction . We considered the unit costs ( in US$ of 2009 ) of salaries , per-diem , insecticides and gasoline calculated for a vertical strategy by Vazquez-Prokopec et al . [5] for the Moreno Department to estimate the overall cost of each modeled scenario . Spatial analyses were performed using PPA ( Chen and Getis 1998 , San Diego State University , San Diego , CA ) and ClusterSeer ( TerraSeer , Ann Arbor MI ) software , whereas non-spatial analyses were performed with STATA 9 . 1 ( Stata Corp ) .
A total of 220 rural communities ( 80% of the total 275 rural communities ) presented entomological data during 1999–2002 ( Figure 1 ) . Most ( 93% ) communities without entomological data were small rural settlements with 1–4 houses . Overall , 29 . 7% of houses ( 857 of 2 , 885 houses ) were found infested with T . infestans . The average prevalence of infestation across communities was 34 . 9% ( SD = 33 . 5% ) . A total of 63 communities ( 28 . 6% ) were negative for domestic infestation , whereas 81 communities ( 36 . 8% ) reported domestic prevalence values of 50% or more . The average distance ( ± SD ) from the centroid of an infested community to the nearest infested community was 4 . 3±3 . 4 km ( Figure 1 ) . Communities located in the southwestern quadrant of the Department presented a significantly higher prevalence of domestic infestation by T . infestans than communities located anywhere in the remaining quadrants ( Mann-Whitney , U = 2 . 45; d . f . = 1; P<0 . 001 ) ( Figure 1 ) . Clustering of domestic infestation ( quantified as the absolute sum of Gi* ( d ) ) was maximized at a distance of 24 km ( Figure S2 ) . At this distance range , a total of 60 ( 27 . 3% ) communities belonged to a unique cluster of high domestic infestation , whereas 45 ( 20 . 5% ) communities belonged to a cluster of low domestic infestation ( Figure 2 ) . The cluster of high domestic infestation occurred in the southwestern extreme of Moreno , whereas the clusters of low domestic infestation in the center and east of the Department ( Figure 2 ) . Clustering of high domestic infestation extended over the neighboring Figueroa Department , proving that the edge effect correction was effective in accounting for spatial variation along the edges . The prevalence of domestic infestation inside the area of positive clustering ( median , Q1–Q3; 41% , 33–65% ) was two times higher than outside the clustering area ( 21% , 0–50% ) ( U = 1 . 346; d . f . = 1; P<0 . 001 ) . We evaluated the effects of various a-priori selected environmental factors on the prevalence of domestic infestation and the membership of a community in a cluster of high domestic infestation by T . infestans . Table 1 shows the results of the multi-model selection approach , ranking the top 10 linear regression models ( of 20 models evaluated ) from best to worst . The Akaike weight ( the probability that model i is the best among all tested models ) identified models 1–4 ( Δi<2 ) as the best supported by the data ( Table 1 ) . Model 1 was the most important in predicting the data , with an Akaike weight of 0 . 35 , followed by model 2 , with a weight of 0 . 21 ( Table 1 ) . The sum of Akaike weights ( a measure of the relative importance of each independent variable in predicting the data ) identified the density of rural houses , NDVI , elevation and the percentage of land covered by crops ( all with Σ ωi ( j ) >0 . 95 ) as the most important in explaining the prevalence of domestic infestation by T . infestans ( Table 1 ) . Model 1 had a pseudo-R2 of 0 . 19 . When the model predictions were compared with the test dataset , a significant deviation between model and data was observed at high ( >60% ) infestation values ( Figure S3 ) , indicating a poor fit to the data . The occurrence of many infested communities with very few ( 1–3 ) houses and high infestation prevalence ( but low overall contribution to the pattern of T . infestans distribution ) explains the model's poor fit at high infestation levels . The Akaike weight identified model 1 ( ωi = 0 . 68 ) as the best logistic model predicting the membership of a community in a cluster of high domestic infestation ( Table 2 ) . Elevation and percentage of landscape modified for soy production ( negatively ) and density of rural houses , mean maximum LST , NDVI , percentage of degraded and deforested lands ( positively ) were significantly associated with the membership of a community in a cluster of high T . infestans domestic infestation . The sum of Akaike weights identified distance to the nearest infested community as the least important factor in explaining membership in a cluster ( Σwi ( j ) = 0 . 25; Table 1 ) . The remaining variables all had Akaike weights higher than 0 . 9 , indicating their strong influence in predicting the data . Model 1 ( Table 3 ) classified correctly 98 . 4% of the communities in the training dataset , had a sensitivity of 93 . 3% , a specificity of 95 . 4% and a pseudo-R2 value of 0 . 72 . Based on the model's sensitivity and specificity , the Youden index identified predicted probabilities higher than 0 . 404 as belonging to a cluster of high domestic infestation . Based on such criteria , a total of 88 . 6% of the 44 communities selected as test dataset were correctly classified by the model ( Table S1 ) . The model's sensitivity and specificity in predicting the training data were 0 . 86 and 0 . 9 , respectively ( Table S1 ) . The regression coefficients of the best models and their respective rasterized GIS layers were used to generate maps describing the predicted prevalence of T . infestans domestic infestation ( Figure 3A ) and the probability of membership in a cluster of T . infestans infestation ( Figure 3B ) . Both maps outline the importance of environmental and demographic factors in defining the western extreme of the Department as a hot-spot ( high-risk area ) of bug infestation . Based on Figure 3B we estimated the high-risk area ( probabilities >0 . 404 , based on Youden index ) to encompass approximately 2 , 736 km2 , or 16% of the 16 , 788 km2 comprising the Moreno Department . We used the risk map to assess the operational costs of reaching high-risk communities by simulating two vector control scenarios ( Figure 4 and Table 4 ) . When a blanket insecticide campaign was simulated , the first communities to be reached ( black dots in Figure 4A ) were located in close proximity to the two cities , in areas of predicted low probability of T . infestans domestic infestation clustering . From such communities , the spraying teams continued moving towards locations of no domestic infestation clustering located in the east ( Figure 4A ) . It took the contiguous strategy 373 days and 2 , 063 km to reach and spray all communities ( Table 4 ) . The spatially targeted strategy , however , lasted for 190 days and covered 846 km ( Table 4 ) . More importantly , our analysis shows that ∼80% of all communities and households slated for spraying could be reached when targeting the areas predicted as high risk of domestic infestation clustering ( representing only ∼16% of the Department's surface ) ( Figure 4B ) . Focusing efforts on the high-risk areas did not increase the average costs of treating each community ( US$317 for blanket and US$312 for targeted interventions ) or household ( US$26 . 4 for blanket and US$25 . 9 for targeted interventions ) , but the reduced area and number of communities that need to be covered provides both a logistically and economically feasible alternative to blanket insecticide spraying , given limited resources .
In areas of low-to-moderate Chagas endemicity , where triatomine bug infestations tend to aggregate in certain households , communities or districts , targeting surveillance and control interventions in areas predicted at high risk can increase the coverage and cost-effectivenes of interventions [42] , [43] . A recent study has found that targeted control of Triatoma dimidiata could be significantly improved when environmental risk factors ( instead of the sole consideration of household factors such as building materials ) are incorporated into proper statistical models [43] . Under this low-transmission context , the success and sustainability of targeted approaches thus relies on thorough epidemiological investigations , proper data collection and analysis and constant assessment of potential changes in the eco-epidemiological dimensions of vector distribution and parasite transmission [42] , [43] . In highly endemic areas such as the Moreno Department and most of the Dry Chaco a true conflict between the need for high coverage of interventions and limited availability of resources exists . Most vertically structured control programs thus fail to reach all communities ( or districts ) at most risk . As an example , Santiago del Estero Province has recently invested significant resources to achieve full insecticide coverage but , given logistic limitations , excluded Moreno and other highly endemic Departments from the current priority areas . Targeting areas of presumed high risk alone while leaving others off the control scheme is not a feasible option . With the limited personnel and resources available , targeting vertical residual insecticide applications on Moreno communities predicted at high risk ( ∼80% of all infested communities slated for full spraying ) would have required half the time and personnel costs than a blanket strategy . Here , we propose that with the remaining resources and time , NCS can implement a community participatory approach [4] , [5] in areas of predicted medium-low infestation risk ( Figure 5 ) . In Moreno , the implementation of a horizontal attack phase ( i . e . , where local villagers were trained in the use and in charge of the application of residual insecticides ) showed significant impacts in reducing bug infestation and parasite transmission but lacked sustainability due to the limited oversight by NCS staff [5] . Implementing a mixed strategy as the one shown in Figure 5 may provide a reasonable tradeoff between campaign costs , insecticide coverage and NCS staff involvement in control activities . Targeted interventions are not a panacea . Relying on model predictions to infer where high risk areas occur may leave many infested communities excluded from the control scheme ( as observed in central Moreno ) . Missing or excluding communities could increase their risk of parasite transmission ( due to relaxation from insecticide pressure ) and contribute to community discomfort due to the perception that control programs do not care about them . Furthermore , bug reinfestation in communities located in the border of high risk areas may proceed faster due to a high likelihood of T . infestans flight or passive dispersal from untreated communities nearby [26] . Transportation models like the one developed in this study can help bridge some of these challenges only when they are based on well developed predictive models and accurate road network information . Only after thorough field validation spatially targeted approaches may be considered as a feasible operational alternative to blanket vertical control in districts like Moreno . Attaining the sustainable control of T . infestans and management of Chagas disease in highly endemic areas such as the economically deprived Gran Chaco not only will require the consideration of biological features of T . infestans and socio-economic attributes of the local population unique to this region , but also the incorporation of novel methods and approaches to help vector control programs design rational and more cost-effective interventions .
|
In the Southern Cone countries of South America , vectorborne transmission of Chagas disease persists in the Gran Chaco region of northern Argentina , Paraguay and Bolivia , where Triatoma infestans is the main vector . More than 50 years of vector control in this region demonstrate that vertically coordinated spraying campaigns aiming at full coverage are unsustainable in practice . Understaffed programmes are challenged by the number of households and vast extension of the endemic area , resulting in suboptimal insecticide coverage and limited and shortlived impacts on vector populations . Our rationale was to recognize this inherent limitation and provide a scientific base for the improvement of current interventions . We quantified the spatial distribution of T . infestans at the Department level and found that environmental and demographic factors can be used to predict the occurrence of areas with high risk of vector infestation that can then be targeted for control . In rural and dispersed areas such as the Gran Chaco , mapping surrogates of T . cruzi transmission risk and incorporating spatial heterogeneities in vector distribution into control and surveillance strategies can contribute to improved delivery of interventions and to the sustainable control and management of Chagas disease .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"public",
"health",
"and",
"epidemiology",
"geography",
"gis",
"zoology",
"entomology",
"epidemiology",
"earth",
"sciences",
"biology",
"population",
"biology",
"spatial",
"epidemiology",
"cartography"
] |
2012
|
Spatial Heterogeneity and Risk Maps of Community Infestation by Triatoma infestans in Rural Northwestern Argentina
|
Resistance to macrolide antibiotics is conferred by mutation of A2058 to G or methylation by Erm methyltransferases of the exocyclic N6 of A2058 ( E . coli numbering ) that forms the macrolide binding site in the 50S subunit of the ribosome . Ketolides such as telithromycin mitigate A2058G resistance yet remain susceptible to Erm-based resistance . Molecular details associated with macrolide resistance due to the A2058G mutation and methylation at N6 of A2058 by Erm methyltransferases were investigated using empirical force field-based simulations . To address the buried nature of the macrolide binding site , the number of waters within the pocket was allowed to fluctuate via the use of a Grand Canonical Monte Carlo ( GCMC ) methodology . The GCMC water insertion/deletion steps were alternated with Molecular Dynamics ( MD ) simulations to allow for relaxation of the entire system . From this GCMC/MD approach information on the interactions between telithromycin and the 50S ribosome was obtained . In the wild-type ( WT ) ribosome , the 2′-OH to A2058 N1 hydrogen bond samples short distances with a higher probability , while the effectiveness of telithromycin against the A2058G mutation is explained by a rearrangement of the hydrogen bonding pattern of the 2′-OH to 2058 that maintains the overall antibiotic-ribosome interactions . In both the WT and A2058G mutation there is significant flexibility in telithromycin's imidazole-pyridine side chain ( ARM ) , indicating that entropic effects contribute to the binding affinity . Methylated ribosomes show lower sampling of short 2′-OH to 2058 distances and also demonstrate enhanced G2057-A2058 stacking leading to disrupted A752-U2609 Watson-Crick ( WC ) interactions as well as hydrogen bonding between telithromycin's ARM and U2609 . This information will be of utility in the rational design of novel macrolide analogs with improved activity against methylated A2058 ribosomes .
Microbial resistance presents a major challenge in the development of novel antibiotics because bacteria are continually developing new resistance mechanisms [1] . The bacterial ribosome is a target for over 60% of antibiotics [2] , [3] , which bind at vital sites within both the 30S and 50S ribosomal subunits and inhibit processes that are essential for cell survival [4]–[10] . One important class of antibiotics is the macrolides , which bind at the beginning of the exit tunnel in the 50S subunit of the bacterial ribosome and block elongation of the nascent polypeptide chain [10]–[12] . For the macrolide class of antibiotics , bacteria achieve resistance by modifying or mutating bases within the binding pocket as well as by other mechanisms such as drug metabolism and overexpression of efflux pumps . Telithromycin is the first of a new class of macrolides , called the ketolides , named for the substitution of the C3 L-cladinose sugar with a ketone . Ketolides have largely addressed resistance due to drug efflux and metabolism yet still remain susceptible to ribosomal modification [13] , [14] , the most clinically relevant of which is modification of A2058 in 23S rRNA ( E . coli numbering throughout ) that confers cross-resistance to antibiotics in the macrolide , lincosamide and streptogramin B classes ( MLS ) [15]–[18] . A2058-based resistance includes mutation of A2058 to G as well as methylation of the exocyclic N6 of A2058 via the expression of erm genes encoding methyltransferases ( Erm methyltransferases ) that add one or two methyl groups and represent the most effective mechanism of macrolide resistance [19]–[21] . Telithromycin and other ketolides have been found to mitigate macrolide resistance due to the A2058G mutation ( Table S1 in Text S1 ) [22]–[24] , though ketolides that are widely effective against Erm methylation-based modifications are not known . Still , telithromycin has shown improved activity against monomethylated ribosomes compared to previous generation macrolides ( Table S1 in Text S1 ) [23] , [25] , which has been attributed to secondary contacts made between telithromycin's alkyl-aryl group and bases A752 and U2609 within the binding pocket [15] , [26]–[28] . Telithromycin's improved activity compared to erythromycin can also partly be explained by its absence of erm gene induction [29] . Erythromycin has been shown to increase levels of Erm methyltransferases in species carrying inducible erm genes [17] , [18] , [25] , [30]–[34] consequently increasing the levels of dimethylated ribosomes , whereas telithromycin and other ketolides have been found to mostly bypass the erm gene induction pathway [29] , [34] , [35] . In a comprehensive study looking at antibiotic activity as a function of mono-/dimethylated 2058 ribosome levels , it was found that in general increased monomethyl A2058 levels alone were not sufficient to mitigate telithromycin activity and that upon increasing the percentage of dimethylated ribosomes via induction with erythromycin was significant telithromycin susceptibility observed [25] . However , species-dependent exceptions were noted and therefore it appears that a combination of factors is involved in enhancing telithromycin's effectiveness in monomethylated ribosomes . Even so , dimethylated ribosomes exhibit the most potent resistance to telithromycin [25] as well as other antibiotics in the MLS class [18] and higher levels of dimethylated ribosomes confer the greatest resistance . Given the widespread transfer of resistant genes between species [36] , [37] it is important to develop analogs with improved affinity toward Erm-mediated resistant phenotypes . Crystal structures of telithromycin bound to E . coli [38] ( as well as T . thermophilus [39] , H . marismortui [15] , and D . radiodurans [10] ) have revealed several interactions that are important for telithromycin binding , as shown schematically in Figure 1 . Multiple van der Waals ( VDW ) contacts between hydrophobic alkyl groups on the macrolactone and residues forming the walls of the exit tunnel are present that contribute to binding . Telithromycin's desosamine 2′-hydroxyl group forms hydrogen bonds with N1 of A2058 , N6 of A2058 , and N6 of neighboring A2059 [10] . In addition , desosamine's 3′-dimethylamino moiety forms a salt bridge with the phosphate of G2505 [10] . And , in the E . coli [38] crystal structure the imidazole-pyridine side-chain ( ARM ) extending from the C11–C12 cyclic carbamate engages in stacking with A752 and U2609 , increasing telithromycin's binding affinity compared to earlier generation macrolides such as clarithromycin , roxithromycin , and azithromycin [15] , [26]–[28] . The present study addresses the molecular details associated with macrolide resistance due to the A2058G mutation and methylation at N6 of A2058 by Erm methyltransferases . To compare these interactions in the wild type and telithromycin-resistant strains , empirical force field-based simulations were performed on truncated versions of the wild type and mutant/modified E . coli 50S ribosomal subunits . The E . coli crystal structure [38] is selected because it is the first crystal structure available for a pathogenic bacterial species and hence most relevant in terms of drug design . To address the buried nature of the macrolide binding site , the number of waters within the pocket were allowed to fluctuate via the use of a combined Grand Canonical Monte Carlo ( GCMC ) /Molecular Dynamics ( MD ) methodology that builds on previous methodological developments [40]–[50] . GCMC/MD has been successfully applied to systems with deeply buried binding pockets [51] , [52] and has been applied to the ribosome to determine the binding free energy of sparsomycin [53] , [54] . It is utilized here as a means to assure proper solvation of the telithromycin binding site . The GCMC water insertion/deletion steps are alternated with MD simulations to allow for relaxation of the entire system . From this GCMC/MD approach information on the nature of the interactions between telithromycin and the 50S ribosome was obtained , yielding a detailed understanding of these therapeutically relevant antibiotic resistance mechanisms at the molecular level .
Altogether , our findings indicate that hydrogen bonding between 2′-OH and A2058 is important for telithromycin activity . Telithromycin maintains activity in the A2058G mutant via hydrogen bonds between the 2′OH and the WC hydrogen bonding groups of guanine , while decreased sampling of short 2′-OH to A2058 distances in the monomethylated species contributes to their lowered activity . In addition , the mutation/methylations are predicted to alter base stacking interactions with 2057 and to a lesser extent with 2059 . These perturbed stacking interactions are communicated to more remote regions of the ribosome that comprise the telithromycin binding pocket thereby contributing to changes in telithromycin's activity in the methylated species . These changes occur through three potential pathways identified in this study: the telithromycin pathway , the G2057 pathway and the A2059 pathway . Analysis indicates the G2057 pathway leads to the largest conformational changes , including alterations of interactions of nucleotides 2058 , 2610 and 2611 with methyl groups on telithromycin and , importantly , perturbation of A752-U2609 base pairing and interactions of those bases with the ARM of telithromycin . In the context of telithromycin activity in wild type and A2058-methylated ribosomes , studies suggest that high levels of A2058-dimethylated ribosomes are required to confer telithromycin resistance [25] . Minimum inhibitory concentration ( MIC ) values for telithromycin are only marginally increased ( ∼4-fold ) in monomethylated ribosomes as opposed to 256-fold for erythromycin [23] . The ability of telithromycin to maintain activity against monomethylated ribosomes has been proposed to result from imidazole-pyridine to A752-U2609 stacking interactions that mitigate the effects of disrupted 2′-OH – A2058 hydrogen bonds [15] , [26]–[28] and our findings coincide with this assessment . While hydrogen bonding does occur between telithromycin's imidazole and 2609 in MAD1 , the ARM and 752–2609 maintain distances and plane angles that are indicative of stacking in a large number of conformations and 752–2609 WC distances are sampled with a higher probability than non base-paired or non-stacked distances . Recently , Melman and Mankin [58] suggested that disruption of the 2′-OH – A2058 N1 hydrogen bond was not the major reason that telithromycin activity is reduced in A2058-dimethylated ribosomes . They found that removal of the 2′-OH from telithromycin did not increase MIC values to the extent that Erm ( A ) expression did . In other words , disruption of the hydrogen bond was not the predominant explanation for loss of telithromycin activity . They propose that the major explanation for reduced telithromycin activity is more likely that the overall structure within the binding site is perturbed upon the introduction of methyl groups onto A2058 as a result of nonbonded interactions between the methyl groups and nearby crystallographic waters . The exocyclic A2058 N6 is within 4 Å of water molecules that are coordinated to Mg+2 that chelates the phosphate groups on G2056 and G2057 . Accordingly , the authors suggest that interactions between the N6-methyl groups and water molecules lead to structural changes in the dimethylated ribosomes that reduce telithromycin activity . The results presented here further suggest that the conformation of RNA in the macrolide binding pocket is perturbed in both the 2058 mutant and modified ribosomes . However , the present results indicate that this is due to altered base stacking interactions with 2057 and 2059 that are propagated to other bases in the G2057 and A2059 pathways ( Figure 5 ) , respectively . To test the hypothesis of Melman and Mankin [58] , the distance between the exocyclic N6 of A2058 and the Mg+2 ion were compared in the WT and mutant/modified systems ( O6 in A2058G mutant ) . In all the systems except the mutant , the interaction is shifted by >1 Å indicating that the crystal structure distance is slightly too short ( Figure S12 in Text S1 ) . The shift to shorter distances for the A2058G mutant is expected given that the exocyclic N6 amine of A is replaced by a keto group in G . The distribution of distances show a high degree of overlap between WT and DMAD indicating that the relative position of the magnesium ion does not change upon introduction of two methyl groups onto A2058 . Moreover , the distribution for DMAD is sharper suggesting that dimethylation restricts the conformational sampling of the 2058 N6 to Mg2+ distance . A particularly interesting result is the flexibility of the ARM in WT and the A2058G mutant versus that occurring in the methylated species . The fluctuations of the ARM are significantly higher than the rest of telithromycin in the WT crystal structure , corresponding to an average RMS fluctuation of 1 . 55 Å at 100 K over the non-hydrogen atoms in the heterocycles in the ARM . Thus , while stacking of the ARM with A752-U2609 is occurring to some extent , this is clearly a dynamic region of the system . The flexibility of the ARM is consistent with kinetic studies of telithromycin binding to the E . coli ribosome [59] , which show that telithromycin's shift from a low to a high-affinity state results from reorganization of the ARM , and may also explain the various conformations of the alkyl-aryl ARM seen in different crystal structures [15] , [26] , [38] . Given that this flexibility is large in both the WT and A2058G mutant and significantly lower in the methylated species , it suggests a scenario where entropic contributions associated with the flexibility of the ARM , and possibly the surrounding environment , makes a favorable contribution to binding . Upon methylation the mobility of the ARM is decreased , thereby contributing to a decrease in binding affinity . However , this is due to increased hydrogen bonding between the ARM and the A752 and U2609 nucleotides , interactions that could be exploited to improve the binding affinity . Essentially , a favorable entropic contribution to binding is being switched to a potentially favorable enthalphic contribution . Indeed , this appears to be the case with the recently published analog , solithromycin , in which the ARM imidazole-pyridine moiety was replaced with a triazolyl-aminophenyl group allowing for additional hydrogen bonding with 752 as well as nearby 748 [60] . Namely , the hydrogen bonds lead to a favorable enthalpic contribution while decreasing the favorable entropic contribution which is consistent with the experimentally observed decrease in the RMS fluctuations from 1 . 5 to 1 . 0 Å for the ARM non-hydrogen atoms upon going from telithromycin to solithromycin . While solithromycin shows improved activity compared to telithromycin against Erm-based modifications , which may be attributed to this hydrogen bonding , the presence of a C2-fluoro group not present on telithromycin may also contribute to improved binding . The C2-fluoro group appears to form a favorable ( 2 . 7 Å ) , hydrophobic interaction with C2611 thereby complicating interpretation of the contribution of ARM hydrogen bonding to affinity . The results presented herein may be used to suggest modifications to telithromycin , or future ketolides , that could improve its binding to Erm-methylated ribosomes . These include modifying telithromycin's ARM to 1 ) engage in hydrogen bonding interactions with 752 , 2609 and adjacent nucleotides , leading to an enthalpic contribution to binding or , on the other hand , 2 ) decrease the potential for such hydrogen bonding and/or increase the conformational flexibility in the ARM , leading to an entropy gain . Another region for potential improvement of binding against the methylated species involve modifying the macrolactone of telithromycin . The present calculations indicate the telithromycin methyl-base interactions to be longer in the modified ribosome . Accordingly , modifications to telithromycin that add steric bulk to the methyl groups nearby 2058 , 2057 , and 2611 may enhance VDW interactions with the ribosome that may also gain from an increased hydrophobic contribution to binding . Such modifications will be guided by previous work showing that ketolides bearing C2 groups larger than F are devoid of activity [61] as well as the results with solithromycin [60] showing that C2-fluorination may increase ketolide activity against Erm-based modifications .
Calculations were performed with the program CHARMM , version C36a6 [62] and the CHARMM additive force field including the protein with the CMAP correction [63]–[65] , nucleic acid [66]–[69] , carbohydrate [70]–[75] , and CGenFF [76] parameters and the TIP3P water model [77] . Coordinates were obtained from the protein database ( PDB ID 3OAT [38] ) , with hydrogens added using the HBUILd facility in CHARMM . Since only the region around the telithromycin binding site was of interest , residues without one or more atoms within 40 Å of the center of system , as defined by the center of mass of telithromycin , were deleted . The system was then overlaid with a 28 Å water sphere and any waters bearing an oxygen within 2 . 8 Å of a solute non-hydrogen atom were deleted . Mutants were generated using this truncated system , with patches applied to A2058 in order to generate the A2058G , N6-monomethyl , and N6 , N′6-dimethyl A2058 variants . Initial guess parameters for the base methylations were obtained from ParamChem [78] , [79] . Bond lengths , angles , and dihedrals specific to the mutations were optimized via comparison to quantum mechanical ( QM ) geometries , water interactions , molecular vibrations and dihedral potential energy scans . Details of the parametrization are included in the Supporting Information ( Text S1 ) . MD simulations were performed using a stochastic boundary-based approach [80]–[84] in which three regions within the sphere were defined . Nucleotides and residues containing one or more atoms within 28 Å comprise the dynamic region , those not in the dynamic region containing one or more atoms within 34 Å comprise the buffer region , and the remaining atoms based on the 40 Å cutoff represent the constrained outer reservoir region . Nucleotides 732 , 696 , 2458 and 2459 are found along the border of the buffer and reservoir regions , and were manually assigned to the reservoir region; nucleotides 766 and 1324 border the dynamic and buffer regions and were assigned to the buffer region . All crystallographic Mg+2 ions within the 40 Å radius sphere were included in the simulation system , yielding a total of 82 ions . Atoms within the reservoir region were fixed for all calculations , while varying harmonic restraints were used on atoms within the buffer and dynamic regions as described below . Water was maintained within the sphere using a spherical , quartic restraining potential as implemented in the MMFP [85] module of CHARMM using a 1 kcal/mol/Å force constant and offset parameter ( P1 ) of 2 . 5 that was applied to the water oxygen atoms . The entire system was first subjected to 250 steps of steepest descent ( SD ) [86] minimization with a harmonic restraint of 5 kcal/mol/Å on non-hydrogen atoms within the dynamic region and a mass-weighted harmonic restraint of 10 kcal/mol/Å on non-hydrogen atoms within the buffer region , followed by 250 steps of Adopted-Basis Newton Rhapson ( ABNR ) [86] using the same restraints . Equilibration consisted of two phases of Grand Canonical Monte Carlo/Molecular Dynamics ( GCMC/MD ) , which is implemented within the MC module in CHARMM [87] . The MC module has been described in detail elsewhere [52] , [87]–[89] , thus the details of the general methodology will only briefly be addressed as they pertain to the simulations presented here . A total of 8997 water molecules were used for the bath of water molecules accessible to the GCMC move set . This number was determined by calculating the number of water molecules that would result in a spherical volume of 35 Å radius with density 0 . 0334 molecules per Å3 , then multiplied by 1 . 5 to guarantee that an excess of water molecules were available . The move set was comprised of rigid body translations , rigid body rotations , as well as insertion/deletions that were performed using an excess chemical potential of −5 . 8 kcal/mol . All moves were equally weighted . The GCMC water pool includes the waters overlaid on the system as described above and the additional water molecules that were placed in a single coordinate set slightly offset from the heterocyclic ARM of telithromycin . All crystallographic waters were set as active throughout the entire simulation . The GCMC waters were initially set as inactive , thereby removing them from all calculations . Each GCMC/MD cycle is defined as 10 , 000 MC steps and 10 , 000 MD steps . Initiation of each MD cycle required assignement of the velocities as those from the previous MD cycle are invalidated by the insertions and deletions during each GCMC cycle . Reassignment of the velocities was done by employing the Langevin integrator for the MD simulations ( see below ) . During the GCMC steps , waters can undergo moves as defined in the move set , in which particle insertions were set to be blocked by all active atoms including hydrogens . The initial hydration phase consisted of 20 GCMC/MD cycles using 5 kcal/mol/Å harmonic restraints on non-hydrogen atoms within the dynamic region and a 10 kcal/mol/Å mass-weighted harmonic restraint on non-hydrogen atoms within the buffer region . For equilibration , the list of active GCMC waters was reset and the harmonic restraints were removed for atoms within the dynamic region and reduced to 2 kcal/mol/Å on non-hydrogen atoms within the buffer region . Iterative GCMC/MD cycles were performed until the number of waters reached adequate convergence ( approximately 5 ns cumulative dynamics simulation time ) . Once the number of waters reached convergence , production consisted of GCMC/MD for a total dynamics simulation time of 25 ns using the same restraints/constraints . Only coordinates from production GCMC/MD were used for analysis . Additional GCMC/MD simulations were performed in which longer MD sampling was performed per cycle ( ie . 100 , 250 , 500 and 1000 ps ) to test convergence of the method with respect to water insertions ( see below and SI ) . As described in the SI , these different simulations of 25 ns sampling each were ultimately combined for the final analysis yielding a total of 150 ns of cumulative MD sampling for each of the studied systems . MD simulations were performed using Langevin dynamics [90] , [91] at 298 K with a friction coefficient of 5/ps . We note that the use of Langevin dynamics ( LD ) is not limited to cases without explicit solvent and in fact is often used in explicitly solvated systems as a temperature control , particularly with the program NAMD [92] , as is the case here . While the explicit presence of water molecules accounts for the frictional effects on the system , the use of LD is justified because the friction coefficient of 5/ps used is much less than that appropriate to mimic the viscosity of water ( ie . ∼60/ps ) . SHAKE [93] was applied to covalent bonds involving hydrogens , and a 2 fs integration timestep was used with the “leapfrog” Verlet integrator [94] . Nonbonded lists were updated heuristically during dynamics with a cutoff of 16 Å , the forces truncated at 12 Å , and a switching function applied to the forces from 10 to 12 Å for both electrostatic and van der Waals energy terms [95] . Coordinates were saved every 10 ps for analysis . Hydrogen bonds were considered present if the hydrogen donor and acceptor atoms came within 2 . 4 Å [96] and have an occupancy greater than 10% .
|
Bacterial resistance to antibiotics is a serious public health problem that requires the continuous development of new antibiotics . Bacteria acquire resistance to macrolide antibiotics by ( 1 ) effluxing the drug from the cell , ( 2 ) modifying the drug , or ( 3 ) modifying the drug target ( i . e . , the 50S subunit of the ribosome ) to abrogate or completely abolish binding . While newer antibiotics are able to avoid the first two mechanisms , they remain unable to overcome resistance due to ribosomal modification , particularly due to methyltransferase ( i . e . , erm ) enzymes . We have applied computer-aided drug design methods designed explicitly for studies of the ribosome to better understand the relationship between modification of the ribosome by erms and the binding of telithromycin , a 3rd generation ketolide antibiotic derived from erythromycin . While we confirm that ribosomal modification leads to decreased binding due to disruption of key interactions with the drug , we find these modifications effect a structural rearrangement of the entire region of the ribosome responsible for binding macrolide antibiotics . This information will be useful in the design of novel antibiotics that are effective against resistant bacteria possessing modified ribosomes .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"computational",
"chemistry",
"chemistry",
"microbiology",
"biology",
"biophysics",
"computational",
"biology"
] |
2013
|
Impact of Ribosomal Modification on the Binding of the Antibiotic Telithromycin Using a Combined Grand Canonical Monte Carlo/Molecular Dynamics Simulation Approach
|
Infection with dengue virus results in a wide range of clinical manifestations from dengue fever ( DF ) , a self-limited febrile illness , to dengue hemorrhagic fever ( DHF ) which is characterized by plasma leakage and bleeding tendency . Although cardiac involvement has been reported in dengue , the incidence and the extent of cardiac involvement are not well defined . We characterized the incidence and changes in cardiac function in a prospective in-patient cohort of suspected dengue cases by serial echocardiography . Plasma leakage was detected by serial chest and abdominal ultrasonography . Daily cardiac troponin-T levels were measured . One hundred and eighty one dengue cases were enrolled . On the day of enrollment , dengue cases that already developed plasma leakage had lower cardiac index ( 2695 ( 127 ) vs 3188 ( 75 ) ( L/min/m2 ) , p = . 003 ) and higher left ventricular myocardial performance index ( . 413 ( . 021 ) vs . 328 ( . 026 ) , p = . 021 ) and systemic vascular resistance ( 2478 ( 184 ) vs 1820 ( 133 ) ( dynes·s/cm5 ) , p = . 005 ) compared to those without plasma leakage . Early diastolic wall motion of the left ventricle was decreased in dengue cases with plasma leakage compared to those without . Decreased left ventricular wall motility was more common in dengue patients compared to non-dengue cases particularly in cases with plasma leakage . Differences in cardiac function between DF and DHF were most pronounced around the time of plasma leakage . Cardiac dysfunction was transient and did not require treatment . Transient elevated troponin-T levels were more common in DHF cases compared to DF ( 14 . 5% vs 5% , p = 0 . 028 ) . Transient left ventricular systolic and diastolic dysfunction was common in children hospitalized with dengue and related to severity of plasma leakage . The functional abnormality spontaneously resolved without specific treatment . Cardiac structural changes including myocarditis were uncommon .
Infection by dengue viruses ( DENV ) results in clinical presentation ranging from asymptomatic infection to a fatal viral hemorrhagic fever characterized by plasma leakage and bleeding [1–3] . Plasma leakage in dengue occurs around the time of defervescence , is localized primarily in the serosal cavities , and usually lasts approximately 48–72 hours . Severe plasma leakage can lead to shock ( dengue shock syndrome ) . Although the principal mechanism of shock is due to decreased intravascular volume , abnormal cardiac function may contribute to cardiovascular compromise . Several studies have reported cardiac involvement in dengue including myocarditis and heart failure [4–10] . Functional impairment and electrocardiographic abnormalities have also been described [11–14] . The incidence and severity of cardiac involvement in dengue has varied between studies , likely as a result of differences in study design , methodology , and populations . Early detection of plasma leakage and fluid replacement is critical in the treatment of dengue . It has been observed that excessive fluid treatment can lead to pulmonary edema in some cases [15–17] . Increased intravascular volume due to fluid intake and reabsorption of fluid from serosal cavities has been thought to be the underlying mechanism . However , it remains possible that abnormalities in cardiac function may also contribute to pulmonary edema . To further our understanding of cardiac involvement in dengue , we undertook serial echocardiographic studies of cardiac function in dengue cases . We characterized the dynamics of cardiac functional indices and analyzed them in the context of the clinical course and the extent of plasma leakage . Our findings indicate that myocarditis is uncommon , but transient functional changes are common and correlate with the extent of plasma leakage . Diastolic dysfunction characterized by impaired left ventricle ( LV ) relaxation was the most prominent abnormality . These findings have practical implications for fluid management in dengue .
The study was approved by the Institutional Review Boards of the Thai Ministry of Public Health , the Walter Reed Army Institute of Research Institutional Review Board . Written informed consent was obtained from the parent or the legal guardian of each study subject . Children less than 15 years of age who were hospitalized for suspected dengue at Queen Sirikit National Institute of Child Health ( QSNICH ) in 2010 to 2012 were enrolled . Criteria for suspected dengue included cases with febrile illness without an obvious focal source of infection and with compatible laboratory findings including leucopenia or thrombocytopenia . Patients with chronic hematologic or immunologic conditions were excluded . Patients were treated according to World Health Organization ( WHO ) guidelines [2] . The patients were encouraged to drink . Intravenous fluid was administered in cases with dehydration and inadequate oral fluid intake . The rate and amount of fluid were adjusted according to the clinical status and the severity of dehydration . In shock cases , 10 ml/kg fluid was given as bolus or within one hour for resuscitation and the rate of fluid was adjusted subsequently according to clinical status following an established treatment guideline at QSNICH . DENV infections were confirmed by RT-PCR to detect viral RNA in plasma obtained on the day of study enrollment and by serology of paired acute and convalescent plasma [18 , 19] Daily complete blood count , plasma albumin and serum aspartate and alanine aminotransferase ( AST , ALT ) levels were obtained . Echocardiographic and chest and abdominal ultrasonographic studies were performed daily . Cases with negative dengue serology were classified as non-dengue febrile illness . Confirmed dengue cases were assigned as dengue fever ( DF ) or dengue hemorrhagic fever ( DHF ) according to the 1997 WHO case definitions [2] . DHF cases were graded according to severity as grades I-IV . Cases were also classified as dengue and severe dengue according to the 2009 WHO guidelines[3] . Severe dengue was defined as cases with 1 ) plasma leakage requiring fluid resuscitation from shock , 2 ) significant bleeding ( defined as cases that required blood transfusion in this study ) , and 3 ) evidence of organ failure including AST or ALT >1000 IU/ml . Case classification was performed after the completion of the study by investigators not involved in patient care . The day of defervescence ( temperature <38°C ) was defined as fever day 0 . Days before and after defervescence were defined as fever day –1 , -2 , and +1 , +2 , etc . The study was approved by the Institutional Review Boards of the QSNICH , Thai Ministry of Public Health , and the Walter Reed Army Institute of Research . Written informed consent was obtained from the legal guardian of each participant . Assent was obtained from children who were at least 7 years old . Daily plasma samples obtained during hospitalization and at the early convalescence follow up visit ( approximately 5 days after hospital discharge ) were assayed for troponin-T levels using the Elecsys Troponin T hs assay ( Roche Diagnostics , Elcsys , Germany ) . Levels >30 pg/ml were considered elevated . Plasma samples collected on the first day of hospitalization and at the early convalescence visit were also analyzed for creatine kinase MB ( CPK-MB ) isoenzyme levels by ELISA ( MyBioSource , U . S . A . ) . The assay range was 0 . 2–60 ng/ml . All tests were performed in batch after study completion and without the knowledge of the clinical diagnosis . Daily ultrasonography was performed as previously described [20] . The vertical dimensions of pleural effusions were measured as the distance between the top of the dome of the diaphragm and the base of the lung visualized by upright midaxillary longitudinal scans of the right hemithorax . The dimensions of ascitic fluids in the perivesicular area were measured in the transverse scan of the lower abdomen . Daily echocardiography was performed by a single cardiologist ( TK ) using a CX50 CompactXtreme ultrasound system ( Philips Healthcare ) . All measurements were obtained on a daily basis without knowledge of the diagnostic laboratory results . Systolic and diastolic blood pressures and electrocardiograms were recorded during the examinations . Routine 2-D echocardiogram and color flow Doppler were obtained in subcostal and apical 4-chamber views . An M-mode scan of the LV obtained from a standard parasternal long-axis view , at the level of the mitral valve ( MV ) tip , was recorded simultaneously with the electrocardiogram . Measurements of LV walls and dimensions were performed in accordance with published guidelines [21 , 22] . Transmitral pulsed-wave Doppler velocities ( peak E- and A-wave velocities ) were measured in the apical four chamber view with the sample volume positioned at the MV . Tissue Doppler Imaging ( TDI ) of the LV was performed using pulsed wave Doppler assessment of the medial and lateral MV annulus . Peak tissue medial and lateral S-wave ( S ) , E-wave ( Ea ) and A-wave ( Aa ) velocities were measured . Myocardial performance index ( MPI ) was calculated from TDI of the LV using the following formula: ( isovolumic contraction time + isovolumic relaxation time ) /ejection time ) . Blood pressure was measured by oscillometric method at the time of echocardiography . An average value from at least three consecutive measurements was calculated . TDI parameters were assessed using published age-specific normal values [23] . Values below the 5th or above the 95th percentile of normal values were deemed abnormal . The inferior vena cava ( IVC ) diameter was measured using the subcostal view , below the level of the hepatic veins . Pericardial effusions , abnormal electrocardiograms , and other anatomical and functional findings were recorded when present . Variables are reported as mean ( SE ) or number ( % ) as appropriate . Analysis of categorical variables was performed using Chi square . Continuous variables were tested with Shapiro-Wilk’s tests for normality of distribution . Normally distributed continuous variables were compared using ANOVA with post hoc analysis or by Student’s t-test . Mann-Whitney U test was used for covariates with non-normal distribution . Correlations were analyzed using Spearman’s analysis for non-parametric data and Pearson’s analysis for parametric data . A p value ≤ 0 . 05 was considered significant . All analyses were performed using SPSS ( version 14 ) .
Between 2010 and 2012 , 861 cases were screened , 320 cases met enrollment criteria; 181 dengue cases and 35 non-dengue cases were enrolled . The reasons for not being enrolled were: 1 ) patients already developed clinical signs of shock at the time of screening , 2 ) failure to obtain informed consent and/or assent , 3 ) the number of cases exceeded the weekly enrollment limit ( 6 cases per week ) . Table 1 shows clinical and laboratory findings on the first day of the study . Dengue cases were classified as 119 DF cases and 12 , 28 , 21 , and 1 cases of DHF Grades I , II , III , and IV , respectively ( Table 1 ) . Twenty-three cases were classified as severe dengue ( SD ) according to 2009 WHO guidelines; all met the DHF case definition . The diagnosis of non-dengue cases were: unspecified viral illness ( 28 cases ) , influenza ( 3 cases ) , respiratory syncytial virus infection ( 1 case ) , pneumonia ( 1 case ) , and sinusitis ( 2 cases ) . Dengue cases were older than non-dengue febrile illness cases ( p = . 002 , ANOVA ) and had higher hematocrit levels on the day of enrollment ( p = . 005 , ANOVA ) . DHF cases had higher levels of alanine aminotransferase ( ALT ) and hematocrit and lower albumin and platelet counts compared to DF cases ( all at p<0 . 05 , ANOVA ) . There were no deaths or intensive care unit admissions , and none received vasopressive or inotropic support . Table 2 shows the hemodynamic and cardiac functions of dengue cases on the first study day . Cases were classified into DF and DHF . To examine cardiovascular functions in the context of plasma leakage , we further divided DHF cases based on the presence of ultrasonographic evidence of plasma leakage at this time point . Differences were observed between DHF cases that already showed plasma leakage at this time point ( n = 36 ) compared to DHF cases that had not yet developed plasma leakage ( n = 24 ) and DF cases ( n = 119 ) . DHF cases with plasma leakage had decreased stroke volume and cardiac index ( CI ) , and elevated systemic vascular resistance ( SVR ) ( p = . 001 , . 003 , . 003 , respectively , Student’s t-test ) . These indices were not different between DF cases and DHF cases that had not yet developed plasma leakage . Diastolic and systolic cardiac indices also differed between groups . MV-E wave and ejection fraction ( EF ) were lower in DHF cases with plasma leakage compared to those without ( p = . 01 , and . 068 ) and to DF ( p = . 001 , and . 006 ) . There were 6 cases with EF < 56%; all had evidence of plasma leakage . There were no differences in age , sex , and viral serotypes of DHF cases with or without plasma leakage at this time point . The above findings suggest that differences in hemodynamic status and LV functions were associated with plasma leakage and decreased intravascular volume . To examine cardiac functions that are less volume dependent , we performed tissue Doppler imaging ( TDI ) of the LV . TDI measurements of myocardial excursion at the MV during diastole and systole revealed lower early LV diastolic motion as indicated by lower Ea velocity in DHF with leakage ( Table 3 ) . E/Ea ratios were reduced in these patients indicating comparatively lower LV filling pressure . Mean myocardial performance index ( MPI ) of the LV was higher ( indicating poorer cardiac performance ) in DHF with leakage compared to the other two groups . The extent of plasma leakage as indicated by the size of pleural effusion as measured on ultrasound examination inversely correlated with early and late LV diastolic wall movement in both the septal and lateral regions ( R = - . 296 , - . 158 , and- . 273 for lateral Ea , lateral Aa , and septal Ea waves , respectively . ( P = . 001 , . 027 , . 001 ) . However , no correlations were observed between the severity of plasma leakage and LV systolic wall movement ( septal S wave and lateral S wave ) . Comparative analysis of hemodynamic and cardiac functional indices observed on the first day of the study in DHF cases who did not have plasma leakage at that time point to the values of these indices observed at the time of subsequent plasma leakage demonstrated evidence of volume contraction and worsening cardiac function ( S1 Table and S2 Table ) . However , the differences were comparatively smaller than those between DHF with or without leakage at the study enrollment . This may be due to the close clinical monitoring and timely fluid intervention in these cases . High proportions of non-dengue and dengue cases demonstrated abnormal TDI parameters including low amplitudes of S waves and Ea waves indicating defects in systolic and diastolic movement of the left ventricle ( Table 4 and S3 Table ) . Forty-two to sixty-three percent of dengue patients had low medial or lateral S waves compared to 21–33% in non-dengue cases but these differences were not statistically significant . However , the proportions of dengue cases with low Ea wave amplitudes were higher than non-dengue cases ( p = . 001 ) with the highest frequency found in DHF cases who already developed plasma leakage . The percentages of cases with abnormally high E/Ea ratio , which indicates elevated LV filling pressure , were higher in DHF cases that had not developed plasma leakage ( 64% ) compared to DF ( 41% , p = . 049 ) and DHF with plasma leakage ( 22% , p = . 001 ) . Fig 1 shows the patterns of hemodynamic and volume indices and fluid intake over the course of the illness of DF and DHF cases . DHF cases had higher heart rate compared to DF early in the course of illness and this persisted throughout the acute illness ( A ) . There was no difference in systolic blood pressure ( B ) , but mean diastolic pressures were higher in DHF cases around the day of fever resolution ( fever day 0 ) and shortly thereafter ( C ) . This coincided with decreased intravascular volume as indicated by a decrease in the average IVC diameter in DHF cases ( D ) . DHF cases had lower CI ( E ) at defervescence ( fever day 0 ) and increased SVR ( F ) on fever day 0 and +1 compared to DF cases . Ultrasonography revealed a gradual increase in the percentage of cases with plasma leakage , with peak incidence on fever day +1 ( G ) . Patients with DHF also received more fluid compared to those with DF starting on fever day 0 and afterward . Evaluation of LV systolic and diastolic functions revealed lower EFs ( Fig 2A ) on fever day -1 and 0 , and lower early component of LV inflow ( MV-E wave ) on fever days 0 and +1 ( Fig 2B ) in DHF compared to DF . The late component LV inflow was also decreased in DHF on fever day +1 ( Fig 2C ) . Seventeen dengue cases had at least one abnormal EF ( <56% ) detected during the illness . They represented 7% , 9% , 15% , and 32% of DF , DHF grade I , II , and DHF Gr III/IV cases , respectively . The relative frequencies of abnormal EFs were higher in DHF grade III/IV compared to DHF grade I/II and DF ( p < . 001 , Chi-square ) . The majority of cases ( 67% ) had low EF detected on only a single day , usually fever day 0 or +1 . The EFs on the day of discharge were improved compared to the lowest EF . Taken together , this indicates that changes in hemodynamic status in DHF were temporally associated with plasma leakage and were characterized by a compensatory autonomic response to contracted intravascular volume which was corrected by fluid replacement . TDI of the MV showed slower septal movement during systole in DHF compared to DF on fever day +1 ( p = . 031 ) ( Fig 3B ) . Early and late diastolic LV annulus motions ( Ea and Aa wave ) were lower in DHF cases compared to DF cases at the end of the febrile period and after ( Fig 3D , 3E and 3F ) . The decreased early diastolic annulus motion was more pronounced in the septal area . The E/Ea ratio , which indicates LV filling pressure , was not different between DHF and DF cases on most days except on fever day +1 when the average E/Ea ratio was significantly lower in DHF cases ( I ) . Twenty to seventy percent of dengue cases had abnormally low diastolic and systolic wall motion during the course of the illness ( Fig 4A–4D ) . The frequencies of abnormally low diastolic wall motion were higher in DHF compared to DF on fever day 0 and after ( Fig 4C and 4D ) . Elevated E/Ea ratios were found in both DF and DHF cases with similar frequencies . None had E/Ea ratio above 15 . Troponin and CPK-MB levels were similar between dengue and non-dengue cases on the day of study enrollment . Plasma levels of cardiac troponin-T were low to non-detectable in most cases with comparable levels found in DF and DHF cases throughout the acute illness period . Consistent with this finding , the mean levels of CPK-MB were similar between DF and DHF . However , 14 . 5% of DHF cases had troponin-T levels >30 ng/L at any time point compared to 5% in DF ( p = 0 . 028 , Chi-square test ) . All of these cases had only one sample with elevated troponin-T levels during the course of illness . These elevated troponin levels were detected later in the course of the illness ( Fig 5 ) . There were no cases with clinical heart failure or conductive defects consistent with clinical myocarditis . Pericardial effusion was detected in 7 DHF cases . The size of the effusion was minimal in 6 cases and substantial in 1 case . Sinus bradycardia was documented in 1 DF case . Four cases of mild mitral regurgitation ( 2 DF and 2 DHF ) and 2 DF cases with tricuspid regurgitation were found . These findings were transient and not clinically significant . Twelve patients ( 4 DF , 8 DHF ) had a follow up examination performed at early convalescence ( < 1 week after discharge ) . Trends toward improvement were found for EF and LV inflow at the MV ( MV-E ) and diastolic medial LV wall movement ( medial Ea wave ) in DHF cases at convalescence compared to fever day +1 ( Table 5 ) .
Much of the literature on cardiac manifestations in dengue consists of case reports and small case series . More recent prospective studies have reported varied incidence of abnormal cardiac findings from approximately 15–27% for myocarditis [8 , 24 , 25] and up to 40% for functional abnormalities[26] . Cardiac performance and hemodynamic status are affected by intravascular volume , cardiac functions and autonomic response . The present study extends our knowledge of cardiac involvement in dengue in several ways . We obtained daily echocardiographic studies over the critical phase of illness in a cohort that included both milder and more severe illness , and analyzed cardiac function data in conjunction with serial assessments of intravascular volume status and plasma leakage . The data show that cardiac functional abnormalities are common in dengue and correlate with disease severity . However , these abnormalities were transient , did not require specific treatment , and were not accompanied by evidence of structural damage to the myocardium . Our study found that cardiac functional abnormalities were related to the severity of plasma leakage . Decreased EF , CI , LV diastolic inflow , and the elevated SVR in DHF are therefore likely to be affected by contracted intravascular volume . TDI indices , which are less preload dependent [27 , 28] , also differed between dengue and non-dengue cases , and between DF and DHF . This was most evident in decreased LV wall movement during early diastole indicating a relaxation defect . The higher frequencies of cases with decreased LV early diastolic wall movement in DF compared to non-dengue cases suggests a dengue specific process independent of plasma leakage . However , the correlation between abnormal LV diastolic movement and the extent of plasma leakage among dengue cases suggests that plasma leakage may also directly contribute to this functional abnormality by reducing intravascular volume . Alternatively , plasma leakage may be a correlate of the disease mechanism that affects cardiac function . The relatively similar E/Ea ratios in DF and DHF despite lower E in DHF reflect this diastolic defect . The pathology underlying this finding is unknown . Nevertheless , the abnormal LV relaxation may further compromise LV filling . Decreased LV relaxation may also contribute to high LV filling pressure when intravascular volume has been restored by intravenous fluid treatment and by reabsorption of effusion fluid , and may lead to the pulmonary edema observed in some DHF cases . The prevalence of subclinical diastolic dysfunction has been reported to increase with age , in individuals with hypertension , LV hypertrophy , and diabetes [29–31] . This has important implications for clinical care considering the shift in the demographics of dengue cases to adults who may have these common co-morbidities . Further studies in adults with dengue will be needed to address this question . Fatal dengue associated with myocardial injury and inflammation has been reported . In a study of adult and pediatric cases from Brazil , the incidence of myocarditis on the basis of clinical diagnosis or elevated biomarkers was approximately 15% [24] . A subset of these cases had abnormal echocardiography or MRI [8 , 24] . Viral antigen has been detected in cardiac myocytes , monocytes , and endothelial cells by immunostaining [7 , 8 , 32] . Contrary to these reports , we did not observe cases with myocarditis based on clinical information and cardiac enzymes . Our findings are consistent with other reports in pediatric dengue cases in Southeast Asia [26 , 33] . The difference in the incidence of myocarditis in various reports may be related to host genetic factors and locally circulating viruses , which may influence tissue tropism and host inflammatory responses . In addition , different temporal patterns of circulating serotypes may lead to distinct dengue immune status that predisposes individuals for more severe cardiac manifestation during a secondary infection . However , differences in age groups , co-morbidities , and study design may also contribute to differences in manifestations and incidence Although sustained elevated enzyme levels were not found , transient increased levels were found in some cases , particularly in DHF , which may indicate subtle myocardial injury in these cases . Cardiac dysfunction has been reported in other viral hemorrhagic fevers such as Crimean-Congo hemorrhagic fever and Puumala virus infection [34–36] . The patterns of cardiac involvement are similar to dengue and are characterized by transient functional impairment and normal cardiac enzyme levels . Another similarity is more severe involvement in adults . These findings suggest common pathophysiologic pathways including endothelial cell activation , perturbation of vascular permeability , and cardiac muscle cell dysfunction . Cardiac dysfunction is also common in sepsis in which TNF-α and nitric oxide are considered to be involved [37 , 38] . These mediators have been reported to be altered in dengue as well [39–41] . These mediators may alter cardiac function and hemodynamic status by their permeability enhancing effects , which lead to decreased preload , as well as direct effects on cardiac myocytes and on vascular resistance . The relative contribution of these mediators will need to be further addressed in relevant animal models . Declining heart rate during defervescence has been observed in dengue and attributed to increased parasympathetic activity[42] . Benign bradyarrhythmias and ectopic beats have been reported in patients with DENV infection following defervescence and were not related to disease severity[43] . The prevalence of cardiac arrhythmia in the present study was lower than previously reported . We did not perform continuous electrocardiographic monitoring and therefore some cases might have been missed . Another limitation was the lack of adult cases that may exhibit distinct cardiac manifestations . Due to the conservative use of intravenous fluid there were no cases with pulmonary edema and fluid overload in this series . However , the relatively normal or elevated E/Ea ratio in DHF cases in the presence of contracted intravascular volume suggests that diastolic dysfunction may predispose some DHF cases to pulmonary edema with aggressive administration of intravenous fluid . In summary , functional cardiac abnormalities in dengue involved both systolic and diastolic functions and correlated with the severity of plasma leakage . Cardiac structural changes such as infarction and myocarditis were not likely the underlying mechanism . Cardiac dysfunction was transient and did not require specific treatment . Decreased LV wall diastolic movement may contribute to abnormal LV filling and the predisposition for pulmonary edema in DHF . These findings underscore the importance of vigilance in fluid management and hemodynamic status monitoring in the treatment of dengue .
|
Dengue is a viral infection with a wide range of symptoms from a self-limiting fever called dengue fever ( DF ) to dengue hemorrhagic fever ( DHF ) which is characterized by leaky blood vessels and bleeding that can lead to shock in severe cases . Abnormal heart function has been reported but the frequencies and the progression of heart involvement are not well defined . In this study children with dengue had serial evaluation of their heart function during the course of the illness . Patients with DHF had comparatively low blood volume at the time of fever resolution and had decreased blood flow into the left lower heart chamber compared to DF cases . Relaxation and contraction of the left side of the heart were also relatively decreased in DHF . These abnormalities may contribute to the clinical response and complications of fluid replacement in dengue .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Evaluation of Cardiac Involvement in Children with Dengue by Serial Echocardiographic Studies
|
Feature-based attention has a spatially global effect , i . e . , responses to stimuli that share features with an attended stimulus are enhanced not only at the attended location but throughout the visual field . However , how feature-based attention modulates cortical neural responses at unattended locations remains unclear . Here we used functional magnetic resonance imaging ( fMRI ) to examine this issue as human participants performed motion- ( Experiment 1 ) and color- ( Experiment 2 ) based attention tasks . Results indicated that , in both experiments , the respective visual processing areas ( middle temporal area [MT+] for motion and V4 for color ) as well as early visual , parietal , and prefrontal areas all showed the classic feature-based attention effect , with neural responses to the unattended stimulus significantly elevated when it shared the same feature with the attended stimulus . Effective connectivity analysis using dynamic causal modeling ( DCM ) showed that this spatially global effect in the respective visual processing areas ( MT+ for motion and V4 for color ) , intraparietal sulcus ( IPS ) , frontal eye field ( FEF ) , medial frontal gyrus ( mFG ) , and primary visual cortex ( V1 ) was derived by feedback from the inferior frontal junction ( IFJ ) . Complementary effective connectivity analysis using Granger causality modeling ( GCM ) confirmed that , in both experiments , the node with the highest outflow and netflow degree was IFJ , which was thus considered to be the source of the network . These results indicate a source for the spatially global effect of feature-based attention in the human prefrontal cortex .
Attentional selection is the mechanism by which a subset of incoming information is processed preferentially . Numerous studies have demonstrated that attentional selection can be based on a spatial location [1–7] . Alternatively , attention can also select specific features independent of their spatial locations [8] . Several studies have demonstrated that attending to different feature dimensions , such as motion and color , enhances the response of their specialized cortical modules , i . e . , human middle temporal area ( MT+ ) and V4 , respectively [9–16] . In addition , other studies have shown that attention can also select specific features within a particular dimension , such as an orientation [17–19] , a color [20 , 21] , and a direction of motion [22–25] . Feature-based attention plays a key role in identifying and highlighting a target during visual search because we often know a target-defining feature but not its exact location . The compelling evidence for the location-independent property of feature-based attention has come from its spatially global effect , as proposed by the “feature-similarity gain model” [23 , 26] , whereby feature-based attention can modulate the gain of cortical neurons tuned to the attended feature not only at the attended location but throughout the visual field . Remarkably , this spatially global effect has been demonstrated in numerous psychophysical [17 , 20 , 27 , 28] , neurophysiological [22–24 , 29] , electroencephalographic ( EEG ) [21 , 30] , magnetoencephalogram ( MEG ) [19 , 31] , and functional magnetic resonance imaging ( fMRI ) [18 , 25 , 32] studies in both striate ( V1 ) and extrastriate visual areas . Both V1 and all of the areas within extrastriate visual cortex ( V2–V4 and MT ) only respond to stimuli presented in the contralateral visual hemifield [33 , 34] . However , during feature-based attention , all of these retinotopically organized areas can be modulated when participants attend to a feature presented anywhere in the visual field . How does feature-based attention modulate neural responses in these brain areas at unattended spatial locations ( i . e . , the spatially global effect ) ? It has been well established that , during spatially directed attention , the enhanced responses in striate and extrastriate visual areas result from top-down feedback from frontoparietal cortical areas [4 , 5 , 35–37] , and previous neurophysiological [16 , 38–40] and brain imaging [13 , 41] studies have suggested that the frontoparietal network is also involved in the top-down control of feature-based attention in the attended location [5 , 8 , 42 , 43] . Moreover , previous neurophysiological studies have shown that some neurons in frontoparietal areas have very large receptive fields [38] . Although the receptive fields of these neurons are not centered in the ipsilateral hemifield , many do extend into the ipsilateral hemifield , especially with longer stimulus presentation times [44 , 45] . This may provide the underlying neural basis for the spatially global effect of feature-based attention . Thus , we hypothesized that the spatially global effect of feature-based attention in striate and extrastriate visual areas might result from top-down feedback from frontoparietal cortical areas . To test this hypothesis , we performed an fMRI experiment and used effective connectivity analysis to examine which area was involved in the spatially global effect of feature-based attention in retinotopically organized visual areas as human participants performed motion- ( Experiment 1 , a speed discrimination task ) and color- ( Experiment 2 , a luminance discrimination task ) based attention tasks . Results indicated that , in both experiments , the respective visual processing areas ( MT+ for motion and V4 for color ) as well as early visual , parietal , and prefrontal areas all showed the classic feature-based attention effect , with neural responses to the unattended stimulus significantly elevated when it shared the same feature with the attended stimulus . Furthermore , effective connectivity analysis using dynamic causal modeling ( DCM ) showed that the spatially global effect in the respective visual processing ( MT+ for motion and V4 for color ) , intraparietal sulcus ( IPS ) , frontal eye field ( FEF ) , medial frontal gyrus ( mFG ) , and V1 was derived by feedback from the inferior frontal junction ( IFJ ) . Complementary effective connectivity analysis using Granger causality modeling ( GCM ) confirmed that , in both experiments , IFJ showed the highest outflow and netflow degree in the network and thus was considered to be the source of the network . Together , our findings indicate a source for the spatially global effect of feature-based attention in the human prefrontal cortex .
Using a block design , Experiments 1 and 2 aimed to measure the feature-based attention effect as human participants performed motion- ( Experiment 1 ) and color- ( Experiment 2 ) based attention tasks , respectively . Both Experiments 1 and 2 consisted of six functional runs . Each run consisted of eight stimulus blocks of 16 s , interleaved with eight blank intervals of 12 s . There were four different stimulus blocks: 2 ( attended feature: Upward/Downward in Experiment 1; Red/Green in Experiment 2 ) × 2 ( feature match: Same/Different ) . In the Same condition , the feature on the ignored side matched the attended feature on the target side ( half the blocks ) ; a Different condition was defined as a mismatch ( half the blocks ) ( Fig 1 ) . Each stimulus block was randomly repeated two times in each run , and consisted of eight trials . On each trial , the stimulus was presented for 0 . 6 s , followed by a 1 . 4-s fixation interval . For attentional control , participants needed to press one of two buttons to indicate a 0 . 2-s speed and luminance change ( increase or decrease ) of the attended stimulus in Experiments 1 and 2 , respectively . The speed and luminance changes were determined by QUEST [46] before scanning to ensure that participants performed equally well for the Same and Different conditions . The change detection thresholds , response accuracies , and reaction times for the Same and Different conditions in Experiments 1 and 2 are shown in S1 Fig . Paired t tests revealed that there was no significant difference ( all p > 0 . 05 ) in all these measurements between the Same and Different conditions in either Experiments 1 or 2 . Regions of interest ( ROIs ) in V1–V4 and MT+ were defined as the cortical regions responding significantly to the stimulus corresponding to the target and ignored sides of the display ( Fig 2A ) . Blood oxygenation level–dependent ( BOLD ) signals were extracted from these ROIs and then averaged according to the feature match ( the Same and Different conditions ) . For each stimulus block , the 2 s preceding the block served as a baseline , and the mean BOLD signal from 5 s to 16 s after stimulus onset was used as a measure of the response amplitude . For each participant , and for each side of the display ( the target and ignored sides ) and each ROI , we computed an attentional modulation index ( IA ) to quantify how much the measured response increased during the Same condition relative to the overall response to the stimuli in the ROI . The index was calculated as follows: IA = ( A Same − A Different ) / ( A Same + A Different ) *100% , where A Same and A Different are the mean response amplitudes ( A ) in the Same and Different conditions , respectively . We hypothesized that if a cortical area shows a feature-based attention effect , the area should show a higher response in the Same condition than that in the Different condition . The IA of this area then should be significantly higher than zero . However , if the cortical area does not show the feature-based attention effect , the IA should not be significantly different than zero . Fig 2 shows the mean BOLD amplitudes in the Same and Different conditions and the corresponding IA of the target and ignored sides . For the target side , V1–V4 and MT+ did not show a significantly higher response in the Same condition than that in the Different condition ( Fig 2B and 2D , left ) , and none of these areas showed an IA significantly different than zero in either Experiment 1 ( V1: t18 = 1 . 014 , p = 0 . 324; V2: t18 = 0 . 266 , p = 0 . 794; V3: t18 = 1 . 186 , p = 0 . 251; V4: t18 = 0 . 811 , p = 0 . 428; MT+: t18 = 1 . 096 , p = 0 . 288 , Fig 2C , left ) or Experiment 2 ( V1: t18 = 0 . 751 , p = 0 . 462; V2: t18 = 0 . 538 , p = 0 . 597; V3: t18 = 0 . 776 , p = 0 . 448; V4: t18 = −0 . 067 , p = 0 . 948; MT+: t18 = 0 . 008 , p = 0 . 994 , Fig 2E , left ) . These findings confirmed that there was no difference in task difficulty or , presumably , attention between the Same and Different conditions . However , for the ignored side of the display , all of these areas showed a significantly greater response in the Same condition than that in the Different condition ( Fig 2B and 2D , right ) , and their IA’s were significantly above zero in both Experiment 1 ( V1: t18 = 4 . 521 , p < 0 . 001; V2: t18 = 4 . 106 , p = 0 . 001; V3: t18 = 4 . 661 , p < 0 . 001; V4: t18 = 3 . 032 , p = 0 . 007; MT+: t18 = 5 . 526 , p < 0 . 001 , Fig 2C , right ) and Experiment 2 ( V1: t18 = 3 . 961 , p = 0 . 001; V2: t18 = 3 . 863 , p = 0 . 001; V3: t18 = 3 . 749 , p = 0 . 001; V4: t18 = 6 . 428 , p < 0 . 001; MT+: t18 = 2 . 755 , p = 0 . 013 , Fig 2E , right ) . These results demonstrated that both striate and extrastriate visual areas showed the classical feature-based attention effect , with responses to the ignored stimulus significantly elevated when it shared the same feature as the attended stimulus . To identify the area showing the largest feature-based attention effect , we submitted the IA in these two experiments to a repeated-measure ANOVA with stimulus side ( the target side and ignored side ) and cortical area ( V1–V4 and MT+ ) as within-participant factors . The main effect of stimulus side ( Experiment 1: F1 , 18 = 15 . 253 , p = 0 . 001; Experiment 2: F1 , 18 = 28 . 443 , p < 0 . 001 ) , the main effect of cortical area ( Experiment 1: F4 , 72 = 4 . 441 , p = 0 . 014; Experiment 2: F4 , 72 = 2 . 620 , p = 0 . 042 ) , and the interaction between these two factors ( Experiment 1: F4 , 72 = 5 . 173 , p = 0 . 004; Experiment 2: F4 , 72 = 5 . 340 , p = 0 . 004 ) were all significant . Thus , these data were submitted to a further simple effect analysis . For all cortical areas ( V1–V4 and MT+ ) , IA on the ignored side was significantly greater than that on the target side in both Experiment 1 ( all t18 > 2 . 267 , p < 0 . 036 ) and Experiment 2 ( all t18 > 2 . 475 , p < 0 . 024 ) . For the target side , the main effect of cortical area was not significant in either Experiment 1 ( F4 , 72 = 0 . 216 , p = 0 . 867 ) or Experiment 2 ( F4 , 72 = 0 . 261 , p = 0 . 847 ) . For the ignored side , however , the main effect of cortical area was significant in both Experiment 1 ( F4 , 72 = 8 . 199 , p < 0 . 001 ) and Experiment 2 ( F4 , 72 = 8 . 059 , p < 0 . 001 ) . Post hoc paired t tests revealed that the IA in MT+ was significantly larger than those in V1 , V2 , V3 , and V4 ( all t18 > 3 . 251 , p < 0 . 044 ) in Experiment 1 , and the IA in V4 was significantly larger than those in V1 , V2 , V3 , and MT+ ( all t18 > 3 . 627 , p < 0 . 019 ) in Experiment 2 . These results indicated that the respective visual processing areas ( MT+ for motion and V4 for color ) showed the largest feature-based attention effect . To examine potential cortical or subcortical area ( s ) that showed a similar feature-based attention effect to these retinotopically organized areas , we performed a group analysis and did a whole-brain search with a general linear model ( GLM ) procedure [47] for cortical and subcortical area ( s ) that showed a significant higher response in the Same condition than that in the Different condition in both Experiments 1 and 2 ( note that the data from the left and right visual fields were combined ) . Statistical maps were thresholded at p < 0 . 01 and corrected by false discovery rate ( FDR ) correction [48] . The results showed that the IPS ( left: −28 ± 1 . 12 , −66 ± 1 . 27 , 39 ± 1 . 19; right: 23 ± 0 . 93 , −68 ± 1 . 14 , 40 ± 1 . 78 ) , FEF ( left: −42 ± 1 . 21 , −5 ± 0 . 94 , 35 ± 1 . 28; right: 40 ±1 . 27 , −5 ± 0 . 89 , 39 ± 2 . 00 ) , IFJ ( left: −43 ± 1 . 62 , 9 ± 1 . 34 , 31 ± 1 . 98; right: 44 ± 1 . 51 , 11 ± 1 . 30 , 29 ± 2 . 04 ) , and mFG ( left: −6 ± 0 . 51 , −2 ± 1 . 47 , 54 ± 0 . 81; right: 6 ± 0 . 46 , 2 ± 1 . 77 , 54 ± 1 . 38 ) demonstrated a greater response in the Same condition than the Different condition in both Experiment 1 ( IPS: t18 = 6 . 686 , p < 0 . 001; FEF: t18 = 2 . 807 , p = 0 . 012; IFJ: t18 = 3 . 253 , p = 0 . 004; mFG: t18 = 4 . 186 , p = 0 . 001 , Fig 3C ) and Experiment 2 ( IPS: t18 = 3 . 650 , p = 0 . 002; FEF: t18 = 5 . 652 , p < 0 . 001; IFJ: t18 = 3 . 979 , p = 0 . 001; mFG: t18 = 4 . 365 , p < 0 . 001 , Fig 3D ) . No significant difference of the IA between Experiments 1 and 2 was found in IPS ( t18 = 0 . 418 , p = 0 . 681 ) , FEF ( t18 = −0 . 027 , p = 0 . 979 ) , IFJ ( t18 = −0 . 039 , p = 0 . 969 ) , or mFG ( t18 = −0 . 314 , p = 0 . 757 ) . Furthermore , we calculated the correlation coefficients between the IA in the respective visual processing areas ( MT+ and V4 in Experiments 1 and 2 , respectively ) and that in these cortical areas across individual participants . In Experiment 1 ( Fig 3E ) , we found that the IA in MT+ correlated significantly with that in IFJ ( r = 0 . 526 , p = 0 . 021 ) and ( marginally ) with that in FEF ( r = 0 . 438 , p = 0 . 061 ) , but not with that in IPS ( r = 0 . 046 , p = 0 . 851 ) or mFG ( r = 0 . 320 , p = 0 . 182 ) . Similarly , for Experiment 2 ( Fig 3F ) , the IA in V4 correlated significantly with that in both IFJ ( r = 0 . 537 , p = 0 . 018 ) and FEF ( r = 0 . 494 , p = 0 . 032 ) , but not with that in IPS ( r = 0 . 283 , p = 0 . 240 ) or mFG ( r = 0 . 207 , p = 0 . 395 ) . These results suggested that the spatially global effect of feature-based attention in the respective visual processing areas ( MT+ for motion and V4 for color ) might derive from feedback projections from FEF and/or IFJ . To further examine which area is the source of the spatially global effect of feature-based attention in MT+ and V4 in Experiments 1 and 2 , respectively , we used DCM analysis [49] to examine functional changes in directional connectivity among the IPS , FEF , IFJ , mFG , and the respective visual processing areas ( MT+ and V4 in Experiments 1 and 2 , respectively ) related to the Same condition . Given the extrinsic visual input into MT+ and V4 in Experiments 1 and 2 , respectively , we defined 15 different models with modulatory input ( the Same condition , Fig 4A ) . The modulatory input could affect feedback from IPS ( Model 1 ) ; from FEF ( Model 2 ) ; from IFJ ( Model 3 ) ; from mFG ( Model 4 ) ; from both IPS and FEF ( Model 5 ) ; from both IPS and IFJ ( Model 6 ) ; from both IPS and mFG ( Model 7 ) ; from both FEF and IFG ( Model 8 ) ; from both FEF and mFG ( Model 9 ) ; from both IFG and mFG ( Model 10 ) ; from IPS , FEF , and IFG ( Model 11 ) ; from IPS , FEF , and mFG ( Model 12 ) ; from IPS , IFG , and mFG ( Model 13 ) ; from FEF , IFG , and mFG ( Model 14 ) ; and from all four areas ( Model 15 ) to MT+ and V4 in Experiments 1 and 2 , respectively . We examined these 15 models for modeling the modulatory effect in the Same condition for each participant . In both Experiments 1 and 2 , we computed the exceedance probability of each model [50 , 51] . The result showed that Model 3 was the best one to explain the modulatory effect in the Same condition in both Experiments 1 ( Fig 4B ) and 2 ( Fig 4E ) . The Same condition significantly increased the feedback connectivity from IFJ to MT+ ( t18 = 3 . 054 , p = 0 . 007 , Fig 4C ) and V4 ( t18 = 2 . 727 , p = 0 . 014 , Fig 4F ) in Experiments 1 and 2 , respectively . Furthermore , across individual participants , we calculated the correlation coefficients between the IA in the respective visual processing areas ( MT+ for motion and V4 for color ) and the effective connection strengths ( the sum of the intrinsic and modulatory connectivities ) from IFJ to MT+ and V4 in Experiments 1 and 2 , respectively . The IA in MT+ and V4 correlated significantly with feedback connectivity from IFJ to MT+ ( r = 0 . 544 , p = 0 . 016 , Fig 4D ) and V4 ( r = 0 . 519 , p = 0 . 023 , Fig 4G ) , respectively . Together , these results support the idea that the spatially global effect of feature-based attention in MT+ ( Experiment 1 ) and V4 ( Experiment 2 ) is derived by feedback from IFJ rather than from IPS , FEF , or mFG . However , it is unclear whether the observed involvement of IFJ in the spatially global effect of feature-based attention is relayed from other areas , namely IPS , FEF , and mFG , which also showed the classical feature-based attention effect ( Fig 3C and 3D ) . To examine this issue , we constructed three families of models with the same modulatory input ( the Same condition ) from IPS , IFJ , FEF , and mFG to MT+ and V4 in Experiments 1 and 2 , respectively . Here , the modulatory input could affect the connection from IFJ to the other three areas ( i . e . , IPS , FEF , and mFG ) in the first model family , or from these three areas to IFJ in the second model family , or the combination of these two families ( i . e . , the third model family ) . Specifically , in the first model family ( Fig 5A ) , the modulatory input could affect the connection from IFJ to IPS ( Model 1 ) to FEF ( Model 2 ) , to mFG ( Model 3 ) , to both IPS and FEF ( Model 4 ) , to both IPS and mFG ( Model 5 ) , to both FEF and mFG ( Model 6 ) , and to all three areas ( Model 7 ) . In the second model family ( Fig 5B ) , the modulatory input could affect the connection from IPS ( Model 1 ) , from FEF ( Model 2 ) , from mFG ( Model 3 ) , from both IPS and FEF ( Model 4 ) , from both IPS and mFG ( Model 5 ) , from both FEF and mFG ( Model 6 ) , and from all three areas ( Model 7 ) to IFJ . In the third model family ( Fig 5C ) , each model ( i . e . , Models 1–7 ) was the combination of corresponding models from the first and second model families . We applied Bayesian model [50] comparison to select the model with the highest exceedance probability within each model family ( model-level inference ) and the model family with the highest exceedance probability ( family-level inference ) . Within each model family , the results showed that Model 2 was the best one to explain the modulatory effect in the Same condition in both Experiments 1 ( Fig 5D ) and 2 ( Fig 5F ) . These results confirmed our correlation analyses ( Fig 3E and 3F ) showing that FEF was more important than IPS and mFG in the spatially global effect of feature-based attention . More importantly , across the model families , the results showed that the first model family had a higher exceedance probability than the other two model families in both Experiments 1 ( Fig 5E ) and 2 ( Fig 5G ) . These results indicate that IFJ may be the source of the spatially global effect of feature-based attention in IPS , FEF , and mFG . Furthermore , additional DCM analyses indicated that the spatially global effect in V1 was also derived by feedback from IFJ rather than by feedback from MT+ ( Experiment 1 ) or V4 ( Experiment 2 ) ; this feedback significantly predicted the spatially global effect in V1 in both experiments ( S5 Fig ) . Note that the contralateral and ipsilateral ROIs to the ignored side in IPS , FEF , IFJ , and mFG were pooled together in the above analyses . We next examined whether there was any difference between the contralateral and ipsilateral ROIs in mediating the spatially global effect of feature-based attention . To do so , we constructed two families of models with the contralateral ROIs only ( the contralateral model family ) and the ipsilateral ROIs only ( the ipsilateral model family ) . We applied a Bayesian model [50] to compare the exceedance probability between the contralateral and ipsilateral model families . The results showed that the contralateral model family had a higher exceedance probability than the ipsilateral model family in all DCM analyses , suggesting that the contralateral ROIs were more important than the ipsilateral ROIs in the spatially global effect of feature-based attention . However , within each model family ( i . e . , the contralateral and ipsilateral model families ) , the results confirmed our previous findings by showing that: ( 1 ) the spatially global effect of feature-based attention in MT+ ( Experiment 1 ) and V4 ( Experiment 2 ) is derived by feedback from IFJ rather than from IPS , FEF , or mFG ( S4 Fig ) ; ( 2 ) IFJ mediates the spatially global effect of feature-based attention in IPS , FEF , and mFG and thus is considered to be a source of the spatially global effect of feature-based attention in these areas ( S5 Fig ) ; and ( 3 ) the spatially global effect of feature-based attention in V1 is dependent on feedback from IFJ rather than MT+ ( Experiment 1 ) or V4 ( Experiment 2 ) ( S7 Fig ) . In addition , we used GCM [52] , a data-driven approach , to further examine which area was a potential source of the spatially global effect for feature-based attention in both experiments . In Experiment 1 , for both contralateral and ipsilateral GCM analyses , our results clearly showed that the node with the highest outflow and netflow degree was node 2 ( i . e . , the node located in IFJ ) , which was thus considered to be the source of the network ( S8A and S8B Fig , left and right ) . The node with the highest inflow degree was node 5 ( i . e . , the node located in MT+ ) , which was thus considered to be the sink of the network ( S8A and S8B Fig , middle ) . Similar results were found in Experiment 2 . For both contralateral and ipsilateral GCM analyses , node 2 located in IFJ showed the highest outflow and netflow degree and thus was considered to be the source of the network ( S8C and S8D Fig , left and right ) . Node 5 located in V4 showed the highest inflow degree and was thus considered to be the sink of the network ( S8C and S8D Fig , middle ) . Together , our GCM results further confirmed our DCM results by showing that IFJ mediated the spatially global effect of feature-based attention in both visual processing and frontoparietal areas .
Numerous psychophysical [17 , 20 , 27 , 28] , neurophysiological [22–24 , 29] , EEG [21 , 30] , MEG [19 , 31] , and fMRI [18 , 25 , 32] studies have indicated the spatially global effect of feature-based attention in visual processing areas , as proposed by the “feature-similarity gain model” [23 , 26] , whereby responses to stimuli that share features with an attended stimulus are enhanced not only at the attended location but throughout the visual field . However , how feature-based attention modulates cortical neural responses at unattended spatial locations remains unclear . Here we performed an fMRI experiment and used effective connectivity analysis to examine this issue as human participants performed motion- ( Experiment 1 , a speed discrimination ) and color- ( Experiment 2 , a luminance discrimination ) based attention tasks . In both experiments , our data indicated that human IFJ mediated cortical neural responses at unattended locations and could be a source of the spatially global effect for feature-based attention in the respective visual processing areas ( MT+ for motion and V4 for color ) . First , IFJ responses showed the classic feature-based attention effect , with neural responses to the unattended stimulus significantly elevated when it shared the same feature with the attended stimulus ( Fig 3C and 3D ) , and this response correlated significantly with the spatially global effect in MT+ ( Experiment 1 , Fig 3E ) and V4 ( Experiment 2 , Fig 3F ) . Second , the DCM analysis indicated that the spatially global effect in MT+ ( Experiment 1 , Fig 4B ) and V4 ( Experiment 2 , Fig 4E ) was derived by feedback from IFJ rather than IPS , FEF , or mFG . Moreover , the increased feedback from IFJ significantly predicted the spatially global effect in MT+ ( Experiment 1 , Fig 4D ) and V4 ( Experiment 2 , Fig 4G ) . Third , the GCM analysis indicated that , in both experiments , the node in IFJ showed the highest outflow and netflow degree and was thus considered to be the source of the network ( S8 Fig ) . Fourth , IFJ not only mediated the spatially global effect of feature-based attention in MT+ ( Experiment 1 ) and V4 ( Experiment 2 ) but also in other visual processing areas ( i . e . , V1–V3 , S3 Fig ) . Moreover , our additional DCM analyses also indicated that the spatially global effect in V1 was also derived by feedback from IFJ rather than by feedback from MT+ ( Experiment 1 ) or V4 ( Experiment 2 ) , and this feedback significantly predicted the spatially global effect in V1 in both experiments ( S6 Fig ) . Altogether , our results indicate that IFJ may be a source of the spatially global effect for feature-based attention in visual processing areas . Our study also found the classical feature-based attention effect in IPS , FEF , and mFG , which showed a greater neural response in the Same condition than that in the Different condition ( Fig 3C and 3D ) , consistent with previous neurophysiological [16 , 38–40] and brain imaging [13 , 41] studies showing feature-based attentional selection in the frontoparietal attention network [5 , 8 , 42 , 43] . Importantly , our DCM analyses suggested that the feature-based attention effect in IPS , FEF , and mFG was derived by feedback from IFJ in both of the two experiments ( the first model family showed the highest exceedance probability , Fig 5 ) . Moreover , our GCM analyses further indicated that , in both experiments , IFJ showed the highest outflow and netflow degree and was thus considered to be the source of the network containing these frontoparietal areas ( S8 Fig ) . These results are consistent with recent neurophysiological studies [38–40] . First , Ibos and Freedman [39 , 40] found that neurons in the lateral intraparietal cortex ( LIP , the monkey homologue of human IPS ) acted as receivers of feature-based attention modulation and were not involved in the generation of top-down feature-based attention signals . Our results confirmed their findings and further indicated that the feature-based attention effect in IPS was derived by feedback from IFJ . Second , Bichot and colleagues [38] found that the ventral prearcuate ( VPA ) region of monkey prefrontal cortex , which could be the homologue of human IFJ [29 , 53] , exhibited the earliest feature-based attention effects , and that inactivation of VPA impaired the animals’ ability to find targets based on their features and simultaneously eliminated the feature-based attention effect in FEF . Our results confirmed their findings by showing that the feature-based attention effect in FEF was also derived by feedback from IFJ . Additionally , both correlation ( Fig 3E and 3F ) and DCM analyses ( Fig 5 ) in our study indicated that FEF was more important than IPS and mFG in feature-based attention , indicating that FEF may also be involved in the control of feature-based attention [16] . Combined with these existing neurophysiological studies [38–40] , we thus speculate that IFJ may be a source of feature-based selection in both the prefrontal ( i . e . , FEF and mFG ) and parietal ( i . e . , IPS ) cortex . In addition , to examine whether there was any difference between the contralateral and ipsilateral hemispheres in the spatially global effect of feature-based attention , we constructed two families of models with the contralateral ROIs only ( the contralateral model family ) and the ipsilateral ROIs only ( the ipsilateral model family ) . The results showed that the contralateral model family had a higher exceedance probability than the ipsilateral model family in all DCM analyses , suggesting that the contralateral ROIs were more important than the ipsilateral ROIs in the spatially global effect of feature-based attention . However , within both the contralateral and ipsilateral model families , the results confirmed the above findings ( S4 Fig and S5 Fig ) and further supported the conclusion that IFJ mediates the spatially global effect of feature-based attention not only in visual processing areas but also in the frontoparietal network . We believe that our fMRI results cannot be explained by any difference in task difficulty or , presumably , attention , between the Same and Different conditions . In Experiments 1 and 2 , participants were asked to detect the speed and luminance change of the attended stimulus , respectively . Speed and luminance changes were determined by QUEST [46] before scanning to ensure that participants performed equally well for the Same and Different conditions . Additionally , our fMRI results also cannot be explained by participants inadvertently shifting their spatial attention to the ignored stimulus in the Same condition , as that would have impaired task performance [7] , but there was no significant performance difference between these two conditions ( S1 Fig ) . Note that , in our study , participants needed to attend to a specific feature on the target side for each block ( Fig 1 ) . However , we did not observe a feature-based attention effect on the target side in any area ( Fig 2C and 2E ) because this attended feature was the Same condition in half of the blocks and it was the Different condition in the other half of the blocks . In the Same condition , the feature , such as the red dots , on the ignored side matched the attended feature ( i . e . , the red dots ) on the target side; a Different condition was defined as a mismatch that the feature on the ignored side was the green dots . In other words , the attended feature on the target side was always red dots between these two conditions . Thus , the BOLD response to the attended stimulus on the target side did not vary with these two conditions , and thus there was no feature-based attention effect on the target side . The most parsimonious account of our results is identification of the IFJ as a source of the spatially global effect for feature-based attention . The IFJ is a region ventrolateral to FEF and is anatomically localized at the intersection of the precentral sulcus and the inferior frontal sulcus [54 , 55] . Anatomical studies have shown that IFJ has connections with sensory , parietal , and prefrontal areas [56 , 57] , and recent resting state functional connectivity data suggest that this brain region functionally interacts with both ventral and dorsal cortical brain structures [58] . Previous studies have suggested that IFJ is involved in many different cognitive processes , including visual search [38] , spatial attention [59 , 60] , switching and Stroop tasks [61] , executive control [62] , working memory for maintaining and updating information [60 , 63–65] , object-based attention induced by feature selection [66 , 67] , and coordination of bottom-up and top-down attention [68] . Here our results extend the function of IFJ by showing a crucial involvement in the global modulation of feature-based attention . IFJ appears to control feature-based attention by actively sending top-down biasing signals for a particular feature to the visual processing areas evoked by the unattended stimulus and also to other frontoparietal areas . Notably , identifying IFJ as a source of the spatially global effect for feature-based attention derives mainly from our DCM and GCM analyses , both of which depend on time-series models of fMRI data for an interpretation of causality [69–71] . This interpretation of causality in our study finds support in previous lesion [38] and transcranial magnetic stimulation ( TMS ) [63 , 72–75] studies showing a causal effect of prefrontal cortical disruption on feature-based attention . Furthermore , previous studies have indicated that color and motion processing engage ventral and dorsal visual processing streams , respectively . Some studies found a feature-general organization in frontoparietal cortical areas by directly comparing brain activity between attention to color and motion [41 , 55] . However , other studies found feature specificity in these frontoparietal cortical areas . For example , some studies found that attending to motion generally evoked larger responses than attending to color in the dorsal attention network [41 , 42] . Moreover , using multivariate pattern analysis ( MVPA ) of fMRI , some studies found attention to color and motion could evoke different patterns of activity in frontoparietal cortical areas [41] . It is important to note that , in those studies , the entire feature stimulus was located at the attended location . However , in our study , half of the feature stimulus was located at an unattended location ( i . e . , the ignored side of the display , see Fig 1 ) , and we did not find this feature specificity in IPS , FEF , IFJ or mFG . Given the limitation of univariate analyses in our study , further work is needed to use MVPA of fMRI or neurophysiological techniques to address whether the spatially global effect of color-based and motion-based attentions is mediated by the same subpopulations within these frontoparietal cortical areas . One should note that our results cannot answer a highly debated question regarding whether spatial and feature-based attention are mediated by the same or different neural mechanisms [76] . Many previous studies have suggested that the frontoparietal network is involved in the top-down control of both spatial [4 , 5 , 35–37] and feature-based attention [5 , 8 , 13 , 16 , 38–43] by mediating the neural response in visual processing areas . Our current results confirmed the role of fronto-parietal cortical areas ( i . e . , IPS , FEF , IFJ , and mFG ) in controlling feature-based attention and identified the IFJ as a source of the spatially global effect of feature-based attention . Several studies have suggested that the IFJ’s function generalizes across both spatial and feature-based attention [60 , 77] . However , our study did not test the difference between spatial and feature-based attention directly and thus cannot address whether these two forms of attention are mediated by the same or different populations of IFJ neurons . In sum , our study implicates for the first time , to the best of our knowledge , the human IFJ as the source for the spatially global effect of feature-based attention . The prominent role of the prefrontal cortex in the spatially global effect of feature-based attention evident here is consistent with recent neurophysiological and brain imaging findings that have begun to address how prefrontal areas directly top-down modulate sensory signals within posterior cortices [78] and how they covertly maintain and manipulate visual object information [79] . Combining our results with earlier studies showing a crucial involvement of IFJ in spatial [59 , 60] , object-based [66 , 67] , bottom-up , and top-down attention [5 , 68] , IFJ may have a very general role in the control of attentional selection and awareness .
All participants gave written informed consent in accordance with a protocol approved by the National Institute of Mental Health ( NIMH ) Institutional Review Board ( NIH Clinical Study Protocol 93-M-0170 ) . A total of 21 adults ( 11 males , 19–26 years old ) participated in both Experiments 1 and 2 . One participant was excluded because of large head motion in the scanner ( >3 mm ) and another participant did not have the stamina to complete the experiments . All were naïve to the purpose of the study . They reported normal or corrected-to-normal vision and had no known neurological , psychiatric , or visual disorders . The stimulus display in both Experiments 1 and 2 was composed of two circular regions ( diameter: 8 . 0° ) in the upper visual field ( centered 8 . 5° to the left and right of the central fixation point ) . One of these regions was attended ( the target side ) and the other was unattended ( the ignored side ) ( Fig 1A ) . The target side in Experiment 1 was comprised of overlapping upward and downward moving dots ( dot speed: 10 . 0°/second , dot diameter: 0 . 186° , dot luminance: approximately 76 . 8 cd/m2 , dot density: 0 . 63/ ( ° ) 2 , each moving direction with 100% coherence ) , while the ignored side was a single field of dots moving either upward or downward ( each moving direction with 100% coherence , Fig 1A ) . The target side in Experiment 2 was comprised of overlapping fields of stationary red ( CIE [1931]: x = 0 . 620 , y = 0 . 348 ) and green dots ( CIE [1931]: x = 0 . 342 , y = 0 . 537 ) , while the ignored side was a single field of red or green dots ( Fig 1B ) . In both Experiments 1 and 2 , to maximally reduce the possibility that participants could focus on a single dot , half of the dots disappeared and were replaced by new dots at different random locations every 100 ms . Using a block design , both Experiments 1 and 2 consisted of six functional runs; three for the target side were in the left visual field ( Fig 1 ) and the other three for the target side were in the right visual field ( note that for each run , the target side was always in one hemifield and the ignored side was in the opposite hemifield ) . Each run consisted of eight stimulus blocks of 16 s , interleaved with eight blank intervals of 12 s . There were four different stimulus blocks: 2 ( attended feature: Upward/Downward in Experiment 1; Red/Green in Experiment 2 ) × 2 ( feature match: Same/Different ) . In the Same condition , the feature on the ignored side matched the attended feature on the target side ( half the blocks ) ; a Different condition was defined as a mismatch ( half the blocks ) ( Fig 1 ) . For example , participants attended to upward moving dots in the target side; the upward and downward moving dots in the ignored side indicate the Same and Different conditions , respectively . Each stimulus block was randomly repeated two times in each run , and the attended stimulus in each stimulus block was indicated by a colored fixation dot: red and green indicated upward moving dots and downward moving dots in Experiment 1 , as well as red dots and green dots in Experiment 2 , respectively . Each stimulus block consisted of eight trials; on each trial , the stimulus was presented for 0 . 6 s , followed by a fixed 1 . 4-s fixation interval , and participants did a 0 . 2-s speed and luminance discrimination task at threshold ( 75% correct , measured by QUEST [46] before scanning ) in Experiments 1 and 2 , respectively . Retinotopic visual areas ( V1 , V2 , V3 , and V4 ) were defined by a standard phase-encoded method developed by Sereno et al . [33] and Engel et al . [34] , in which participants viewed rotating wedge and expanding ring stimuli that created traveling waves of neural activity in the visual cortex . A block-design scan was used to localize the ROIs in V1–V4 and MT+ corresponding to the target and ignored stimuli ( Fig 1 ) . In both Experiments 1 and 2 , the localizer scan consisted of 12 stimulus blocks of 12 s , interleaved with 12 blank intervals of 12 s . In the stimulus block , participants were asked to press one of two buttons to indicate the random luminance change ( increase or decrease ) of the stimulus . Whereas Experiment 1 consisted of two different stimulus blocks: stationary dots and moving dots , Experiment 2 consisted of two different stimulus blocks: gray dots and colored dots . MRI data were collected using a 3T Siemens Trio scanner with a 32-channel phase-array coil . In the scanner , the stimuli were rear-projected via a video projector ( refresh rate: 60 Hz; spatial resolution: 1 , 280×800 ) onto a translucent screen placed inside the scanner bore . Participants viewed the stimuli through a mirror located above their eyes . The viewing distance was 115 cm . BOLD signals were measured with an echo-planar imaging sequence ( TR: 2 , 000 ms; TE: 30 ms; FOV: 192×192 mm2; matrix: 64×64; flip angle: 70; slice thickness: 3 mm; gap: 0 mm; number of slices: 34; slice orientation: axial ) . The bottom slice was positioned at the bottom of the temporal lobes . A 3D MPRAGE structural dataset ( resolution: 1×1×1 mm3; TR: 2 , 600 ms; TE: 30 ms; FOV: 256×224 mm2; flip angle: 7; number of slices: 176; slice orientation: sagittal ) was collected in the same session before the functional scans . Participants underwent three sessions—one for retinotopic mapping and ROI localization and the other two for Experiments 1 and 2 , respectively . Note that the MRI data analysis , whole-brain group analysis , and DCM of this study closely followed those used by our previous studies [6 , 67] and therefore , for consistency , we largely reproduce that description here , noting differences as necessary . The anatomical volume for each participant in the retinotopic mapping session was transformed into the Talairach space [80] and then inflated using BrainVoyager QX . Functional volumes in all three sessions for each participant were preprocessed , including 3D motion correction , linear trend removal , and high-pass ( 0 . 015 Hz ) [81] filtering using BrainVoyager QX . Head motion within any fMRI session was <3 mm for all participants . The images were then aligned to the anatomical volume from the retinotopic mapping session and transformed into Talairach space . The first 8 s of BOLD signals were discarded to minimize transient magnetic saturation effects . A GLM procedure was used for the ROI analysis for each participant . For each side ( i . e . , the target and ignored sides , Fig 1 ) , the ROIs in V1–V3 were defined as regions that responded more strongly to the stationary gray dots than to the blank screen ( p < 10−3 , uncorrected ) . The ROI in V4 was defined as regions that responded more strongly to the stationary colored dots than to the stationary gray dots ( p < 10−3 , uncorrected ) . The ROI in MT+ was defined as regions that responded more strongly to the moving dots than to the stationary dots ( p < 10−3 , uncorrected ) . BOLD signals were extracted from these ROIs and then averaged according to the Same and Different conditions . For each stimulus block , the 2 s preceding the block served as a baseline , and the mean BOLD signal from 5 s to 16 s after stimulus onset was used as a measure of the response amplitude . For each ROI and each participant , we computed an IA to quantify how much the measured response increased during the Same condition relative to the overall response to the stimuli in the ROI . The index was calculated as follows: IA = ( A Same − A Different ) / ( A Same + A Different ) *100% , where A Same and A Different are the mean response amplitudes ( A ) in the Same and Different conditions , respectively . The index is positive whenever the mean response in the Same condition is greater than that in the Different condition . In the whole-brain group analysis , for both Experiments 1 and 2 , a fixed-effects general linear model ( FFX-GLM ) was performed for each participant on the spatially non-smoothed functional data in Talairach space . The design matrix consisted of two predictors ( the Same and Different conditions ) , which were modeled as epochs using the default BrainVoyager QX`s two-gamma hemodynamic response function . Six additional parameters resulting from 3D motion correction ( x , y , z rotation and translation ) were included in the model . First , we calculated fixed effects analyses for each participant for the two predictors . Second , a second-level group analysis ( n = 19 ) was performed with a random-effects GLM to calculate the contrast between the two predictors . Statistical maps were thresholded at p < 0 . 01 and corrected by FDR correction [48] . To further examine which area is involved in the spatially global effect of feature-based attention in MT+ and V4 in Experiments 1 and 2 , respectively , we applied DCM analysis [49] in SPM12 to our fMRI data in both experiments . For each participant and each hemisphere , using BrainVoyager QX , V4 and MT+ voxels were identified as those activated by the colored and moving dots at a significance level of p < 0 . 005 , respectively; all IPS , FEF , IFJ , and mFG voxels were identified as those activated by the stimulus block at a significance level of p < 0 . 005 . The mean Talairach coordinates of these voxels and the standard errors across participants for the left and right hemispheres in IPS were [−28 ± 1 . 12 , −66 ± 1 . 27 , 39 ± 1 . 19] and [23 ±0 . 93 , −68 ± 1 . 14 , 40 ± 1 . 78] , respectively; those in FEF were [−42 ± 1 . 21 , −5 ± 0 . 94 , 35 ± 1 . 28] and [40 ±1 . 27 , −5 ± 0 . 89 , 39 ± 2 . 00] , respectively; those in IFJ were [−42 ± 1 . 62 , 9 ± 1 . 34 , 31 ± 1 . 98] and [44 ± 1 . 51 , 11 ± 1 . 30 , 29 ± 2 . 04] , respectively; and those in mFG were [−6 ± 0 . 51 , −2 ± 1 . 47 , 54 ± 0 . 81] and [6 ± 0 . 46 , 2 ± 1 . 77 , 54 ± 1 . 38] , respectively . For each participant and each hemisphere , these Talairach coordinates were converted to Montreal Neurological Institute ( MNI ) coordinates using the tal2mni conversion utility ( http://imaging . mrc-cbu . cam . ac . uk/downloads/MNI2tal/tal2mni . m ) . In Statistical Parametric Mapping ( SPM ) , for each of these areas , we extracted voxels within a 4-mm sphere centered on the most significant voxel and used their time series for the DCM analysis . The estimated DCM parameters were later averaged across the two hemispheres using the Bayesian model averaging method [50] . DCMs have three sets of parameters: ( 1 ) extrinsic input into one or more regions , ( 2 ) intrinsic connectivities among the modeled regions , and ( 3 ) bilinear parameters encoding the modulations of the specified intrinsic connections by experimental manipulations [49] . The third set of parameters is used to quantify modulatory effects , which reflect increases or decreases in connectivity between two regions given some experimental manipulation , compared with the intrinsic connections between the same regions in the absence of experimental manipulation . fMRI data were modeled using GLM , with regressors for the Same condition , and a second condition comprising all visual inputs ( i . e . , the Same and Different conditions ) . This second condition was added specifically for the DCM analysis to be used as the extrinsic visual input . Given the extrinsic visual input into MT+ and V4 in Experiments 1 and 2 , respectively , we defined 15 different models with the modulatory input ( the Same condition ) . The modulatory input could affect feedback from IPS ( Model 1 ) ; from FEF ( Model 2 ) ; from IFJ ( Model 3 ) ; from mFG ( Model 4 ) ; from both IPS and FEF ( Model 5 ) ; from both IPS and IFJ ( Model 6 ) ; from both IPS and mFG ( Model 7 ) ; from both FEF and IFG ( Model 8 ) ; from both FEF and mFG ( Model 9 ) ; from both IFG and mFG ( Model 10 ) ; from IPS , FEF , and IFG ( Model 11 ) ; from IPS , FEF , and mFG ( Model 12 ) , from IPS , IFG , and mFG ( Model 13 ) ; from FEF , IFG , and mFG ( Model 14 ) ; and from all four areas ( Model 15 ) to MT+ and V4 in Experiments 1 and 2 , respectively ( Fig 4A ) . We examined these 15 models for modeling the modulatory effect by the Same condition and fit each of these 15 models for each participant . Using a hierarchical Bayesian approach [50] , we compared the 15 models by computing the exceedance probability of each model , i . e . , the probability to which a given model is more likely than any other included model to have generated data from a randomly selected participant . In the best model ( Model 3 ) , we examined the modulatory effect by the Same condition . Moreover , to examine whether the observed involvement of IFJ in the spatially global effect of feature-based attention is relayed from the other three areas , namely IPS , FEF , and mFG , we constructed three families of models with the same modulatory input ( the Same condition ) from IPS , IFJ , FEF , and mFG to MT+ and V4 in Experiments 1 and 2 , respectively . Note that the modulatory input could affect the connection from IFJ to the other three areas ( i . e . , IPS , FEF , and mFG ) in the first model family , or from these three areas to IFJ in the second model family , or the combination of these two families ( i . e . , the third model family ) . Specifically , in the first model family ( Fig 5A ) , the modulatory input could affect the connection from IFJ to IPS ( Model 1 ) , to FEF ( Model 2 ) , to mFG ( Model 3 ) , to both IPS and FEF ( Model 4 ) , to both IPS and mFG ( Model 5 ) , to both FEF and mFG ( Model 6 ) , and to all three areas ( Model 7 ) . In the second model family ( Fig 5B ) , the modulatory input could affect the connection from IPS ( Model 1 ) , from FEF ( Model 2 ) , from mFG ( Model 3 ) , from both IPS and FEF ( Model 4 ) , from both IPS and mFG ( Model 5 ) , from both FEF and mFG ( Model 6 ) , and from all three areas ( Model 7 ) to IFJ . In the third model family ( Fig 5C ) , each model ( i . e . , Models 1–7 ) was the combination of corresponding models from the first and second model families . We applied Bayesian model [50] comparison to select the model with the highest exceedance probability within each model family ( model-level inference ) and the model family with the highest exceedance probability ( family-level inference ) . Eye movements were recorded with an EyeLink 1000 Plus system ( SR Research , Ltd . , Mississauga , Ontario , Canada ) in a psychophysics lab ( the scanner did not have an applicable eye tracking system ) . Recording ( 500 Hz ) was performed when participants performed the same task as Experiments 1 and 2 . S2 Fig shows that participants’ eye movements were small and statistically indistinguishable between the Same and Different conditions .
|
Attentional selection is the mechanism by which relevant sensory information is processed preferentially . Feature-based attention plays a key role in identifying an attentional target in a complex scene , because we often know the features of the target but not its exact location . The ability to quickly select the target is mainly attributed to enhancement of responses to stimuli that share features with an attended stimulus , not only at the attended location but throughout the whole visual field . However , little is known regarding how feature-based attention modulates brain responses at unattended locations . Here we used fMRI and advanced connectivity analyses to examine human subjects as they performed either motion- or color-based attention tasks . Our results indicated that the visual processing areas for motion and color showed the feature-based attention effect . Effective connectivity analysis showed that this feature-based attention effect was derived by feedback from the inferior frontal junction , an area of the posterior lateral prefrontal cortex involved in many different cognitive processes , including spatial attention and working memory . Further modeling confirmed that the inferior frontal junction showed connectivity features supporting its role as the source of the network . Our results support the hypothesis that the inferior frontal junction plays a key role in the spatially global effect of feature-based attention .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"open",
"science",
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"functional",
"magnetic",
"resonance",
"imaging",
"social",
"sciences",
"light",
"electrophysiology",
"neuroscience",
"electromagnetic",
"radiation",
"magnetic",
"resonance",
"imaging",
"cognitive",
"psychology",
"luminance",
"brain",
"mapping",
"visible",
"light",
"vision",
"eyes",
"neuroimaging",
"research",
"and",
"analysis",
"methods",
"imaging",
"techniques",
"head",
"physics",
"psychology",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"anatomy",
"science",
"policy",
"color",
"vision",
"physiology",
"biology",
"and",
"life",
"sciences",
"ocular",
"system",
"sensory",
"perception",
"physical",
"sciences",
"cognitive",
"science",
"attention",
"neurophysiology"
] |
2018
|
The role of inferior frontal junction in controlling the spatially global effect of feature-based attention in human visual areas
|
The Wnt receptor Ryk is an evolutionary-conserved protein important during neuronal differentiation through several mechanisms , including γ-secretase cleavage and nuclear translocation of its intracellular domain ( Ryk-ICD ) . Although the Wnt pathway may be neuroprotective , the role of Ryk in neurodegenerative disease remains unknown . We found that Ryk is up-regulated in neurons expressing mutant huntingtin ( HTT ) in several models of Huntington's disease ( HD ) . Further investigation in Caenorhabditis elegans and mouse striatal cell models of HD provided a model in which the early-stage increase of Ryk promotes neuronal dysfunction by repressing the neuroprotective activity of the longevity-promoting factor FOXO through a noncanonical mechanism that implicates the Ryk-ICD fragment and its binding to the FOXO co-factor β-catenin . The Ryk-ICD fragment suppressed neuroprotection by lin-18/Ryk loss-of-function in expanded-polyQ nematodes , repressed FOXO transcriptional activity , and abolished β-catenin protection of mutant htt striatal cells against cell death vulnerability . Additionally , Ryk-ICD was increased in the nucleus of mutant htt cells , and reducing γ-secretase PS1 levels compensated for the cytotoxicity of full-length Ryk in these cells . These findings reveal that the Ryk-ICD pathway may impair FOXO protective activity in mutant polyglutamine neurons , suggesting that neurons are unable to efficiently maintain function and resist disease from the earliest phases of the pathogenic process in HD .
Cell stress response factors are important for cells to maintain function in a large variety of normal and pathological contexts , including diseases linked to proteotoxicity [1] . Among these factors , the FOXO family of Forkhead transcription factors is central to longevity and cellular homeostasis [2] , [3] . Additionally , FOXO factors may be important to the regulation of neuron survival in neurodegenerative diseases such as Huntington's disease ( HD ) [4] , [5] , a dominantly inherited CAG repeat disorder caused by expanded polyglutamines ( polyQ ) in the N-terminal portion of huntingtin ( HTT ) and characterized by striatal and cortical degeneration [6] . FOXO may indeed promote neuron survival in simple models of HD [7] and cellular proteostasis in simple models of Alzheimer's disease ( AD ) [8] . Interestingly , the canonical Wnt pathway component β-catenin may functionally interact with FOXO in oxidative stress signaling [9] and has a neuroprotective role in models of the early phases of the pathogenic process in HD [10] . This suggests that the canonical Wnt pathway may interact with the FOXO pathway to promote diseased-neuron function and survival , which is in line with the notion that neuronal differentiation factors such as Wnts may promote adult neuron survival [11] . However , Wnt signaling effectors may be impaired in HD and AD [12]–[14] , raising the possibility that Wnt pathways may have a dual role in the regulation of neurodegeneration . Here , we hypothesized that alteration of Wnt pathways might antagonize the FOXO pathway to compromise cell stress response and neuronal resistance during the earliest phases of the pathogenic process in neurodegenerative disease such as HD . To test for this hypothesis , we used Caenorhabditis elegans transgenics that recapitulate an early phase of mutant HTT toxicity , namely neuronal dysfunction before cell death [15] . At the young adult stage , these animals show a dramatic loss of response to light touch produced by polyQ-expanded exon-1 like HTT fused to GFP in touch receptor neurons [15] . To assess the mechanisms that underlie the dysfunction of these neurons , we performed a microarray analysis of primary neurons upon Fluorescence Activated Cell Sorting ( FACS ) of embryonic cells . This analysis emphasized the deregulation of neuronal differentiation genes , notably genes that are up-regulated in expanded-polyQ nematodes and in the brain of HD patients such as Ryk . Ryk is an evolutionary-conserved Wnt receptor ( lin-18 in C . elegans ) that is important to neurogenesis and axon guidance [16]–[18] and that is involved in the regulation of developmental/postdevelopmental processes such as planar cell polarity [19] , [20] , regeneration [21] , and hematopoietic repopulation [22] . Further investigation revealed that loss-of-function ( LOF ) of lin-18/Ryk in polyQ nematodes and reduction of Ryk levels in mouse striatal cells derived from HdhQ111 knock-in mice [23] strongly protected from expanded polyQ/mutant HTT . Neuroprotection by lin-18 LOF in expanded-polyQ nematodes , a cell-autonomous process , required the neuroprotective factor daf-16/FOXO [7] , suggesting that lin-18 represses the neuroprotective activity of daf-16 in these animals . The intracellular domain of Ryk ( Ryk-ICD ) , a γ-secretase cleavage product that translocates in the nucleus to control neurogenesis [16] , [17] , was found to bind to the FOXO partner β-catenin , suggesting that Ryk-ICD may trigger the repression of FOXO by increased levels of Ryk in mutant polyQ neurons . In support of this mechanism , Ryk-ICD overexpression was sufficient to repress the transcriptional activity of FOXO3a , a protein that promotes the survival of mutant htt striatal cells . Additionally , LIN-18 ICD expression was sufficient to suppress neuroprotection by lin-18 LOF in expanded-polyQ nematodes . This mechanism was further supported by results in mutant htt cells showing that ( i ) Ryk-ICD overexpression abolished the protective activity of β-catenin , which is consistent with the possibility that an excess of Ryk-ICD may bind and block this survival protein; ( ii ) reducing presenilin PS1 levels ( which is protective ) compensated for the cytotoxicity of full-length Ryk but not that of Ryk-ICD , implicating this γ-secretase in the toxic effects of Ryk; and ( iii ) nuclear levels of endogenous Ryk-ICD were increased compared to normal htt cells , corroborating a role for the Ryk-ICD pathway in triggering abnormal Ryk signaling in mutant htt cells . Finally , Ryk was suggested to have a pathological role in HD as emphasized by the early stage ( before or during the onset of pathology ) increase of Ryk in expanded-polyQ nematodes [7] and the neostriatum of 140CAG knock-in mice [24] . Collectively , these results suggest that Ryk and its ICD fragment may reduce the ability of mutant polyQ neurons to handle cell stress and maintain function by repressing FOXO protective activity , which may occur during the earliest phases of the pathogenic process in HD .
To explore the pathways that underlie the early phases of expanded-polyQ neurotoxicity , we performed a microarray analysis of mRNAs extracted from C . elegans touch receptor cells . To this end , we used transgenic nematodes expressing polyQ-expanded ( 128Q ) and normal ( 19Q ) N-terminal HTT fused to GFP under the control of the mec-3 promoter [15] , and transgenic nematodes expressing only GFP under the control of the same promoter as a control . In this model , expanded polyQ expression produces a strong level of neuronal dysfunction not found in normal polyQ animals , namely the loss of response to light touch [15] . GFP-positive cells were purified by cell sorting from primary cultures of embryonic cells prior to mRNA extraction and microarray analysis . Forty-one genes were deregulated in 19Q cells compared to cells expressing GFP only ( Table S1 ) . A total of 2 , 070 genes were deregulated in 128Q cells compared to 19Q cells ( Table S2 ) . Interestingly , only 18 of the 2 , 070 genes were also deregulated in 19Q nematode cells , suggesting that our microarray analysis has provided clean and specific information on the transcriptomic effects of expanded-polyQ expression . To analyze the biological content of these data , we used several methods including Gene Ontology analysis , Gene Set Enrichment Analysis ( GSEA ) , and a powerful network-based method based on Fourier analysis ( see Text S2 ) . In contrast to the GO analysis ( Figure S1 ) , the GSEA and Fourier analyses highlighted several processes previously suspected to be altered in HD ( see Text S1 , Figure S2 , Figure S3 , and Tables S3–S5 ) , suggesting that nematode data are relevant to HD pathogenesis . Additionally , cell differentiation pathways such as Wnt signaling were emphasized as components potentially involved in expanded-polyQ neuron dysfunction , a trend also emphasized by the network-based analysis of data resulting from a large-scale functional RNAi screen in our expanded-polyQ nematodes [25] . Among the up-regulated Fourier modules , module 40 ( Wnt/TGF-β signaling ) was of particular interest ( Figure S4 ) . This module suggested that lin-18/Ryk , a Wnt receptor important during neurogenesis and axon guidance [26] , [27] , is up-regulated in neurons expressing expanded polyQs , which was confirmed by qRT-PCR ( Table S5 ) . To enhance the prioritization of candidate genes , we focused on pathways and processes that were highlighted by GSEA and by the Fourier analysis and that contained evolutionary conserved druggable [28] genes ( Table S6 ) . Up-regulated module 40 was again pointed out , as it contained the largest number of genes in common with the GSEA and Fourier analyses ( see Wnt , cell cycle , and TGF-β in Table S4 ) as well as 3 of the 25 genes deregulated in 128Q nematode cells and in the caudate nucleus of HD patients [29] , among which was lin-18/Ryk . In the up-regulated Fourier module 40 ( Figure S4 ) , the conserved lin-18 gene was of interest in the Wnt pathway as a druggable gene that may be deregulated in the touch receptor cells of expanded-polyQ nematodes and caudate nucleus of HD patients . Although target gene activation is an option for developing disease-modifying strategies , target inhibition is usually regarded as a more easily achievable approach . Interestingly in this respect , neuronal dysfunction was abolished by lin-18/Ryk LOF in 128Q nematodes with no effect in 19Q nematodes and no change in transgene expression ( Figure 1A ) . These results suggested that lin-18 up-regulation is toxic to 128Q neurons and that Ryk inhibition may provide protection from mutant polyQ cytotoxicity . Having observed that neuronal dysfunction was suppressed by lin-18/Ryk LOF in 128Q nematodes , we tested if Ryk inhibition could decrease the cell death caused by full-length mutant HTT . To this end , we used striatal cells derived from the HdhQ111 knock-in mice [23] . Mutant htt ( 109Q/109Q ) striatal cells are abnormally susceptible to cell death as induced by serum deprivation [23] , a phenotype suitable for identifying modifiers of cell vulnerability [30] . As indicated by qRT-PCR and Western blotting , Ryk mRNA and protein levels are two times higher in 109/109Q cells compared to normal ( 7Q/7Q ) cells ( Figure 1B ) . To test whether reducing Ryk levels may promote mutant htt striatal cell survival , we subjected 109Q/109Q and 7Q/7Q cells to Ryk siRNA treatment . Ryk siRNA treatment enhanced the survival of 109Q/109Q cells with no effect on 7Q/7Q cells , an effect unrelated to a change in HTT expression ( Figure 1C ) . This was consistent with neuroprotection by lin-18/Ryk LOF in 128Q nematodes , further suggesting that Ryk has a pathological role in HD neurons . We next sought to examine the mechanisms underlying neuroprotection by Ryk inhibition . We first tested whether neuroprotection by lin-18/Ryk LOF in expanded-polyQ nematodes ( see Figure 1A ) occurred in a cell-autonomous manner . The expression of lin-18 cDNA in touch receptor neurons ( 4 ng/µl ) using the mec-3 promoter abolished neuroprotection by lin-18 LOF in 128Q nematodes with no effect detected in 19Q nematodes as observed in two independent arrays per polyQ genotype ( Figure 2A ) , indicating that neuroprotection by lin-18 LOF is cell autonomous . Ryk participates in canonical Wnt signaling [26] and there is a functional interaction between β-catenin , a downstream effector of canonical Wnt , and FOXO in oxidative stress signaling [9] . We thus tested whether lin-18 required bar-1/β-catenin and daf-16/FoxO activity to modulate neuronal dysfunction in 128Q nematodes . In 19Q nematodes , LOF of lin-18 ( see Figure 1A ) , daf-16 ( see [1] ) , and bar-1 ( see [10] ) had no effect on touch response . In 128Q nematodes , bar-1 LOF exacerbated the loss of touch response ( Figure 2B ) , with no change in transgenic protein expression ( Figure 2C ) , suggesting that bar-1 normally protects touch neurons from 128Q . The same applied to daf-16 LOF ( Figure 2B/C ) , as previously observed [7] . In 128Q;lin-18 ( e620 ) animals , neuroprotection by lin-18 LOF was reduced by LOF of bar-1 ( Figure 2B ) with no change in transgenic protein expression ( Figure 2C ) , suggesting that lin-18 activity requires bar-1 . However , the effect of bar-1 LOF in 128Q;lin-18 ( e620 ) nematodes was partial , suggesting a role for other genes . In contrast , neuroprotection by lin-18 LOF was suppressed by daf-16 LOF ( Figure 2B ) , an effect unrelated to a change in transgenic protein expression ( Figure 2C ) , indicating that neuroprotection by lin-18 LOF fully requires daf-16 . Given that daf-16 is normally neuroprotective ( Figure 2B ) , this indicated that lin-18 may repress daf-16 activity in 128Q nematodes , which led us to investigate the mechanisms by which Ryk may repress FOXO activity . Ryk signals through multiple mechanisms such as the canonical Wnt pathway and the nuclear translocation of its cleaved ICD to regulate neuronal differentiation [16] . This suggests that one mechanism for Ryk to repress FOXO in mutant polyQ neurons might be the deregulation of canonical Wnt , a neuroprotective pathway [11] . However , LOF of pop-1/TCF does not appear to modulate 128Q neurotoxicity in C . elegans nematodes ( unpublished data ) , suggesting that the outcome of the canonical Wnt pathway is not critical to mutant polyQ neuron survival . We thus hypothesized that another mechanism for Ryk to repress FOXO in mutant polyQ neurons could involve a noncanonical mechanism , namely the binding of the Ryk-ICD to FOXO or its partner β-catenin . It was previously shown that β-catenin associates with Ryk [31] . Consistent with these findings , we also observed that β-catenin binds to Ryk by using constructs that code for Ryk with a Myc tag at the C-terminal end or an uncleavable Ryk mutant ( Ryk: EGFRRc , a chimeric construct in which the transmembrane region was replaced with that of the EGF receptor ) [16] . These constructs were transfected into 293T cells ( derived from human kidney cells ) and cell lysates subjected to immunoprecipitation using an anti-Myc antibody followed by immunoblotting . In the cells expressing Ryk or uncleavable Ryk , immunonoprecipitation of Myc-tagged Ryk pulled down endogenous β-catenin ( Figure 3A ) , as expected . To examine if Ryk-ICD binds to β-catenin , Flag-tagged Ryk-ICD constructs were transfected into 293T cells . β-catenin was detected in the anti-Flag immunoprecipitate , suggesting that Ryk-ICD is sufficient to associate with β-catenin ( Figure 3B ) . Similar experiments aimed at testing for binding of Ryk-ICD to FOXO3a , a neuroprotective FOXO protein [32] , did not detect any interaction ( unpublished data ) . Together , these results suggested that Ryk might signal onto FOXO through binding of its ICD fragment to the FOXO co-factor β-catenin . We then sought to test whether an excess of Ryk may repress FOXO transcriptional activity . We also tested whether an excess of the γ-secretase–cleaved ICD of Ryk may have a similar effect , as the Ryk-ICD fragment associates with β-catenin ( see Figure 3B ) , a protein that promotes FOXO transcriptional activity [9] . We examined the activity of FOXO3a , a protein that is neuroprotective in models of motor neuron disease [32] , also protecting against cell death associated with mutant HTT in striatal cells from HdhQ111 knock-in mice ( Figure 4A/B ) . From here , we further examined the molecular relationships between FOXO3a , β-catenin , and Ryk in mutant polyQ toxicity . To this end , constructs encoding FOXO3a together with a Forkhead responsive element ( FHRE ) -luciferase reporter and internal control reporter were utilized to transfect mouse striatal cells that express normal HTT . Reducing β-catenin levels reduced luciferase activity compared to control cells ( Figure 4C/D ) , which is consistent with the ability of β-catenin to promote FOXO transcriptional activity in mouse cells [9] . Overexpressing Ryk reduced luciferase activity to levels comparable to those observed for β-catenin siRNA treatment ( Figure 4C ) , suggesting that Ryk up-regulation can repress FOXO3a transcriptional activity . Western blotting indicated that this effect might be attributable to full-length Ryk and to a Ryk C-terminal fragment ( Ryk CTF ) of ∼47 kDa ( Figure 4D ) . However , the CTF fragment is too large to be the γ-secretase cleavage product of Ryk ( Ryk-ICD , ∼40 kDa ) . The size of this fragment suggests that Ryk CTF contains the ICD plus the transmembrane domain and a portion of the extracellular domain , which is consistent with the possibility that other proteases might cleave the extracellular domain of Ryk near the transmembrane domain [33] . The Ryk-ICD fragment was not visible , which might be attributable to a very short half-life of this fragment similar to what has been observed for the γ-secretase–cleaved intracellular fragment of transmembrane proteins such as the amyloid precursor protein ( APP ) [34] . Interestingly , Ryk-ICD overexpression produced a similar reduction of luciferase activity compared to Ryk overexpression ( Figure 4C/D ) , indicating that Ryk-ICD is sufficient to repress FOXO3a transcriptional activity . Furthermore , overexpressing a mutant form of Ryk that cannot undergo γ-secretase cleavage ( Ryk: EGFRRc ) showed no effect on luciferase activity ( Figure 4C ) . Mutant Ryk was detected as one fragment corresponding to full-length Ryk and one fragment corresponding to Ryk CTF ( Figure 4D ) . In contrast to cells transfected with wild-type Ryk , the full-length Ryk precursor appeared to be more abundant in cells transfected with mutant Ryk ( Figure 4D ) . This suggests that blocking the γ-site of Ryk may also block Ryk CTF generation , an effect that was previously shown for APP [35] and that further points to putative similarities between Ryk and APP processing . Collectively , these results suggested that the generation of the Ryk-ICD fragment is important for the cytotoxic action of Ryk . We next tested whether the Ryk ICD has a role in mutant polyQ cytotoxicity as predicted by its ability to repress FOXO transcriptional and neuroprotective activity . In 128Q nematodes , lin-18/Ryk LOF is neuroprotective , and this effect is cell autonomous , as LIN-18 overexpression in touch receptor neurons suppresses the neuroprotective effect of lin-18 LOF ( see Figure 2A ) . Interestingly , overexpression ( at a dose of 4 ng/µl ) of the LIN-18 ICD was sufficient to suppress the protective effect of lin-18 LOF on touch response in 128Q nematodes with no effect detected in 19Q nematodes as tested in two independent arrays per polyQ genotype ( Figure 5A ) . Although overexpressing LIN-18 ICD at higher doses ( 40 ng/µl ) showed a trend towards exacerbation of 128Q cytotoxicity but did not reach statistical significance relative to control ( Figure S5 ) , overexpressing LIN-18 ICD at 40 ng/µl produced cytotoxicity in 19Q nematodes ( Figure S5 ) , suggesting that LIN-18 ICD has dose-dependent effects in polyQ;lin-18 nematodes . The overexpression ( 4 ng/µl ) of LIN-18 ICD was also sufficient to suppress the protective effect of lin-18 LOF on axonal swelling ( Figure 5B ) , accompanying the loss of touch response in 128Q nematodes [15] . Given that overexpressing the Ryk-ICD fragment is sufficient to reduce FOXO transcriptional activity ( Figure 4C/D ) , this further suggested that the Ryk-ICD fragment mediates Ryk cytotoxicity . Next , we tested whether overexpressing the Ryk-ICD fragment may be cytotoxic in striatal cells derived from HdhQ111 knock-in mice [23] and whether this involves β-catenin activity . To this end , striatal cells were transfected with either a construct coding for Myc-tagged Ryk-ICD or a construct coding for V5-tagged β-catenin , or both , and these cells were subjected to serum deprivation . The overexpression of Ryk-ICD showed no effect on normal htt cell survival ( Figure 5C/D ) . These results are consistent with the absence of toxicity for LIN-18 and LIN-18 ICD overexpression in normal-polyQ nematodes ( see Figure 2A , Figure 5A ) , supporting a model in which Ryk up-regulation represses a mechanism ( FOXO activity ) that specifically promotes mutant polyQ neuron survival . Although mutant htt cells have high endogenous Ryk levels ( see Figure 1B ) , overexpressing Ryk-ICD potentiated cell mortality ( Figure 5C/D ) . This effect was moderate , however , suggesting that Ryk toxicity is close to a maximum in these cells . Consistent with the pro-survival role of β-catenin , overexpressing β-catenin reduced the cell death of mutant htt cells with no effect detected in normal htt cells ( Figure 5C/D ) . The protective effect of β-catenin overexpression in mutant htt cells was abolished by Ryk-ICD overexpression ( Figure 5C/D ) , suggesting that Ryk-ICD may antagonize β-catenin activity . There was no significant difference between the mortality of mutant htt cells transfected with Ryk-ICD alone and that of cells co-transfected with Ryk-ICD and β-catenin ( Figure 5C/D ) . Nonetheless , cells co-transfected with Ryk-ICD and β-catenin showed similar levels of mortality compared to cells transfected with the empty vector ( Figure 5C/D ) , suggesting that β-catenin overexpression may compensate for Ryk-ICD cytotoxicity to that extent . As such , these results might just reflect the parallel activity of two proteins with antagonistic properties in the regulation of cell survival . However , given that Ryk-ICD binds to β-catenin ( see Figure 3 ) , these results supported a model in which a functional cross-talk between Ryk-ICD and β-catenin contributes to Ryk-ICD cytotoxicity . Our results in C . elegans neurons and mouse striatal cells suggest a model in which the repression of FOXO3a activity by Ryk in mutant polyQ cells is mediated by the Ryk-ICD fragment , a γ-secretase cleavage product . The γ-secretase complex has been previously implicated in HD through its role in HTT cleavage and production of N-terminal fragments CpA and CpB [36] . Here , we sought to investigate the protective role of the γ-secretase complex relative to the pathological activity of Ryk in mutant polyQ cells . To this end , we tested whether decreasing the activity of the γ-secretase complex might promote the survival of mutant htt mouse striatal cells and whether this effect might compensate for Ryk toxicity in these cells . At this point in our study , we elected to measure caspase 3/7 activity instead of counting picnotic nuclei in order to perform faster and complementary measures of striatal cell mortality . We observed that reducing the expression of presenilins PS1 or PS2 enhanced the viability of mutant htt striatal cells , with no effect detected in normal htt cells ( Figure 6A/B ) . Most importantly , reducing the levels of PS1 compensated for the cytotoxic effect of overexpressing full-length Ryk but not that of overexpressing Ryk-ICD in mutant htt cells ( Figure 6C ) . The compensatory effects of reducing PS1 levels on Ryk cytotoxicity was accompanied by an apparent increase of full-length Ryk levels as inferred from Western blot analysis ( Figure 6D ) . A decrease of the CTFs that may be generated by the sequential proteolysis of Ryk such as Ryk-CTF and Ryk-ICD [33] was also apparent , however to a lesser extent ( Figure 6D ) . Given that Ryk-ICD may result from PS1 cleavage , as previously investigated in PS1-deficients cells [16] , these results suggested that Ryk toxicity may be triggered by the production of Ryk-ICD in mutant htt cells , which led us to analyze endogenous Ryk-ICD levels in these cells . The model supported by our data for Ryk to be toxic in HD suggests that Ryk-ICD levels might be increased in mutant polyQ cells . To explore this possibility , we used a newly obtained rabbit polyclonal Ryk-ICD antibody ( anti-RykICD ) [33] to perform immunocytostaining of mouse striatal cells followed by confocal analysis . This antibody was raised against the Ryk-ICD fragment , and it recognizes Ryk species that contain the ICD including full-length Ryk and the Ryk-ICD fragment [33] . Depending on the type of cells and level of expression in these cells , this antibody may detect one or more of all possible Ryk species . The detection of Ryk species is also dependent on the methods used for expression analysis ( see Western blot in Figure S6 ) and likely to be dependent on epitope accessibility . Immunocytostaining of mouse striatal cells indicated that Ryk-ICD immunoreactivity was primarily localized in the nucleus , with little signal detected in the cytoplasm and no signal detected at the membrane ( Figure 7A/B ) . Additionally , the intensity of nuclear staining was decreased if Ryk expression is reduced ( Figure S7 ) . These results suggested that full-length Ryk might be rapidly processed to produce intracellular fragments that localize to the nucleus . Interestingly , nuclear Ryk-ICD levels appeared to be increased by about 25% in mutant htt cells compared to normal htt cells , with no change detected in the cytoplasm ( Figure 7A/B ) . This effect was observed for normal and mutant htt cells cultured on the same slides and for cell nuclei having similar sizes between genotypes . We further investigated the possibility that nuclear Ryk-ICD levels may be increased by using an internal control for improved comparison of Ryk-ICD levels between normal and mutant htt cell nuclei . We elected to use Pol2 immunostaining as a second antigen , as we observed that Pol2 levels are similar between normal and mutant htt cells . Nuclear Ryk-ICD levels were increased by about 40% in mutant htt cells in these experiments ( Figure 7C/D ) . Collectively , these observations suggested that nuclear Ryk-ICD levels may be significantly increased in mutant htt cells , corroborating the C . elegans and striatal cell data on the toxic effects of Ryk-ICD overexpression ( see Figures 5 and 6 ) and further supporting a model in which the ICD of Ryk triggers the detrimental effects of Ryk increase in mutant polyQ cells . In 128Q nematodes , lin-18/Ryk is up-regulated at an early stage of touch receptor neuron genesis ( see Figure S4 , Table S5 ) —that is , before the onset of pathology ( loss of touch response ) as detected in late larvae/young adults [7] . This led us to explore whether Ryk might be increased before or at an early stage of pathology in HD mice . Although there are a number of fragment and full-length genetic mouse models of HD , each possessing useful experimental outcomes that can be used to provide a greater understanding of the human disease process , full-length knock-in HD mice such as 140CAG mice [24] may provide the best possible genetic comparability to human HD . Western blot analysis of protein extracts from the striatum of 140CAG mice using a N-terminal Ryk antibody showed that Ryk can be detected as two bands , one band corresponding to full-length Ryk and a weaker band of ∼28 kDa ( Figure S8A ) . The weaker band likely corresponds to an extracellular fragment of Ryk that is generated by proteolytic cleavage near the transmembrane domain , illustrating again ( see Figure 4D ) the possibility that Ryk may be cleaved by several proteases [33] . Densitometric analysis of Ryk showed an increase of immunofluorescence activity in the neostriatum of 140CAG HD mice at 8 mo of age ( Figure S8B ) . To explore the chronological features of Ryk increase in the striatum of 140CAG mice , we performed immunohistochemical analysis and observed a significant age-dependent increase of Ryk levels at 2 , 4 , 6 , and 8 mo of age ( Figure S8C ) . Because running wheel and sensorimotor deficits may begin at 4 mo in the 140CAG HD mice with climbing deficits detected at 1 . 5 mo [24] , this observation suggested that Ryk might be increased prior to the onset or during the early phases of motor decline in these mice . We next explored striatal Ryk expression in HD brains . Human HD brain tissue specimens were age-matched and postmortem interval-matched . Densitometric analysis of Ryk showed a significant increase of immunofluorescence in grade 1/2 HD patients ( Figure S9A ) . In contrast to large cholinergic neurons and medium-sized NADPH-diaphorase aspiny neurons , medium-sized spiny GABAergic projection neurons of the HD striatum are affected early and most severely [37] . To further explore Ryk immunoreactivity within the striatum of HD patients , we performed double immunofluorescence for Ryk in combination with either calbindin or NOS immunoreactivities , labeling degenerating and relatively unaffected striatal neurons , respectively . Ryk co-localized with calbindin-positive striatal neurons , with no cross-reactivity with spared NOS-positive neurons ( Figure S9B ) , suggesting that Ryk immunoreactivity might correlate with the selective neuronal loss observed in HD . Additionally , we tested whether Ryk immunoreactivity might correlate with HD severity . Little positive Ryk immunoreactivity was present in the control specimens , with marked expression of Ryk immunoreactivity in all HD tissue samples in neurons , neuropil , and other cell types with increased degrees of neuropathological severity ( Figure S9C ) . Densitometric analysis showed grade-dependent increases in Ryk immunoreactivity ( Figure S9C ) . Finally , we used combined Ryk GFAP immunofluorescence to test whether Ryk might be increased in cells other than neurons for the most severe HD grade ( grade 4 ) , in which few medium spiny neurons are left in the caudate nucleus . Compared with age-matched normal specimens , Ryk immunofluorescence was increased within astroglia from caudate nucleus grade 4 HD , but not on a one-to-one basis ( Figure S10 ) . Although Ryk immunofluorescence signals were observed in the GFAP-negative area , and although there were glial figures without co-localized Ryk , Ryk immunoreactivity appeared to be predominantly from astrocytes ( Figure S10 ) , a possibility also suggested by the densitometric analysis of Ryk immunoreactivity in mouse striata ( Figure S8B ) . Collectively , these results suggested that Ryk may have discrete patterns of increased expression in striata and caudate nuclei , an aspect that will the subject of future studies that include additional Ryk antibodies and brain samples .
Wnt signaling regulates several developmental processes , including synaptic differentiation , as well as adult neurogenesis [11] . The activation of canonical and noncanonical Wnt pathways may be neuroprotective in neurodegenerative disease as indicated by studies of Aβ oligomer toxicity [11] . Studies of polyglutamine expansion [7] and Aβ toxicity [8] indicates that the same applies to the activation of the FOXO pathway , a pathway that is central to cell survival/longevity [38] as well as neuronal cell homeostasis [39] . Here , we found that Ryk , a Wnt receptor important to axon guidance [40] , [41] , is increased in mutant polyQ neurons and represses FOXO neuroprotective activity , highlighting a pathological situation in which the Wnt and FOXO pathways interact to promote neuron dysfunction and degeneration . The Ryk protein is involved in canonical Wnt to regulate neurite outgrowth , being required for TCF-driven transcription through binding to Frizzled and Dishevelled [26] . This raised the possibility that Ryk might repress FOXO activity via the alteration of the canonical Wnt/Fzd–β-catenin pathway . However , Wnt ligands such as cwn-1/WNT4 , cwn-2/WNT5 , and egl-20/WNT16 and the transcription factor pop-1/TCF did not appear to modulate 128Q neurotoxicity in C . elegans , suggesting that canonical Ryk signaling may not contribute to the toxicity of Ryk in mutant polyQ neurons . Besides its ability to signal through the planar cell polarity pathway [19] , [20] , the Ryk protein can also signal through another noncanonical pathway , namely the nuclear translocation of its ICD ( Ryk-ICD ) , a γ-secretase cleavage product that is required for cortical neurogenesis [16] . Here , we provide a model in which the repression of FOXO by Ryk in mutant polyQ neurons may primarily occur through the nuclear translocation of the Ryk-ICD fragment . Our data noticeably indicate that ( i ) the free Ryk-ICD fragment may be sufficient to repress FOXO activity in mutant polyQ cells , ( ii ) the lin-18/Ryk ICD fragment is sufficient to suppress the protective effects of lin-18 knock-out on the function of polyQ-expanded nematode neurons , and ( iii ) the free Ryk-ICD fragment may bind to β-catenin , and there is a functional cross-talk between Ryk-ICD and β-catenin in the modulation of Ryk-ICD cytotoxicity . These data support the possibility that abnormally high amounts of Ryk-ICD may prevent the activity of β-catenin , a survival protein [9] , [14] , [42] that promotes FOXO transcriptional activity [9] and protects against the early phases of mutant polyQ cytotoxicity [10] . This model is further supported by our data linking Ryk and γ-secretase PS1 in the modulation of mutant htt striatal cell viability , and by our data on the increase of endogenous Ryk-ICD levels in the nucleus of these cells . Collectively , these observations point to the Ryk-ICD fragment as an important mediator of the detrimental effects of Ryk increase . The γ-secretase complex is involved in the role of Ryk in cortical neurogenesis [16] and has been previously implicated in HD through its effect on HTT proteolysis and production of N-terminal HTT species [36] . Our data suggest that γ-secretases may also have a role in the regulation of cell stress response in HD by mediating the detrimental effects of Ryk increase on FOXO3a activity . Here , it is noticeable that aph-1 , the C . elegans homologue of APH-1B [43] , may be up-regulated by expanded polyQs , as suggested by our microarray data from touch receptor cells , which might further contribute to the toxicity of the Ryk-ICD pathway in mutant polyQ cells . Future research will build on these observations to further investigate the mechanisms that may underlie the pathological role of the Ryk-ICD/FOXO3a pathway in HD . Besides testing whether Ryk cytotoxicity might be modulated by mammalian Wnt ligands or Ryk inhibitory antibodies [33] and studying whether an excess of the Ryk-ICD fragment might tether β-catenin away from FOXO , it will be interesting to identify the FOXO3a transcriptional targets that may be deregulated by Ryk in mutant polyQ neurons and understand how this may impact brain longevity mechanisms as the pathogenic process develops in HD . The type and activity of FOXO target genes may greatly depend on the cellular context in which FOXO operates [4] , [44] , which may also be true in HD , as suggested by the study of context dependence in FOXO interaction networks across mouse and human datasets [25] . Interestingly , the transcriptomic signature of 128Q nematode touch receptor cells contains 84 genes that were previously identified as putative daf-16/FOXO targets in C . elegans , including 31 genes that are highly conserved in mice ( Table S7 ) , suggesting that FOXO target genes may be altered by mutant polyQ expression . The deregulation status of most of these 31 genes appeared to be greatly context-dependent , as evaluated by using entropy-based feature selection across 14 HD-associated conditions including murine striatum and human ( postmortem brains , blood samples , induced pluripotent stem cells ) datasets ( Table S7 ) [5] , which emphasizes the importance of context ( e . g . , cell type , time requirement ) in future analyses of the physiological impact of abnormal Ryk signaling on FOXO activity in HD . Lin-18/Ryk is up-regulated in 128Q nematode touch cells before the loss of touch response—that is , prior to pathology in C . elegans . In the neostriatum of 140CAG mice , Ryk increase is first detected during the early phases of pathology in these mice , which further links increased expression of Ryk with the early phases of HD pathology . The alteration of Ryk may also be associated with the human disease , as suggested by the grade-dependent increase of Ryk immunoreactivity in the human HD caudate nucleus . However , although the increase of Ryk detected in the neostriatum of 140CAG mice is unlikely to result from abnormal Ryk degradation or age-dependent accumulation of Ryk as it was observed in young mice , Ryk immunoreactivity in postmortem HD brains might be influenced by several factors such as alteration of cyto-architectural structure and disease-unrelated factors . Similarly to other membrane proteins , Ryk might be processed by several proteases that cleave Ryk in the vicinity of its transmembrane domain [33] , which might generate Ryk C-terminal and N-terminal fragments with distinct localization and half-life in brain tissues . Selective Ryk antibodies such as inhibitory anti-Ryk antibody RWD1 were recently described [33] , and these antibodies will be useful in additional studies of Ryk expression in the HD mouse and human HD brain . The Ryk protein is solely represented in several genomes , and reducing its pathological levels is anticipated to ameliorate neurological disease , which makes this receptor an attractive candidate for therapeutic intervention . Ryk is acutely induced in models of CNS injury and in concert with Wnt signaling proteins inhibits axon regeneration [45] , [46] , suggesting that Ryk inhibition may promote axon regeneration upon injury . Our findings identify the inhibition of Ryk-ICD signaling as a therapeutic strategy to restore cell stress response and neuronal function in HD and perhaps in other neurodegenerative diseases . The up-regulation of Ryk prior to pathology in 128Q nematodes , the up-regulation of Ryk at the mRNA and protein levels in mutant htt striatal cells derived from HdhQ111 mice , and the converging evidence for nuclear Ryk-ICD to mediate Ryk neurotoxicity by altering FOXO activity suggest that the inhibition of the Ryk-ICD pathway may have therapeutic potential in HD . In summary , our study reveals that Ryk activity is altered in mutant polyQ neurons , with Ryk up-regulation promoting neuronal dysfunction via the early-stage repression of FOXO neuroprotective activity . Regardless of a possible contribution from other mechanisms , our findings support a model in which the ICD of Ryk , a γ-secretase cleavage product , may be important for the neurotoxic action of Ryk increase by altering the protective activity of the β-catenin/FOXO complex . These data highlight a pathological process in which neurodevelopmental pathways may be altered during the early stages of the pathogenic process in neurodegenerative disease to repress cell stress response and “neuronal longevity” mechanisms such as those controlled by FOXO and its co-factors . Importantly , our data reveal that neurons may be unable to develop an efficient FOXO-mediated survival response against the earliest stages of mutant HTT toxicity . This suggests that the early-stage restoration of neuronal resistance capacity through the stimulation of cell-stress response networks and mechanisms that are under FOXO control might efficiently delay the pathogenic process in HD , which may have significant implications for the prioritization of disease-modifying strategies and identification of disease modifiers .
All the animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh . Work involving human brain tissue samples was approved by the institutional review board and the Committee for Oversight of Research Involving the Dead at the University of Pittsburgh . Nematode strains were handled using standard methods [47] . The integrated polyQ strains used in this study were previously described [15] , [48] and all assays were performed blindly . Some nematode strains used in this work were provided by the Caenorhabditis Genetics Center , which is funded by the National Institutes of Health National Center for Research Resources . The complete list of strains used in this study is shown in Table S8 . Constructs encoding LIN-18 were generated as follows . The lin-18 cDNA was obtained from wild-type animals by RT-PCR , using lin-18_attB5 ( 5′-GGGGACAACTTTGTATACAAAAGTTGATGTTCATCAGCAAAGAGGA ) and lin-18_attB2 ( 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTATTAGATGTATTGACTGAGT ) primers . These primers contain , respectively , attB5 and attB2 sequences for recombination in the pDONR221-P5-P2 vector , using the Gateway system ( Invitrogen ) . In parallel , we produced a clone , in pDONR221-P1-P5r , containing the promoter of mec-3 , mec-3p , using primers RV3 ( 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTCCTGCAGGTACCCGGAGTAGTTG ) and RV4 ( 5′-GGGGACAACTTTTGTATACAAAGTTGTGGCGCGCCAATGCGCGAAATTGTG GCTACTC ) . Both clones were used to assemble mec-3p and lin-18 , using Gateway technology , in the destination vector pDEST-AN [49] , which is suitable for C . elegans transgenesis . To produce the construct coding for the LIN-18 ICD , we used the same strategy . Briefly , we used as a reverse primer lin-18_attB2 , and a forward primer located after the sequence that encodes LIN-18 ICD , RV61 ( 5′-GGGGACAACTTTGTATACAAAAGTTGAAAATGTTCAAGCGCTCTAAAAAAGAAGA ) . Constructs were verified for sequence integrity and were then injected at 4–40 ng/µl in 19Q;lin-18 ( e620 ) and 128Q;lin-18 ( e620 ) nematodes , together with pPD118 . 33 ( a plasmid containing myo-2p::GFP ) at 10 ng/µl as a marker to follow transgenesis and pUC18 as a DNA carrier to a final total DNA concentration of 100–150 ng/µl , using standard methods . We isolated at least two independent strains from each construct based on GFP expression in the pharynx to perform touch assays and axonal swelling assays . Touch response assays were carried out as described [7] . Touch tests involved scoring for the response to light touch at the tail by using a fine hair . Touch test were performed by scoring 10 touches at the tail of the animal for a minimum of 200 animals per genotype . Ordinarily , wild-type animals will respond by backing away from the touch . The responses were recorded for every animal such that , for example , 3 responses out of 10 at the tail is given as 30% responsiveness , and the mean values for responsiveness were retained for comparison of nematode groups . Touch tests were performed by two groups of two experimenters , and maximum baseline variation was 9% . Axonal swelling was scored as previously described [7] . Briefly , 128Q nematodes were mounted on agar pads and immobilized using levamisole 20 mM , prior to examination on a 40× objective of a Leica Axioplan microscope , equipped with fluorescence . A total of more than 150 animals per strain were examined for axonal swelling in PLM neurons . Animals containing at least one swollen axon were scored as positive . For strains expressing extrachromosomal arrays , only animals expressing the transgenic marker ( i . e . , GFP expressed in the pharynx under the control of myo-2p ) were assayed . Extraction of protein from whole worms and Western blotting was conducted using standard methods [50] and the following primary antibodies: GFP antiserum ( Abcam , 1∶5 , 000 ) and actin antibody ( Molecular Probes , 1∶5 , 000 ) . Secondary antibodies used were as follows: goat–anti-rabbit IgG HRP-conjugated ( Abcam , 1∶10 , 000 ) and goat–anti-mouse IgG HRP-conjugated ( Biorad , 1∶10 , 000 ) . Proteins were detected by using ECL+ ( ECL for actin ) and evaluated by densitometry . Unpaired t tests were used for statistics . Embryonic cells were obtained as previously described [51] . Briefly , embryos were isolated from gravid adults following lysis in a hypochlorite solution . Eggshells were removed by incubation in 0 . 5 ml chitinase/chymotrypsine ( 1 U/ml and 3 , 000 U/ml , respectively , in egg buffer ) for 20 min . Following resuspension in egg buffer , the embryos were dissociated by 0 . 25% trypsin treatment for 5 min and resuspended in L-15 supplemented with antibiotics and 20% FBS ( L15-CM ) . Cells were plated on TESPA ( 4% , Sigma ) coated glass plates at a density of ∼300 , 000 cells/cm2 and maintained in L15-CM media . Cells were incubated at 20°C overnight . Wild-type ( N2 ) cells were isolated and treated similarly . The materials and methods used for FACS sorting , RNA extraction , microarray analysis , microarray data analysis , and RT-PCR are described in Text S1 . We used striatal cells homozygous for normal ( 7Q/7Q ) or mutant ( 109Q/109Q ) htt derived from HdhQ111 knock-in mice [23] and handled them as previously described [30] . Low-passage ( P9–P11 ) cell lines were used in all experiments . We used jetPEI-FluoR for transfection with cDNA , jetSI-ENDO for siRNA assays , and JetPrime for co-transfection with siRNA and cDNA as indicated by the manufacturer ( PolyPlus Transfection ) . The siRNAs ( si-Ryk , si-Foxo3a ) and scramble RNAs were obtained from QIAGEN and Operon , respectively . Mixes of 3–4 different siRNA sequences per gene ( each sequence at 25 or 33 nM ) were systematically tested for modulation of cell survival and target gene expression , followed by the evaluation of individual siRNA sequences at optimal concentration ( 100 nM ) . Effects on cell survival were considered to be reliable if two different siRNAs showed similar effects on target expression and cell survival and if the scramble RNAs ( 100 nM; unique sequence that does match with any sequence in the mouse genome ) did not show any effect . The active siRNA sequences shown in the figures are as follows: si-Ryk , 5′-GCAAATTAGTAGAAGCCAA-3′ ( 100 nM ) ; si-Foxo3a , 5′-GCTTCATGCGCGTTCAGAA-3′ ( 100 nM ) ; siPS1 , 5′-GGAGCATTCTAACGAGTGA-3′ ( 100 nM ) ; and siPS2 , 5′-CTATCAAGTCTGTGCGTTT-3′ ( 100 nM ) . The corresponding scramble RNAs shown in the figures are as follows: Ryk , 5′-GACGAAAGACCTATAATGA-3′ ( 100 nM ) ; Foxo3a , 5′-GAGCTGTACCGATGACCTT-3′ ( 100 nM ) ; PS1 , 5′-ACGTAGTCAATTCGGAGAG-3′ ( 100 nM ) ; and PS2 , 5′-GGTATTCCGTATTCGTCTA-3′ ( 100 nM ) . For transfection with cDNA constructs , we used 1 µg pcDNA3 . 1-FOXO3a-HA ( a gift from Lisa Ellerby ) , pcDNA3 . 1/nV5-DEST-β-catenin ( Addgene ) , or pcDNA3 . 1-Myc-Ryk ICD . The Ryk-ICD fragment ( amino acids 239 to 595 ) was amplified from mouse cell RNAs by RT-PCR using the Ryk-ICD att-B1 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTATCAGACTGAGGCTCCCAGGGCAG-3′ and Ryk-ICD att-B2 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTGCCGCCACCATGGAACAAAAACTTATTTCTGAAGAAGATCTGAAAAGGATTGAACTGGATGACG-3′ primers . The reverse primer contained a sequence ( italicized ) coding for a Myc tag . PCR products were subcloned into the pDONR221 vector and then into the final destination vector pcDNA3 . 1 using Gateway technology and verified for sequence integrity . Cell mortality assays were performed as described previously by counting picnotic nuclei [52] or measuring caspase 3/7 activity [53] . Briefly , low passage number ( P9–P12 ) 7Q/7Q and 109Q/109Q cells were subjected to a 24-h serum deprivation 48 h after cell transfection . Cells were then fixed and subjected to DAPI staining , and cell mortality was scored by counting picnotic versus normal nuclei in DAPI- and jetPEI-FluoR– , or JetSI-ENDO–positive cells . Alternatively , caspase assays were used and transfected cells were plated on collagen-coated 96-well plates for 48 h . After 24-h serum starvation , the activation of caspase 3/7 was measured in cells using the Apo 3/7 HTS High Throughput Screen Assay kit ( Cell Technology ) . The activity of caspase 3/7 was measured using a Tecan infinite F500 microplate reader , with excitation and emission wavelengths of 485 and 535 nm , respectively . Caspase assays were performed using six replicates per point and data expressed as dRFU/min/mg of protein . For qRT-PCR analysis , we used the forward 5′-TGAGAGCTGACACACCCAA and reverse 5′-CACTTCGCAAGTCGTTCTTC primers for amplification of Ryk mRNAs and forward 5′-TTTGCCGCGAGCCG and reverse 5′-TAACCTGGTTCATCATCGCTAATC primers for amplification of HPRT mRNAs as a control . For Western blotting , proteins were extracted as previously described [30] , separated by SDS-PAGE , and analyzed by Western blotting using the following primary antibodies: mouse anti-HTT ( 4C8 , Chemicon , 1∶5 , 000 ) , rabbit anti-RYK ( Abgent , 1∶100 ) , mouse anti-FOXO3a ( Cell Signaling , 1∶1 , 000 ) , rabbit anti-PS1 and rabbit anti-PS2 ( Cell Signaling , 1∶1 , 000 ) , rabbit anti-Myc tag ( Cell Signaling , 1∶1 , 000 ) , mouse anti-V5 tag ( InVitrogen , 1∶2 , 000 ) , and mouse anti-actin ( MP Biomedicals , 1∶5 , 000 ) . Secondary antibodies used were as follows: goat–anti-mouse IgG HRP-conjugated ( Biorad , 1∶10 , 000 ) and goat–anti-rabbit IgG HRP-conjugated ( Biorad , 1∶10 , 000 ) . Proteins were detected by using ECL+ ( ECL for actin ) and evaluated by densitometry . Statistical analyses were performed using unpaired t tests . To test for interaction between Ryk and β-catenin , transfection of 293T cells with wild-type Myc-tagged Ryk , Flag-tagged RYK ICD , or an uncleavable Ryk mutant ( Ryk , EGFRRc ) [16] was performed using a calcium phosphate precipitation method [54] . Cells were lysed in a lysis buffer containing 25 mM Tris-HCl , pH 7 . 4 , 150 mM NaCl , 5 mM EDTA , 1% Triton X-100 , 10 mM sodium pyrophosphate , 10 mM β-glycerophosphate , 1 mM sodium orthovanadate , 10% glycerol , and protease inhibitors ( Roche ) . For immunoprecipitation , cell lysates were incubated with a specific antibody for 2 h at 4°C and then with Protein A/G agarose beads ( Pierce ) overnight . Immunoprecipitates were eluted using SDS sample buffer and separated using 8% or 10% SDS-PAGE . After blocking , the blots were incubated with a primary antibody and then with a peroxidase-conjugated secondary antibody . The bound secondary antibody was then detected using enhanced chemiluminescence ( ECL ) reagent ( Santa Cruz Biotechnology ) . To test for an interaction between Ryk ICD and FOXO3a , mouse striatal cells were transfected with Foxo3a-HA and Myc-tagged Ryk-ICD using Jet PEI ( PolyPlus Transfection ) as described by the manufacturer . Cells were lysed in PBS 0 . 1% tween-20 supplemented with PMSF and protease inhibitors ( Roche ) . For immunoprecipitation , cell lysates were incubated with a specific antibody for 2 h at 4°C and then with fastflow Protein A agarose beads ( Sigma ) overnight at 4°C . Immunoprecipitates were eluted using SDS sample buffer and separated using 4%–12% SDS-PAGE ( InVitrogen ) . The Ryk-ICD fragment ( amino acids 239 to 595 ) construct used in these experiments was the same as above ( see striatal cell vulnerability assays ) . Normal htt mouse striatal cells derived from HdhQ111 mice and used at low passage numbers ( P9–P12 ) were seeded in 96-well plates , at a density of 25 , 000 cells/well , and were co-transfected using Amaxa technology with 0 . 5 µg of the plasmid ( s ) pcDNA3 . 1-Foxo3a-HA ( a gift from Lisa Ellerby ) together with pcDNA3 . 1-Myc-Ryk ( see below ) , pcDNA3 . 1-Myc-Ryk ICD ( see above ) , pcDNA3 . 1-Myc-uncleavable-Ryk ( see above ) , 100 nM of β-catenin siRNAs ( QIAGEN ) or scramble RNAs ( Operon ) , 0 . 5 µg of the luciferase reporter ( FHRE-luc , Addgene ) [55] , which contains three canonical FOXO binding sites , and 50 ng of the Renilla luciferase construct ( Promega ) . The pcDNA3 . 1-Myc-Ryk construct was generated as follows: the Myc-Ryk coding sequence was amplified from pCMV6-Myc-Ryk ( Origene ) by using the primers forward 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTGCCGCCACCATGCGCGCGGGCCGGGGC and reverse 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTATCAGACGTAGGCTCCCAGGGCAG and subcloned into pcDNA3 . 1 , and the resulting plasmid was verified for sequence integrity . The pmaxGFP construct ( Amaxa ) was used as a control for cotransfection . Transfection efficiency was greater than 80% . One day after transfection , cells were lysed in 20 µl of lysis buffer and the luciferase and Renilla luciferase activities assayed using Stop & Glo reagents ( Promega ) according to the manufacturer's protocol . For the β-catenin siRNA experiments , effects were considered to be reliable if two different siRNAs showed similar effects on β-catenin expression and luciferase activity and if a scramble RNA did not show any effect . The active β-catenin siRNA and the scramble RNA that are shown in the figures are 5′-GATAGAAATGGTCCGATTA-3′ and 5′-GTGTGAATGCATGAAACTA-3′ , respectively . Endogenous Ryk-ICD levels were analyzed in 7Q/7Q and 109/109Q mouse striatal cells [23] by using the rabbit Ryk-ICD antibody anti-RykICD [33] . Low-passage ( P9–P11 ) cell lines were grown in normal conditions ( no serum starvation ) and treated with Ryk siRNAs or scramble RNAs as described above ( see striatal cell mortality assays ) using Lab-Tek eight-chamber glass slides ( BD Biosciences ) . For comparison of Ryk-ICD levels between 7Q/7Q and 109/109Q cells , cells were grown on the same slides . Cells were fixed and permeabilized using the Cytofix/Cytoperm Kit ( BD Biosciences ) following the manufacturer's protocol . Cells were then incubated with anti-RykICD ( 1∶100 ) followed by incubation with the anti-rabbit secondary antibody Alexa Fluor 488 ( Invitrogen , 1∶500 ) and subjected to diamidino-2-phenylindole ( DAPI ) staining . Alternatively , cells were co-incubated with anti-RykICD ( 1∶100 ) and the Pol2 antibody 7C2 ( 1∶500 ) followed by co-incubation with the anti-rabbit secondary antibody Alexa Fluor 555 ( Invitrogen , 1∶500 ) and the anti-mouse secondary antibody Alexa Fluor 488 ( 1∶500 ) , and cells were then subjected to DAPI staining . Fluorescent signals were quantified using a confocal microscope ( Leica TCS SP5 ) and images analyzed using ImageJ . For each of the focal planes , anti-RykICD signals were quantified from the nucleus ( DAPI staining ) . The comparison of 7Q/7Q and 109/109Q cells was based on cell nuclei that have a size of 150–250 pixels in each of the cell lines . To assess cytoplasmic expression , nuclear anti-RykICD signals were blackened , and the remaining signals were quantified . Cytoplasmic analysis was performed only if nuclear anti-RykICD signals were detected in the same confocal plane and cell . The methods used for immunocytochemistry and fluorescent immunocytochemistry analyses of mouse tissue samples and human tissue specimens are described in Text S2 . Western blot membranes were exposed to films . After exposure , films were scanned at high resolution ( 1 , 200 dpi ) using a regular scanner ( Epson ) , and scans ( tiff files ) were opened into Photoshop and Western blot images processed using PowerPoint and Photoshop ( Microsoft ) for constructing final figures . Alternatively , Western blot membranes were scanned at high resolution ( 692 dpi ) using a luminescent image analyzer ( FUJIFILM LAS-4000 ) , and scans ( tiff files ) were opened into ImageJ and Western blot images processed as previously mentioned for constructing final figures . Statistical analysis of variance ( ANOVA ) , t tests , and Welch's t tests were performed using Prism . One-way ANOVA was followed by correction for multiple testing by Tukey's Multiple Comparison Test . The statistical analysis of microarray data and statistical methods used for biological annotations are described in Text S2 .
|
Neuronal cell decline in neurodegenerative disease can be caused by inherited mutations and involves neuronal dysfunction followed by neuronal death . The ability of neurons to cope with the chronic stress induced by mutant protein expression may determine the course of their decline and eventual demise . Although the pathophysiological importance of these stress responses has been previously shown , very little is known about the signaling networks that regulate neuronal homeostasis during the early presymptomatic—but pathogenic—phases of a neurodegenerative disorder such as Huntington's disease ( HD ) . In particular , it remains unclear whether neuronal differentiation factors regulate stress response pathways during neurodegenerative disease and how this might impact the overall capacity of neurons to cope with stress and maintain their function . Here , we show that the Wnt receptor Ryk , a protein known to be important for neurogenesis , is increased in different animal models of HD , before or during the early phases of the disease process . Interestingly , increased levels of Ryk repress activity of the FOXO proteins—a family of transcription factors that play a role in cell survival/longevity and in neuronal homeostasis and protection . Ryk represses FOXO protective activity , possibly directly , through its intracellular domain , a product of γ-secretase–mediated cleavage previously implicated in the birth of new cortical neurons . This highlights the regulation of HD neuron survival by a Ryk-dependent pathway that is distinct from canonical Wnt/Ryk signaling . From our findings , we postulate that neurons are unable to develop an efficient FOXO-mediated survival response during the very early , pathogenic phases of HD .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"models",
"neurobiology",
"of",
"disease",
"and",
"regeneration",
"medicine",
"and",
"health",
"sciences",
"model",
"organisms",
"neurology",
"biology",
"and",
"life",
"sciences",
"neuroscience",
"research",
"and",
"analysis",
"methods"
] |
2014
|
The Wnt Receptor Ryk Reduces Neuronal and Cell Survival Capacity by Repressing FOXO Activity During the Early Phases of Mutant Huntingtin Pathogenicity
|
Chagas disease , considered a neglected disease by the World Health Organization , is caused by the protozoan parasite Trypanosoma cruzi , and transmitted by >140 triatomine species across the Americas . In Central America , the main vector is Triatoma dimidiata , an opportunistic blood meal feeder inhabiting both domestic and sylvatic ecotopes . Given the diversity of interacting biological agents involved in the epidemiology of Chagas disease , having simultaneous information on the dynamics of the parasite , vector , the gut microbiome of the vector , and the blood meal source would facilitate identifying key biotic factors associated with the risk of T . cruzi transmission . In this study , we developed a RADseq-based analysis pipeline to study mixed-species DNA extracted from T . dimidiata abdomens . To evaluate the efficacy of the method across spatial scales , we used a nested spatial sampling design that spanned from individual villages within Guatemala to major biogeographic regions of Central America . Information from each biotic source was distinguished with bioinformatics tools and used to evaluate the prevalence of T . cruzi infection and predominant Discrete Typing Units ( DTUs ) in the region , the population genetic structure of T . dimidiata , gut microbial diversity , and the blood meal history . An average of 3 . 25 million reads per specimen were obtained , with approximately 1% assigned to the parasite , 20% to the vector , 11% to bacteria , and 4% to putative blood meals . Using a total of 6 , 405 T . cruzi SNPs , we detected nine infected vectors harboring two distinct DTUs: TcI and a second unidentified strain , possibly TcIV . Vector specimens were sufficiently variable for population genomic analyses , with a total of 25 , 710 T . dimidiata SNPs across all samples that were sufficient to detect geographic genetic structure at both local and regional scales . We observed a diverse microbiotic community , with significantly higher bacterial species richness in infected T . dimidiata abdomens than those that were not infected . Unifrac analysis suggests a common assemblage of bacteria associated with infection , which co-occurs with the typical gut microbial community derived from the local environment . We identified vertebrate blood meals from five T . dimidiata abdomens , including chicken , dog , duck and human; however , additional detection methods would be necessary to confidently identify blood meal sources from most specimens . Overall , our study shows this method is effective for simultaneously generating genetic data on vectors and their associated parasites , along with ecological information on feeding patterns and microbial interactions that may be followed up with complementary approaches such as PCR-based parasite detection , 18S eukaryotic and 16S bacterial barcoding .
Chagas disease ( American trypanosomiasis ) is caused by the protozoan parasite Trypanosoma cruzi . Considered a neglected disease by the World Health Organization , it is widespread in the Americas , where an estimated 70 million people are at risk of contracting the infection [1] . The disease is most prominent in poor , rural communities of South and Central America , where the disruption of sylvatic ecosystems and precarious socioeconomic conditions aid the establishment of domestic and peridomestic vector populations [1 , 2 , 3] . The infective agent , Trypanosoma cruzi , is genetically diverse and widely dispersed in the Americas [4 , 5 , 6 , 7] . Multiple strains are distributed from the southern United States to northern Argentina , and are ancestrally linked to sylvatic and/or domestic transmission cycles depending on their habitat affiliation [4 , 8 , 9] . From an epidemiological standpoint , T . cruzi sensu lato ( s . l . ) is the most important group of parasitic trypanosomes strains , comprising T . cruzi cruzi , which causes Chagas disease in humans , and T . cruzi marinkellei , a strain uniquely found in South American bats [5 , 10 , 11] . Within T . c . cruzi , seven Discrete Typing Units ( DTUs ) have been characterized ( TcI-VI and TcBat ) [4 , 11 , 12 , 13] . All DTUs can cause disease in humans; however , their relative abundance varies among ecological and geographical niches , and they show variation in clinical epidemiology and prevalence in domestic ecotopes [12] . TcI is the predominant DTU in the Americas , found in arboreal Rhodnius species from Central America to Ecuador , and in sylvatic and domestic Triatoma from the southern United States to northern Argentina [4 , 5 , 13] . It is also reported in other Triatominae genera such as Meccus , Mepraia and Panstrongylus , and its genetic diversity is consistent with its long evolution in the continent , dating between 3–4 MYA [4 , 14] . TcIV , a DTU hypothesized as an ancestral hybrid between TcI and TcII , is the only other DTU that has been detected in vector and human specimens in Central America [4 , 13 , 15] . Although there are 84 reports of humans infected with TcIV from six countries , there is evidence that this DTU is of sylvatic origin and exclusively associated with sylvatic vectors [4] . In addition to T . cruzi diversity , the genetic structure of the vector , driven by geographical and ecological factors , is also likely to play an important role in determining human infections . To date , more than 140 species of New World triatomines have been described [16 , 17 , 18] and a small number of species have been reported from Asia . The majority are associated with sylvatic habitats , but species such as Triatoma infestans and Rhodnius prolixus have adapted to domestic and peridomestic niches [7 , 16 , 19 , 20 , 21 , 22 , 23] . Furthermore , species like T . dimidiata are in the process of domiciliation , establishing multi-generational colonies in human households , therefore increasing the risk of T . cruzi transmission to humans [23] . In Central America , R . prolixus was the predominant Chagas disease vector until successful eradication of the vector in 2010 [21] . In its place , endemic triatomines including T . dimidiata have colonized vacant peridomestic and domestic habitat niches and have slowly changed the dynamics of disease transmission in these ecotopes [24 , 25 , 26 , 27 , 28] . Triatoma dimidiata is widely distributed from Mexico to Perú in sylvatic , peridomestic and domestic habitats [26 , 29 , 30] . It is morphologically highly variable across this range , with phenotypic variation among sylvatic and domestic ecotopes , as well as geographical niches [23 , 30] . Population genetic analyses using various molecular markers have yielded conflicting assessments of the extent and importance of genetic structuring across its geographical distribution; nevertheless , most studies agree that it is genetically diverse [17 , 24 , 26 , 27 , 29 , 31] . The microbial community colonizing the vector’s gut may further influence parasite transmission to vertebrate hosts . When the parasite is ingested in a blood meal , the parasite moves into the midgut , where availability of glucose moderates its transformation to replicative epimastigotes [32 , 33] . In the midgut , the parasite attaches to the cuticle wall prior to differentiating into a metacyclic form [33] . Although the composition and physiological role of gut bacteria in triatomines are largely unknown , bacterial communities can significantly modify glucose levels in anaerobic environments such as the gut , facilitating or impeding colonization of the insect’s digestive tract by pathogens such as T . cruzi [34 , 35 , 36 , 37] . Some bacterial species have been shown to directly inhibit colonization by T . cruzi in Triatoma and Rhodnius spp . ( e . g . , S . marecescens ) [35 , 38] , either in their native form , or as introduced transgenics in the gut of triatomines under laboratory conditions [39 , 40 , 41] . At the same time , T . cruzi infection may be capable of decreasing the microbial population in the gut and modifying the nitrite/nitrate production important for triggering defense metabolic cascades [42] . As a vector-borne disease , domestic and sylvatic transmission cycles are dependent on the diversity and availability of vertebrates , both as blood meals for the vector and as potential hosts [43] . Trypanosoma cruzi is most commonly transmitted to mammalian hosts via contamination of a wound or mucous membrane by the parasite-contaminated feces of the vector , and/or by direct ingestion of an infected insect [5 , 33] . In domestic ecotopes , humans and dogs are presumed to serve as both the primary blood meals of the vector and the main mammalian source of the parasite; however , there are numerous peridomestic hosts ( e . g . small ruminants , rodents , pigs ) that may be important contributors to disease recurrence [3 , 16 , 20 , 25 , 44 , 45] . Accidental introduction of the vector into or near houses may happen through movement of human belongings like clothes or blankets , movement of chickens carrying early instar nymphs or transportation of infested wood or palm leaves [16 , 44] . In addition , local wildlife populations in peridomestic or sylvatic environments , such as bats , rodents and opossums , may serve as parasite reservoirs [20 , 25 , 46] . Given the diversity of interacting biotic elements involved in the epidemiology of Chagas disease , having simultaneous information on parasites , vectors , gut fauna and hosts would facilitate identifying how they interact to influence disease risk . Although genetic studies are typically focused on a single target organism at a time , reduced representation sequencing methods such as Restriction-site Associated DNA sequencing ( RADseq ) provide an affordable way to simultaneously sequence mixed-DNA specimens without relying on taxon-specific primers or probes [47] . When combined with a bioinformatics pipeline designed to identify and assign sequences back to their taxonomic source , such approaches may be ideally suited to explore complex , multi-factorial systems such as T . cruzi transmission cycles [48 , 49] . RADseq also typically generates sufficient SNP loci to resolve relationships across multiple spatial and temporal scales , allowing a uniform protocol for producing data that can be meaningfully compared across studies [50 , 51] . Although RADseq has been used to assess the population genomics of individual disease vectors ( e . g . , Anopheles spp . , [52]; Aedes aegypti , [53] ) , it has not yet been reported for mixed-species analyses . In this study , we develop a RADseq-based analysis pipeline for analyzing mixed-species DNA derived from T . dimidiata abdominal DNA . The ideal approach would be cost-effective , feasible with samples of varying age and quality , and capable of resolving vector and parasite population processes across spatial scales , from within-village dispersal to broad biogeographic and ecological differentiation . To evaluate whether the method was effective across this spatial range , we used a nested spatial sampling design for T . dimidiata , starting with multiple insects within and among individual villages , to samples collected from increasingly greater distances across major biogeographic regions in Central America . Sample results helped determine the utility of RADseq genotyping for simultaneous assessment of: ( 1 ) the prevalence of T . cruzi infection in the vector and its phylogenetic characterization in the region , ( 2 ) the population genetic structure of T . dimidiata , ( 3 ) the gut microbial community structure associated with T . cruzi infection of the vector , and ( 4 ) the blood meal history of the vector . We demonstrate that the method can effectively separate genomic information of parasite , vector , microbiome and blood meal , even without a sequenced genome for T . dimidiata .
Sixty-one adult T . dimidiata were collected by the Laboratorio de Entomoligía Aplicada y Parasitología ( LENAP ) at San Carlos University of Guatemala and the Centro de Investigación y Desarrollo en Salud ( CENSALUD ) at Universidad de El Salvador from 1999 to 2013 , representing a range of age and preservation conditions for evaluating the effect of specimen quality on sequencing yield ( Table 1 ) . Specimens were captured alive in domestic environments , transferred to a laboratory setting for microscopic examination for T . cruzi and placed in vials containing 95% ethanol + 5% glycerol within two days of capture . The exceptions were the specimens from the towns of El Chaperno and El Carrizal , collected in 2012 and 2013 ( Table 1 ) , which were examined by microscopy and placed in 95% ethanol ( no glycerol ) within a few hours of collection . To assess infection status , the abdomen of each insect was compressed to obtain fecal droplets that were diluted with 1 drop of saline solution and examined by a trained observer under the microscope at 220–400X for 5 minutes for active trypanosomes . The specimens placed in ethanol + glycerol were stored at room temperature at LENAP until being transported to Loyola University New Orleans or the University of Vermont in 2012 and 2013 , respectively . Once in the United States , the insects were stored at -20°C until DNA was extracted for sequencing . Specimens from El Chaperno and El Carrizal were stored in ethanol at room temperature for less than one week before being transported to University of Vermont , where they were maintained at -20°C . To measure the spatial resolution at which RADseq markers are able to resolve the genetic structure of T . dimidiata and T . cruzi , three nested geographical spatial scales of sampling were selected: a ) individual villages , including five in the neighboring regions of Chiquimula , Jutiapa , and Santa Ana; b ) within-country regions , including three in Guatemala , and one in El Salvador; and c ) countries across Central America , including Guatemala , Belize , El Salvador and Nicaragua ( Table 1 , Fig 1 ) . We extracted DNA from the 61 specimens from the three posterior segments of the abdomen or four surface-sterilized legs ( Table 1 ) ; the latter included the attached muscle , and served as “insect-only” controls . Tissues were flash-frozen by submerging the vials in liquid nitrogen , manually homogenized using sterilized pestles and DNA extracted using a modified Qiagen DNeasy ( Burlington , Vermont ) tissue extraction protocol . Modifications included an overnight Proteinase K digestion at 56°C , followed by an RNAse digestion at 37°C for 30 minutes using 1 . 5 uL of 4mg/mL RNAse to reduce RNA contamination . DNA was quantified using a Qubit spectrophotometer ( Burlington , Vermont ) , and quality was assessed by electrophoresis on a 1 . 5% agarose gel stained with ethidium bromide . Only specimens with a minimum yield of 1 , 000 ng of DNA and a single , high-molecular weight band were considered suitable for sequencing; of the original 61 specimens , 32 ( 20 abdomens and 12 legs ) met these minimal requirements . To verify the reproducibility of the retrieved genetic markers ( SNPs ) , for one insect specimen we included high-quality DNA isolated from two different body parts ( abdomen and leg tissue , JUCA-02A and JUCA-02L; Table 1 ) . RADseq library preparation was conducted using the restriction enzyme SbfI ( 8-base cutter: 5′—CCTGCA↓GG—3′ , 3′—GG↓ACGTCC—5′ ) at Floragenex ( Portland , Oregon ) following the methods of Baird et al . [47] . RAD libraries were barcoded by individual , and multiplexed in a 24-plex format on an Illumina GAIIx / HiSeq Analyzer . The raw sequencing reads were 100 bp in length , including the inline 5-bp barcode and 8-baseSbfI recognition sequences . We used FastX-trimmer in the FastX-toolkit to remove the barcodes , recognition sites , and FastQ-quality-filter to remove sequences with any base having a confidence score below 10 [54] . The DNA recovered from a T . dimidiata abdomen represents a mixture of DNA from the parasite ( if present ) , the insect vector , possibly one or more vertebrate blood meals , and the microbial community residing in the gut , internal tissues and on the cuticle . We designed a custom bioinformatics pipeline to separate these DNA sources and analyzed them individually for either SNP genotypes ( T . dimidiata , T . cruzi ) or taxonomic identification ( blood meal , microbes ) ( Fig 2 ) . We mapped the trimmed sequences from all 32 specimens against six T . cruzi reference genomes downloaded from the NCBI genome database ( May , 2016 ) using Bowtie 1 . 1 . 2 [55] . These included a subset of DTUs: two representatives of TcI ( ACCN: AODP01000000 , ADWP02000000 ) , one of TcII ( ACCN: ANOX01000000 ) , and two of TcVI ( ACCN: AAHK01000000 , AQHO01000000 ) . We also included T . cruzi marinkellei ( ACCN: AHKC01000000 ) , which served as the phylogenetic out-group . The 12 samples of T . dimidiata leg tissue were also mapped to the T . cruzi genomes in order to filter out any possible T . cruzi contamination from handling , with only the unmapped reads from this step used in downstream analyses . Mapping success was negligible ( < 8 reads ) for all of the leg samples . Because there is no sequenced genome for T . dimidiata , we used the sequences derived from leg tissue to assemble a reference set of RAD-tags most likely to be derived from the T . dimidiata genome . Using the 12 legs , we used the denovo_map pipeline in Stacks to obtain a putative set of T . dimidiata loci [56] ( Fig 2 ) . The parameters of the alignment were set at 3X depth of coverage for the initial stack , with a maximum of two mismatches among trimmed sequences of a single individual . Once the first stack was formed with primary reads that met the parameters , we allowed a maximum of 4 mismatches when aligning the secondary reads ( those reads that did not meet the cut-off to align in the first stack ) , and a maximum of 3 mismatches per nucleotide across both the primary and secondary reads [56] . Once the alignment yielded a raw catalog , tags were retained if: ( a ) at least half of the specimens had a read for the locus , ( b ) there were between 0 and 3 SNPs present across the reference sequences and ( c ) there were no more than two haplotypes for any individual specimen at the locus . A total of 6206 loci fitting these criteria were used as a custom index in Bowtie against which all 32 specimens were mapped to obtain individual , vector-specific reads ( Fig 2 ) . SNP genotypes for both T . cruzi and T . dimidiata were called using the Stacks ref_map pipeline [56] . Because the number of reads retrieved for the vector were an order of magnitude higher than for the parasite ( see Results ) , we set the parameters for the vector to a maximum of six mismatches between loci and a depth of coverage of 3X , while for the parasite we also allowed up to 6 mismatches but retained calls at 1X depth of coverage . We excluded any locus with missing data in at least 18 of the 32 specimens for T . dimidiata and 10 of the 13 T . cruzi-positive specimens for T . cruzi . With the remaining unmapped reads , we ran a BLAST search query of the nt database for potential blood-meal sources and microbiota , using an e-value cutoff of 0 . 001 , a query coverage minimum of 85 bp ( 97% ) , and only retaining the top hit that mapped to each sequence ( Fig 2 ) . Exploratory mapping to other databases ( e . g . , RefSeq ) yielded fewer hits than the nt database and were not included in the final pipeline . When the sequence mapped equally well to multiple taxa , the first species returned by the BLAST algorithm was retained; although species identity in such cases was not well supported , identification was consistent across all reads of identical sequence within and among specimens . Information on the mean e-value cutoffs by taxonomic group is provided as S1 Table . Because genomic reference sequences were available for only three of the six DTUs , two approaches were used to assign a putative DTU to the T . cruzi-positive specimens . First , we identified the total set of reads for each specimen that mapped successfully to any one or more of the T . cruzi reference genomes and then mapped this set of reads to each genome individually to determine relative mapping success . For comparison , we generated in-silico RAD-tags from the six reference genomes using a custom python script that identified all occurrences of the restriction enzyme recognition sequence in the genome and retrieved the 87 bp directly up- and down-stream of the cut site . These were mapped against each of the six reference genomes using the same Bowtie protocol as with the field specimen data to obtain expected mapping success for a given DTU . Two main patterns of mapping success were found across the entire DNA specimen set ( see Results ) ; for each distinct subset , we ran one-way ANOVA and a post-hoc Tukey’s range test using the stats package in R [57] to test whether the mapping success was biased toward a particular reference genome . Second , we used the SNP genotypes generated with Stacks to reconstruct phylogenetic relationships among the in-silico genomes and the field specimens with MEGA version 7 , using Maximum Likelihood with a nucleotide p-distance substitution model and 10 , 000 bootstrap permutations [58] . To infer the population genetic structure of T . dimidiata , we performed a k-means clustering analysis , and classified the individuals by a discriminant analysis of principal components ( DAPC ) using the Adegenet package for R [59] . To prevent biases associated with missing data , specimens with >50% missing SNPs were excluded from the analysis ( i . e . , CHGU-01 and CHCE-01 ) ; one additional specimen ( UnID ) did not have precise geo-location information and was also excluded . Using the 29 remaining specimens , we identified the best number of genetic clusters using the k-means cluster algorithm from the find . clusters function in Adegenet and selected the value of k that minimized the Bayesian Information Criterion ( BIC ) value , setting the maximum number of potential clusters to 16 , and retaining a total of 25 principal components based on the cumulative variance explained by the eigenvalues . We also calculated the fixation index ( Fst ) , nucleotide diversity ( pi ) , observed ( Hetob ) and expected ( Hetex ) heterozygosity among clusters using the Populations function in Stacks [56] . To compare bacterial species richness across specimen types ( infected abdomens , non-infected abdomens and legs ) , we used the rarefaction function in the Vegan package in R to estimate asymptotic species richness for each specimen [60 , 61] . Specimen types were compared using an ANOVA with post-hoc Tukey’s pairwise comparisons in the R Stats package . To compare gut bacterial community composition as a function of infection status , we ran a non-metric multidimensional scaling ( NMDS ) weighted Unifrac ordination analysis with the default number of dimensions ( k = 2 ) using the phyloseq package in R [62] . Because they do not contain gut tissue , leg specimens were excluded from this analysis . Bacterial phylogenetic relationships were retrieved from the SILVA 123 ribosomal living tree , pruned to the set of taxa present in the specimens using the prunedTree function in the Picante package [63 , 64 65] . The matrix of counts is available in S3 Table . To assess significance of clusters , we performed a post-hoc permutation analysis of 999 repetitions embedded in the NMDS function . To distinguish actual vertebrate blood meals from possible contamination due to handling and/or false-positive BLAST hits from multiple taxon matches , we identified the chordate species identified by the largest number of sequencing reads ( the "top-hit" species ) for each specimen . Representation of the top-hit species within a specimen was expressed as a percent of the total possible hits ( i . e . , the total number of reads that had not mapped to either the parasite or vector ) . Leg specimens were used to determine the expected background representation of chordate hits . Putative blood meals were called for those specimens with a top-hit representation statistically above the background , identified with an outlier test using the Tukey boxplot method for skewed data [66] , with the upper outlier threshold defined by the Tukey range of Q3+1 . 5*IQR , the Inter-Quartile Range ( S4 Table ) .
We obtained a total of 164 . 1 million unfiltered reads across all specimens . There was no difference in the number of raw reads between leg and abdomen , or among specimens obtained in different collection years . After quality filtering , 70 . 69% of reads were retained , with an average of 3 . 25 million reads per specimen ( ± 652 , 000 ) . Analysis with the mixed-species pipeline produced subsets of reads corresponding to all of the expected taxonomic groups ( parasite , vector , blood meal and bacteria ) ( Fig 3 ) . Although the majority of reads ( 63% ) could not be assigned to a particular source , both the vector ( 20% ) and the parasite when present ( 1% ) were represented by sufficient mapped reads to approach saturation of SNP recovery ( Figs 3a , 4a and 4b ) . In our internal control ( Table 1 ) , the leg specimen ( JUCA-03 ) was over-represented compared to the abdomen ( JUCA-02 ) from the same insect , yielding 60 . 8% more trimmed reads than the abdomen . This difference affected the number of mapped reads ( 37 . 87% higher ) , mean depth of coverage ( 222 . 8X for leg versus 98X for abdomen; Fig 4c ) , and number of called SNPs ( 19% higher ) ; however , for the 15 , 611 loci called across both genotypes , only eight ( 0 . 05% ) were different between the two tissue types . Thirteen of the 20 abdomens mapped to at least one of the six available T . cruzi reference genomes; however , four of these specimens yielded fewer than 100 mapped reads , with no polymorphic loci ( Fig 5 ) . These specimens were omitted from further T . cruzi analysis . Eight of the 12 leg specimens did not map to any of the T . cruzi genomes , while four legs mapped to at least one genome with a range of 1–7 reads and no polymorphic loci . The nine T . cruzi-positive abdomens yielded an average of 150 , 994 ±118 , 089 mapped reads , corresponding to 6 , 377 unique genomic locations , with a total of 6 , 405 SNPs ( Fig 5 ) . The median depth of coverage was 8 . 7X , ranging from 4 . 7X to 181 . 9X; there was no relationship between the mean depth of coverage and the number of SNP genotypes successfully called per specimen ( Fig 4D ) . Detection of infection status via fecal microscopy and RADseq were significantly associated ( Fisher’s Exact test , p = 0 . 0018 ) ( Fig 5 ) . All six specimens positive for T . cruzi by microscopy were also positive by RADseq . Seven additional T . cruzi-positive specimens were detected by RADseq but not by microscopy , including three with high read abundance and the four that yielded <100 reads . Among the positive specimens identified solely by this method , the abdomen internal control , JUCA-02A , yielded a total of 8 , 610 T . cruzi reads . In contrast , the leg control extracted from the same insect , JUCA-02L , yielded only 7 T . cruzi reads . Genome mapping comparisons indicated that the nine T . cruzi isolates from the T . dimidiata abdomens included two distinct parasite DTUs ( Table 2 ) . Patterns of mapping success fell into two distinct groups; one encompassed the geographical range from Petén to Nicaragua ( i . e . JUCA-01 , PTN-01 , PTN-02 , NIC-01 , JUCA-02 , JUCH-04 , SASA-01 ) , while a second group included Belize ( BLZ-01 ) and an unidentified specimen from Guatemala ( UnID ) ( Table 2 ) . Specimens from the first group were most similar to the TcI DTU ( >92% mapping success to TcI-AODP , >74% TcI-ADWP ) , followed by TcVI ( <64% ) , TcII ( <46% ) and T . c . marinkellei ( <12% ) ( Table 2 ) . This was consistent with the TcI in-silico specimen , which mapped more successfully to the TcI reference genome than to any other DTU . Specimens from the second group mapped most closely to TcVI , consistently mapping >91% of their reads to the two available TcVI genomes , followed by TcII ( <76% ) , TcI ( <70% ) and T . c . marinkellei ( <12% ) , respectively ( Table 2 ) . This pattern was most similar to the TcVI in-silico reads , although compared to the TcVI in-silico tags , mapping success of the field specimens was lower for the TcVI genomes and higher for TcI and TcII ( Table 2 ) . Phylogenetic reconstruction also supported the existence of two DTUs ( Fig 6 ) . Although most specimens clustered with strong bootstrap support into a single clade with the two TcI genome references , the exceptions were BLZ-01 and UnID , which formed a distinct cluster , sister to TcI and distinct from the clade that includes the TcVI and TcII reference genomes ( Fig 6 ) . All leg and abdomen samples mapped successfully to the T . dimidiata reference catalog , with an average of 610 , 013 ± 80 , 410 mapped reads , corresponding to 19 , 577 ± 4 , 389 tags , and a total of 25 , 710 T . dimidiata SNPs across the 32 specimens . Of these , individual villages contained from 9–27% of the total allelic variation , resulting in over 1900 informative SNPs even at the smallest spatial scale assayed ( Table 3 ) . As the scale was increased from villages to regions , polymorphism was detected at an increasing proportion of SNPs , with the region of Jutiapa containing nearly 50% of the total number identified across the entire area of the study . K-means clustering and posterior DAPC revealed four main clusters corresponding to their geographical distributions among the 29 T . dimidiata individuals included in the analysis ( two excluded for low SNP counts , and one for which location data were not available ) ( Fig 7 ) . Madriz , Nicaragua ( NIC ) , Quiché , Guatemala ( QUI ) and La Bendición , El Salvador ( SABE ) were clustered in one group; the two northern sites , Río Frío , Belize ( BLZ ) and Petén , Guatemala ( PTN ) , were clustered in a second group; all individuals from Chiquimula , Guatemala ( CHAM , CHCE , CHGU and CHPR ) were isolated in a third cluster; and the remaining specimens from the region of Santa Ana , El Salvador and Jutiapa , Guatemala ( SACH , SAJU , SASA , JUBR , JUCA , JUCH and JUYU ) were grouped in a fourth cluster ( Fig 7 ) . The Fst values between clusters were greater than zero in all pair-wise comparisons; cluster 3 , which groups all individuals from Chiquimula , was the most differentiated , with pair-wise Fst ranging from 0 . 142 to 0 . 222 compared to 0 . 062 to 0 . 083 for all pair-wise combinations not involving cluster 3 ( Table 4 ) . Nucleotide diversity and observed heterozygosity were highest in cluster 4 ( El Salvador + Jutiapa ) compared to other clusters , despite the relatively small geographic area encompassed by this cluster ( Table 4; Fig 1 ) . Across all clusters , the expected heterozygosity tended to be higher than the observed ( Table 4 ) . For the 16% of reads with a significant BLAST hit ( e-value < 0 . 001 ) , 68% mapped to bacteria , 21% mapped to chordates , and the remaining 11% mapped to archaea , insects , protozoa , viruses , fungi and nematodes ( Fig 3b ) . Among chordates , 59% matched to known mammalian T . cruzi hosts , including dogs , humans , rodents , cats , swine , ruminants and opossum ( Fig 3b ) . Domestic birds , including chickens , ducks , and turkeys , constituted 30% of the bird BLAST reads . Within the viruses , 94% were bacteriophages . Fungal hits included entomopathogenic strains in the orders Hypocreales ( e . g . , Beauveria and Metarhizium ) and Entomophthorales ( e . g . , Zoophthora and Entomophaga ) typically used for biological control . Human and rodent parasitic nematodes , in the genera Angiostrongylus , Heligmosomoides , Haemonchus , Parastrongyloides , and Strongyloides constituted 94% of the nematode community and were found across all 32 specimens , while entomopathogenic nematodes from the genus Steinernema constituted 5% of the nematode mapped reads ( Fig 3b ) . Bacterial species richness varied significantly across specimen types ( F2 , 29 = 4 . 15 , p = 0 . 019 ) . Infected abdomens with T . cruzi contained significantly more bacterial species than non-infected abdomens ( post-hoc Tukey test , p <0 . 01 ) ( Fig 8 ) , but there was no difference in species richness between the leg specimens and either infected or non-infected abdomens . We identified 1 , 142 putative bacterial species across all abdomens . The reads from the subset of T . cruzi-infected abdomens mapped to 1 , 006 bacterial species , with 49% unique to a single specimen and 28% present across more than 50% of the infected abdomens . SNPs from non-infected abdomens mapped to 508 bacterial species , with 70% of the species mapping to a single specimen; however , only 12 species ( 2 . 4% ) from four genera ( Bacillus , Enterobacter , Ralstonia , and Alcaligenes ) were shared by more than 50% of the uninfected specimens . Unifrac analysis of gut bacterial community composition grouped specimens based on both geographic location and infection status . The first NMDS axis , explaining 47 . 2% of the variance , separated most regions from Guatemala and Belize from Quiché , Guatemala and El Salvador . The second NMDS axis , explaining 28 . 9% of the variance , separated Jutiapa from Chiquimula , Guatemala ( Fig 9 ) . Infected specimens from all sites were clustered around the origin . Permutation tests determined three statistically significant clusters: ( 1 ) non-infected specimens from Jutiapa , Guatemala ( p = 0 . 031 ) , ( 2 ) non-infected specimens from Chiquimula ( p = 0 . 028 ) , and ( 3 ) infected-specimens from multiple locations ( p = 0 . 043 ) ( Fig 9 ) . Five abdomens returned chordate reads for a single top-hit species at an order of magnitude higher than the background threshold calculated from the leg controls . Top hits for these specimens included chicken , dog , duck and human ( Fig 10 ) . Reads that matched chordates were present in all 32 specimens , including both abdomens and legs . The top hits had an exceedingly low representation in most specimens ( median = 0 . 035% of reads; Fig 10 ) ; these included human ( n = 23 ) , domestic birds ( chickens and ducks ) ( n = 5 ) , dog ( n = 1 ) , fish ( n = 1 ) , ruminant ( n = 1 ) and frog ( n = 1 ) ( S5 Table ) .
Our results suggest that RADseq can be used to simultaneously investigate T . cruzi infection and phylogenetic reconstruction of DTUs , population genetic structure of T . dimidiata , parasite-microbial interactions in the gut of the vector , and predominant blood meal source . For vector-borne diseases that involve multiple interacting species , methods that can produce data on an entire community can be used to leverage a single genetic study to address multiple biological questions across a range of taxa . Although the approach has some limitations , there was sufficient information to identify biologically meaningful patterns of genetic and community structure at a range of spatial scales , from individual villages to across Central America . Furthermore , the modest minimal requirements of 2–3 million reads to recover sufficient data on all taxa also makes RADseq a relatively economical method , with expected sequencing costs in 2016 of ~$30/specimen using current sequencing technologies ( e . g . , HiSeq 2000 ) . Notably , the method can be successful even for specimens preserved for considerable periods prior to sequencing , although careful assessment of DNA quantity and quality is critical for recovering sufficient high-quality read information from target taxa . RADseq successfully identified T . cruzi infection across multiple DTUs ( Fig 5 ) , with higher sensitivity than microscopy . The sensitivity of the method is important for surveys of parasite prevalence in natural populations , as T . cruzi infection intensity within vectors can range from high to exceedingly low representation of the parasite in the hindgut , and can vary across populations , species , physiological condition of the vector , anti-microbial activity in the gut and haemolymph , and co-occurrence of other pathogens and symbionts [67 , 68 , 69] . In general , molecular methods such as PCR-based detection have proven more sensitive compared to microscopy , but replicability of PCR methods is dependent on the volume of parasitic DNA extracted from the hindgut , the extraction protocol , and the DNA region that the probes amplify [70 , 71] . Given the low representation of the parasite across all specimens ( 1% of all trimmed reads ) , T . cruzi is likely to be more readily detected in RADseq libraries prepared with longer restriction enzymes that cut in fewer recognition sites , allowing higher depth of coverage across the parasite genome ( 6–8 bases , e . g . SbfI or PstI ) . Careful dissection to maximize the representation of parasite-rich tissues such as the lower abdomen and anus may also assist in T . cruzi recovery by preventing overrepresentation of the vector during sequencing . When T . cruzi is found , the genome-wide sampling provided by RADseq , in combination with the availability of reference genomes , also provides an effective tool for T . cruzi DTU identification and phylogenetic reconstruction . The two DTUs identified among the nine infected specimens clustered into two clear clades , with strong bootstrap support and branch lengths between clades ~10-fold longer than that within each DTU ( Fig 6 ) . The more common of these closely matched TcI , the DTU expected to be the most common in circulation in Central America [7 , 11 , 13] . The identity of the second DTU is unclear , as it did not cluster with any of the DTUs for which sequenced reference genomes are available . The two DTUs most commonly found in Central America are TcI and , less frequently TcIV , for which a reference genome was not available ( previously TcIIa ) [72–78] . As additional references become available , the power of the RADseq mapping approach to positively assess DTU identities throughout the Americas should progressively increase . Despite the absence of a sequenced reference genome for T . dimidiata , this study effectively identified SNP markers useful for understanding vector population structure . Even with relatively strict filtering criteria , using a small set of vector-only reference specimens to create a species-specific catalog yielded tens of thousands of SNP markers ( Figs 2 and 4a ) , and the low BLAST mapping to other insects ( 0 . 06% of all trimmed reads ) suggests that the method captured a substantial proportion of the true T . dimidiata tags in the mixed-DNA specimens . The SNP dataset was sufficiently large to enable population-genetic analysis across spatial scales with a single methodology , with thousands of variable loci present within individual villages that increased with each successive increase in spatial scale included ( Table 3 ) . Such flexibility is a considerable advantage over traditional markers , such as microsatellites or multi-locus gene sequencing , which are each most appropriate for questions at a particular temporal or spatial scale but uninformative for others . Even with the limited sampling included here , patterns of allelic variation successfully resolved biogeographic structure at multiple geographic scales ( Fig 7 ) , yielding four distinct genetic clusters corresponding to departmental and regional geographic divisions . As in previous studies , our results suggest moderate levels of differentiation within T . dimidiata across this region of Central America [24 , 26 , 27] , although clearly more comprehensive sampling focused on thorough biogeographic coverage will be needed to evaluate these patterns further . Although informative SNP markers were identified across all villages and departments in the present study , genetic variability was not consistent across space , with a range of 9–30% of loci showing polymorphisms at the village scale for samples that in all cases but one were collected in the same year for each village and with similar sample sizes ( Table 3 ) . This likely represents underlying variation in genetic diversity across the range of T . dimidiata; it is important to note that the current study focused on a portion of the species’ range , and thus it is not clear whether the variation and genetic structuring suggested here will extend to other regions or vector species . Even when variability was relatively low , however , the scale of genomic coverage afforded by techniques such as RADseq yielded a large absolute number of SNPs from the perspective of population-genetic analysis , and thus should facilitate effective SNP discovery for all but the most genetically uniform populations and species . RADseq can also reveal biologically interesting comparative patterns of microbiome variation that can subsequently be explored with more in-depth metagenomic approaches . From this study , two main drivers of gut bacterial community structure are evident . First , bacterial communities were strongly locally structured , with distinct species assemblages even between Jutiapa and Santa Ana , whose vector populations are not differentiated ( Figs 7 and 9 ) . Whether this is true spatial patterning , or reflects temporal , seasonal or other environmental variation among sites at the point of sampling or during processing cannot be determined from these data; however , this is an interesting avenue for future research . Second , T . cruzi parasitic infection significantly increases the diversity of bacteria ( p <0 . 01 ) , introducing a common additional set of infection-associated microbiota across the entire region ( Fig 9 ) . These patterns are consistent with recent literature demonstrating shifts in bacterial diversity across vector genera , by geographic location , and parasitic infection status [35 , 79] . How T . cruzi interacts with gut microbes is a promising area of future research in this system , as infection prevalence is highly variable across Central America and may be affected by the ability of native microbial communities to resist colonization [40 , 79] . Further studies of infection-associated bacterial taxa may also reveal important aspects of the transmission cycle . Infection may facilitate bacterial colonization due to modification of the immune response of the vector or changes in the gut lining [33 , 38]; alternatively , successful infection may be the end result of bacterial compositional changes associated with insect condition , health or other factors that make the gut environment more favorable for T . cruzi attachment [32 , 34 , 35 , 36 , 37 , 41 , 79] . Although RADseq can identify community patterns , it is likely to be poor for species-level identification of individual taxa such as bacterial symbionts that are not anticipated a priori . True species identity often could not be ascertained with confidence due to database limitations and lack of sequence specificity; a significant drawback of RADseq is the short read length , which can make it difficult to assign taxonomic identity with precision . Of the set of reads that did not map to either the parasite or vector , significant BLAST hits were returned for 20 . 1% of the queried reads ( Fig 3a ) . Even in the subset of reads with a significant hit , the likelihood that the taxon returned was the true DNA source depended on its representation in the nt database as well as the degree of evolutionary conservation of the genomic region . This was most evident in reads assigned to chordates , which occasionally returned species that clearly were not locally available , including model organisms ( e . g . , zebrafish ) and Old-World relatives of putative blood meals ( e . g . , gorilla ) . These were rare ( ~1% ) , and appear to represent highly conserved loci with close matches to a diverse set of taxa; because species calls were made without regard to how much better the top hit matched the query than the subsequent taxa; loci with equally-close matches to multiple taxa returned results that were consistent across runs but essentially arbitrary with respect to the species listed first . It is more difficult to assess the degree to which misassignment occurred in other taxonomic groups . With an undirected sequencing approach like RADseq , sequencing reads from the gut microbiome are an automatic consequence of targeting tissues harboring T . cruzi . Whether RADseq is sufficient for answering microbial community questions , however , is likely dependent on the type of information required . If the goal is to identify species that interact with T . cruzi or influence its transmission ( e . g . , Serratia marescens [38] ) or produce novel or functionally important chemical compounds , alternative next-generation sequencing methods such as shotgun metagenomic , transcriptomic and/or meta-barcoding methods could provide higher specificity and quantitative precision . This is less of a critical issue for community composition analysis , however , because the Unifrac procedure incorporates phylogenetic relationships into the distance measure , linking specimens even when minor sequence differences lead to different species calls . Given that ( 1 ) triatomines can live for several months in starvation , ( 2 ) the vast majority of insects sampled here were adults , which ingest proportionally smaller blood meals than nymphs , and ( 3 ) many field studies have found that specimens are often starved at the moment of collection , it was not surprising that we were able to confirm putative sources of blood meal from just 25% of the abdomens analyzed [80–83] . Nevertheless , the fact that contamination from human handling was uniformly present across samples , the RADseq approach was arguably least effective at resolving vector-feeding patterns , and is likely to be useful only for very recent or large blood meals . Minimizing handling , along with surface-sterilizing and extracting DNA under sterile conditions are advisable for minimizing such sources of ambiguity . In addition to background contamination , the strict DNA quality requirements for next-generation sequencing technologies likely introduce biases against detecting blood meals . Although using abdomen DNA has the tremendous advantage of investigating mixed taxa , the use of abdomens presents the challenge of obtaining high-quality DNA that has not been degraded by digestion . Previous studies targeting blood meals using species-specific primers recommended the use of PCR-based assays targeting small size amplicons of nuclear DNA to detect unique blood meals instead of a catchall method [82–84] . In our experience , obtaining high-quality DNA from the hindgut of adult T . dimidiata was challenging , with a total of 61 insects required to obtain the final 32 specimens . Even among these specimens , sequencing yield ranged from 489 , 656 to 18 , 878 , 597 reads , a 38-fold range . Many DNA specimens excluded from sequencing were characterized by a strong second band of degraded DNA at 100-200bp , possibly a degraded blood meal , in addition to the expected high-molecular weight band ( S1 Fig ) . The degradation from blood meal digestion is compounded by the challenge of field preservation , storage , and transport of specimens from remote areas with limited infrastructure . Although not enough specimens were tested to allow statistical comparisons , higher extraction success tended to be achieved when specimens were collected closer to the extraction date than those collected 3+ years earlier . Additionally , the time delay between DNA extraction and sequencing was kept to a maximum of one month to maintain the quality of the specimens and avoid DNA degradation during storage . A benefit of using taxonomically , non-specific sequencing approaches like RADseq is the potential for discovery of unexpected taxa that may be of ecological or epidemiological importance . One such finding was the common presence of entomopathogenic fungi ( 22% of fungi hits ) . Although none of the specimens showed visual evidence of cuticular fungal germination , the presence of Beauveria , Metarhizium , Zoophthora , and Entomophaga , both in the abdomens and legs , suggest possible latent infection of the vectors by spores waiting for environmental cues that can trigger germination [85] . Although the fungal inoculation sources are unknown , the presence of the entomopathogenic genera across tissues and specimens suggests a wide distribution of spores regardless of the local environment in which the triatomine was collected [86] . We also found a low signal of entomopathogenic nematode species from the family Steinernematidae . Additionally , the BLAST search revealed a wide range of common mammalian parasitic nematodes from the genera Angiostrongylus , Heligmosomoides , Haemonchus , Parastrongyloides and Strongyloides ( Fig 2 ) . Although to some extent this may be a result of transfer from humans to the bug during handling , this result raises the possibility that T . dimidiata may harbor and/or transmit such parasites as a passive carrier of infective free-living larvae or eggs [87] . This is a meaningful finding because of the potential of co-transmission of additional human pathogens , which has been previously documented in other vectors such as Aedes aegypti and A . albopictus [88] . The role of a triatomine vector could either involve the cutaneous transportation of the nematode as it moves from dirt crevices to the skin of mammalian host or by gut transportation; eventually defecating eggs near open wounds , eyes , or areas prone to oral contamination [89 , 90] . It is unlikely that the vector can acquire the nematodes from a blood meal source given that only the genus Strongyloides is known to have a non-reproductive larval stage in the human bloodstream , and even in this case , it is cutaneously transmitted , remaining in the bloodstream only in transition to the small intestine [91] . The detection of other human pathogenic nematodes opens new avenues of research to study the role of triatomines in the context of vector-aided transmission . Although the aim of this study was not to reveal community patterns beyond the parasite , vector and microbiota , our findings can potentially lead to community-based studies of entomopathogenic fungi and nematodes , human parasitic nematodes and other taxa with relevant association to disease transmission complexes .
Overall , our results show that a mixed-DNA approach can provide simultaneous information on the community of biotic factors involved in T . cruzi transmission . RADseq can provide informative SNP marker sets for taxonomic and biogeographic analysis for both vector population genetic structure and parasite evolutionary history . It also has a strong potential to retrieve information about the community ecology and diversity of microbiota; and although it is limited at revealing quantitative details of vector feeding history , this method may be useful for identifying recent vertebrate hosts . For all of these areas of inquiry , a broad-based sequencing approach can reveal novel patterns that can be followed up with complementary approaches ( e . g . , proteomics , metagenomics ) . Testing this mixed-DNA sequencing method with different vectors and disease models will help to determine its reproducibility in other systems where multiple organisms interact in tightly-integrated and complex ways .
|
Chagas disease is caused by the parasite Trypanosoma cruzi , which is spread by triatomine kissing bugs . There are many biotic factors that influence the risk of disease transmission , including the strain of the parasite , the vector movement patterns , the community of microbes interacting with the parasite inside the vector's gut , and the availability of suitable vertebrate hosts . DNA from all of these species can be found in the gut of an infected bug , providing an opportunity to investigate all of them simultaneously by genetically analyzing this single tissue . In this study , we developed a DNA-based method to retrieve , separate , and analyze genetic information from the abdomens of 32 T . dimidiata kissing bug vectors collected across Central America . We found two distinct strains of T . cruzi , and four T . dimidiata genetic clusters associated with environmental and geographical characteristics . These populations harbored different bacterial gut communities that were augmented by specifically infection-associated bacteria when the vector was infected by the parasite . In some cases , we could identify what the insect had recently fed on , including chicken , duck , dog and human . Having simultaneous information on all of these organisms may help to fine-tune control strategies that influence the risk of T . cruzi transmission .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusions"
] |
[
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"abdomen",
"geographical",
"locations",
"vector-borne",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"invertebrate",
"genomics",
"north",
"america",
"protozoans",
"central",
"america",
"guatemala",
"infectious",
"diseases",
"animal",
"genomics",
"people",
"and",
"places",
"trypanosoma",
"cruzi",
"trypanosoma",
"eukaryota",
"blood",
"anatomy",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"organisms"
] |
2018
|
Uncovering vector, parasite, blood meal and microbiome patterns from mixed-DNA specimens of the Chagas disease vector Triatoma dimidiata
|
Fungal pathogens exploit diverse mechanisms to survive exposure to antifungal drugs . This poses concern given the limited number of clinically useful antifungals and the growing population of immunocompromised individuals vulnerable to life-threatening fungal infection . To identify molecules that abrogate resistance to the most widely deployed class of antifungals , the azoles , we conducted a screen of 1 , 280 pharmacologically active compounds . Three out of seven hits that abolished azole resistance of a resistant mutant of the model yeast Saccharomyces cerevisiae and a clinical isolate of the leading human fungal pathogen Candida albicans were inhibitors of protein kinase C ( PKC ) , which regulates cell wall integrity during growth , morphogenesis , and response to cell wall stress . Pharmacological or genetic impairment of Pkc1 conferred hypersensitivity to multiple drugs that target synthesis of the key cell membrane sterol ergosterol , including azoles , allylamines , and morpholines . Pkc1 enabled survival of cell membrane stress at least in part via the mitogen activated protein kinase ( MAPK ) cascade in both species , though through distinct downstream effectors . Strikingly , inhibition of Pkc1 phenocopied inhibition of the molecular chaperone Hsp90 or its client protein calcineurin . PKC signaling was required for calcineurin activation in response to drug exposure in S . cerevisiae . In contrast , Pkc1 and calcineurin independently regulate drug resistance via a common target in C . albicans . We identified an additional level of regulatory control in the C . albicans circuitry linking PKC signaling , Hsp90 , and calcineurin as genetic reduction of Hsp90 led to depletion of the terminal MAPK , Mkc1 . Deletion of C . albicans PKC1 rendered fungistatic ergosterol biosynthesis inhibitors fungicidal and attenuated virulence in a murine model of systemic candidiasis . This work establishes a new role for PKC signaling in drug resistance , novel circuitry through which Hsp90 regulates drug resistance , and that targeting stress response signaling provides a promising strategy for treating life-threatening fungal infections .
Microbial survival depends critically upon coordination of sensing environmental stimuli with control of the appropriate cellular responses . As a consequence , microbes have evolved elaborate mechanisms to sense and respond to diverse environmental stresses , including oxidative stress , osmotic stress , thermal stress , changes in pH , and nutrient limitation [1] , [2] . Signal transduction cascades integrate recognition and response to these stresses as well as to challenges imposed by exposure to various small molecules that are a ubiquitous presence in the environment . Small molecules can have a dramatic effect on cellular signaling , mediate communication between microbes , or exert potentially lethal toxicity [3] , [4] , [5] , [6] , [7] . Many natural products are produced by microbes in competitive communities and can lead to selection for enhanced capacity to tolerate these agents . Since natural products and their derivatives are extensively used in medicine and agriculture [8] , [9] , the evolution of resistance to these agents can have profound consequences for human health . The evolution of drug resistance in fungal pathogens poses considerable concern given that invasive fungal infections are a leading cause of human mortality worldwide , especially among immunocompromised individuals . The frequency of such infections is on the rise in concert with the growing population of patients with compromised immune systems due to chemotherapy , transplantation of organs or hematopoietic stem cells , or infection with HIV [10] , [11] . The leading fungal pathogen of humans is Candida albicans , which ranks as the fourth most common cause of hospital acquired infectious disease and is associated with mortality rates approaching 50% [12] , [13] , [14] . There is a very limited repertoire of antifungal drugs with distinct targets for the treatment of fungal infections , in part due to the close evolutionary relationships between these eukaryotic pathogens and their hosts [15] , [16] . Most of the antifungal drugs in clinical use target the biosynthesis or function of ergosterol , the main sterol of fungal membranes [2] , [17] , [18] . The therapeutic efficacy of most antifungal drugs is compromised by the emergence of drug resistant strains , superinfection with resistant strains , and by static rather than cidal activities that block fungal growth but do not eradicate the pathogen population . To improve clinical outcome it will be necessary to develop new antifungal drugs with different mechanisms of action and to discover drugs that improve the fungicidal activity of current antifungals . The molecular basis of antifungal drug resistance is best characterized in the context of the azoles through studies with C . albicans and the model yeast Saccharomyces cerevisiae . The azoles have been the most widely deployed class of antifungal drugs for decades and inhibit lanosterol 14α-demethylase , encoded by ERG11 , resulting in a block in ergosterol biosynthesis , the accumulation of a toxic sterol intermediate , and cell membrane stress [2] , [17] , [18] . The azoles are generally fungistatic against Candida species and many patients are on long-term therapy , creating favorable conditions for the emergence of resistance . Despite the evolutionary distance between C . albicans and S . cerevisiae , mechanisms of azole resistance are largely conserved [19] . Resistance can arise by mechanisms that minimize the impact of the drug on the fungus , such as the overexpression of multidrug transporters or alterations of the drug target that prevent the drug from inhibiting its target . Alternatively , resistance can arise by mechanisms that minimize drug toxicity , such as loss of function of the ergosterol biosynthetic enzyme Erg3 , which blocks the production of the toxic sterol that would otherwise accumulate when the azoles inhibit Erg11 . Recent studies have established that basal tolerance of wild-type strains and resistance due to mechanisms that mitigate drug toxicities without blocking the effect of the drug on the cell are often dependent upon stress responses that are critical for survival of azole-induced cell membrane stress [2] , [18] . The key regulator of cellular stress responses implicated in both basal tolerance and resistance to azoles is Hsp90 [2] , [18] , [20] . Hsp90 is an essential molecular chaperone that regulates the stability and function of a diverse set of client proteins , many of which are regulators of cellular signaling [21] , [22] , [23] . In S . cerevisiae and C . albicans , inhibition of Hsp90 function blocks the rapid evolution of azole resistance and abrogates resistance that was acquired by diverse mutations [24] , [25] . A central aspect of Hsp90's role in the emergence and maintenance of azole resistance is that it enables calcineurin-dependent stress responses that are required to survive the membrane stress exerted by azoles . In both yeast species , Hsp90 physically interacts with calcineurin keeping it in a stable conformation that is poised for activation [26] , [27] . Inhibition of calcineurin function phenocopies inhibition of Hsp90 function , abrogating azole resistance of diverse mutants [24] , [25] . This has led to the model that calcineurin is the key mediator of Hsp90-dependent azole resistance . Notably , in C . albicans both Hsp90 and calcineurin have recently been demonstrated to regulate resistance to the echinocandins , the only new class of antifungals to reach the clinic in decades; they inhibit the synthesis of ( 1 , 3 ) -β-D-glucan , a key component of the fungal cell wall [20] , [27] . Another key cellular stress response pathway implicated in basal tolerance to antifungal drugs is the protein kinase C ( PKC ) cell wall integrity pathway , though it has only been implicated in tolerance to drugs targeting the cell wall . Central to the core of this signaling cascade is Pkc1 , the sole PKC isoenzyme in S . cerevisiae that is essential under standard growth conditions and regulates maintenance of cell wall integrity during growth , morphogenesis , and response to cell wall stress [28] , [29] , [30] , [31] . Signals are initiated by a family of cell surface sensors that are coupled to the small G-protein Rho1 , which activates a set of effectors including Pkc1 . Pkc1 signaling has been the focus of extensive study in S . cerevisiae where it is known to regulate multiple targets , most notably the mitogen-activated protein kinase ( MAPK ) cascade comprised of a linear series of protein kinases including the MAPKKK Bck1 , the MAPKKs Mkk1/2 , and the MAPK Slt2 that relays signals to the terminal transcription factors Rlm1 and Swi4/Swi6 . While Pkc1 is not essential in C . albicans [32] , the Pkc1-activated MAPK cascade is conserved in C . albicans with Bck1 , Mkk2 , and the Slt2 homolog Mkc1 [33] . In both species , components of the Pkc1 signaling cascade have been implicated in mediating tolerance to the stress exerted by the echinocandins that target the fungal cell wall [34] , [35] , [36] , [37] . Here , we embarked on a drug screen of 1 , 280 pharmacologically active compounds to identify molecules that abrogate azole resistance of both an S . cerevisiae resistant mutant and a C . albicans clinical isolate . We identified a key role for PKC signaling in mediating crucial responses to azoles as well as to other drugs targeting the ergosterol biosynthesis pathway , including allylamines and morpholines . Pkc1 regulated responses to azoles at least in part via the MAPK cascade in both species via multiple downstream effectors . Strikingly , inhibition of Pkc1 function phenocopied inhibition of Hsp90 or calcineurin . In S . cerevisiae , compromise of PKC signaling blocked calcineurin activation in response to ergosterol biosynthesis inhibitors , providing a compelling mechanism for the impact on drug resistance . In C . albicans , we found that Pkc1 and calcineurin independently regulate resistance via a common target . The complexity of interactions linking PKC signaling , Hsp90 , and calcineurin was further illuminated as genetic reduction of C . albicans Hsp90 resulted in destabilization of Mkc1 thereby blocking its activation . Deletion of C . albicans PKC1 rendered the fungistatic ergosterol biosynthesis inhibitors fungicidal and attenuated virulence in a murine model of systemic disease . Our findings establish an entirely new role for PKC signaling in basal tolerance and resistance to ergosterol biosynthesis inhibitors , a novel mechanism through which Hsp90 regulates drug resistance , and that targeting Pkc1 provides a promising therapeutic strategy for life-threatening fungal infections .
To identify compounds that enhance the efficacy of the azole fluconazole we screened the LOPAC1280 Navigator library . Our initial screen used an S . cerevisiae strain with azole resistance due to deletion of ERG3 . This resistance phenotype is exquisitely sensitive to perturbation of stress response pathways [24] , [25] . To enhance the activity of library compounds , this azole-resistant mutant also harbored deletion of PDR1 and PDR3 , transcription factors that regulate the expression of numerous multidrug transporters which efflux structurally diverse compounds from the cell [38] . The library was initially screened at 25 µM in defined RPMI medium at 30°C in the presence of 8 µg/ml fluconazole , which reduces growth of this strain by less than 50% . The compounds that reduced growth by greater than or equal to 50% relative to the fluconazole-only controls were re-screened at 12 . 5 µM in the presence and absence of fluconazole to distinguish those that enhance the activity of fluconazole from those that are simply toxic on their own . This screen identified 185 compounds that enhanced the efficacy of fluconazole ( data not shown ) . To prioritize compounds with synergistic activity with fluconazole against a clinical isolate of C . albicans , we then screened the 185 compounds at 12 . 5 µM for activity against an isolate from an HIV-infected patient undergoing fluconazole treatment , both in the presence and absence of fluconazole at 8 µg/ml . The capacity of this clinical isolate to grow in the presence of high concentrations of azole is critically dependent upon cellular stress responses [25] , despite the fact that it has increased expression of the multidrug transporter Mdr1 relative to a drug-sensitive isolate recovered from the same patient at an earlier time point [39] , [40] , [41] . This secondary screen identified seven compounds that had little toxicity on their own but which enhanced the efficacy of fluconazole ( Figure 1A ) . One hit from our screen , brefeldin A , was recently confirmed to exhibit potent synergy with antifungals against Candida and Aspergillus [42] . Strikingly , three of the seven hits were characterized as inhibitors of protein kinase C ( PKC ) . PKC governs the cell wall integrity signaling pathway so named for its role in regulating cell wall integrity during growth , morphogenesis , and exposure to stress in fungi [29] , [30] , [31] . In both S . cerevisiae and C . albicans , the PKC signaling cascade is known to regulate cellular responses crucial for survival of exposure to antifungal drugs targeting the cell wall , such as the echinocandins [34] , [35] , [36] , [37] . Since the PKC inhibitors identified in our screen were characterized in mammalian cells [43] , [44] , we next turned to other pharmacological inhibitors of PKC whose mode of action had been validated in fungi . Cercosporamide was identified as a selective Pkc1 inhibitor through C . albicans Pkc1-based high-throughput screening and was shown to exhibit potent synergy with echinocandins [45] . We purified cercosporamide from the fungus Cercosporidium henningsii following standard protocols [46] . As a positive control , we tested the impact of a concentration gradient of cercosporamide on growth in the presence of a fixed concentration of the echinocandin micafungin that causes less than 50% inhibition of growth on its own and confirmed that cercosporamide had the expected synergistic activity with micafungin against the clinical C . albicans isolate ( Figure 1B ) . Using a comparable assay , we determined that cercosporamide also enhanced the activity of fluconazole ( Figure 1B ) , validating the results from our screen . We further confirmed our pharmacological findings with another PKC inhibitor characterized in fungi , staurosporine [47] , [48] . Both cercosporamide and staurosporine enhanced the efficacy of antifungals targeting the cell wall , micafungin , and those targeting the cell membrane ( Figure 1C ) , including fluconazole and the morpholine fenpropimorph , which inhibits Erg2 and Erg24 [49] . While staurosporine enhanced the efficacy of another ergosterol biosynthesis inhibitor that inhibits Erg1 [49] , the allylamine terbinafine , cercosporamide did not ( Figure 1C ) . The lack of effect of cercosporamide on terbinafine tolerance is likely an artifact of an inactivating drug-drug interaction given that mutants that are hypersensitive to terbinafine are rendered resistant by cercosporamide ( data not shown ) . In S . cerevisiae , PKC1 is essential [50] , thus we used a strain harboring only a temperature-sensitive ( ts ) pkc1-3 allele [51] and assayed tolerance to three ergosterol biosynthesis inhibitors fluconazole , fenpropimorph , and terbinafine . Growth of the wild-type strain and the pkc1-3 ts mutant was assayed over a gradient of drug concentrations relative to a drug-free control at either the permissive temperature ( 30°C ) or at a more restrictive temperature , but where the pkc1-3 ts mutant was still able to thrive in the absence of antifungals ( 35°C ) . At the permissive temperature , the wild type and the pkc1-3 ts mutant had comparable tolerance to all three drugs tested ( Figure S1A ) . At the restrictive temperature , the pkc1-3 ts mutant was hypersensitive to all three drugs ( Figure 2A ) . The same trend was observed when a dilution series of cells was spotted on solid medium with a fixed concentration of drug ( Figure S1B ) . To determine if reduction of Pkc1 function rendered the fungistatic ergosterol biosynthesis inhibitors fungicidal we used tandem assays with an antifungal susceptibility test performed at the restrictive temperature followed by spotting onto rich medium without any inhibitors . The wild-type strain was able to grow on rich medium following exposure to all concentrations of drug tested ( Figure 2B ) ; compromise of Pkc1 function in the pkc1-3 ts mutant enhanced cidality of all three drugs with the most severe effect for fluconazole and fenpropimorph . Thus , reduction of Pkc1 activity increases sensitivity to drugs targeting the cell membrane and enhances cidality of these otherwise fungistatic agents . Despite the simple linear schematic commonly used to illustrate the architecture of the Pkc1 cell wall integrity pathway ( Figure 2C ) , there is evidence for additional Pkc1 targets [30] and multiple cases of cross talk with other stress response pathways [28] . We next sought to determine if the effects of Pkc1 on tolerance to ergosterol biosynthesis inhibitors are due to signaling via the downstream MAPK cascade . S . cerevisiae mutants lacking the MAPKKK Bck1 or the terminal MAPK Slt2 were hypersensitive to all three ergosterol biosynthesis inhibitors tested in both a liquid antifungal susceptibility assay measuring growth of a fixed concentration of cells across a gradient of drug concentrations ( Figure 2A ) and a spotting assay of a dilution of cells on solid medium with a fixed concentration of drug ( Figure S1B ) . Deletion of the MAPK components also rendered these fungistatic drugs fungicidal . Thus , Pkc1 enables tolerance to ergosterol biosynthesis inhibitors via the MAPK cascade in S . cerevisiae . In C . albicans , PKC1 is not essential though it does share a high degree of sequence conservation with S . cerevisiae PKC1 and has a conserved role in regulating cell wall integrity through a conserved MAPK cascade [32] , [33] . To genetically validate the role of C . albicans PKC1 in tolerance to drugs affecting the cell membrane , we constructed a pkc1Δ/pkc1Δ mutant . Homozygous deletion of PKC1 rendered the strain hypersensitive to all three ergosterol biosynthesis inhibitors tested in liquid static susceptibility assays ( Figure 3A ) as well as on solid medium ( Figure S2A ) . Comparable results were obtained in well-aerated shaking liquid cultures ( data not shown ) . Restoring a wild-type PKC1 allele under the control of the native promoter to the native locus restored drug tolerance ( Figure S2 ) . To determine if deletion of C . albicans PKC1 renders the ergosterol biosynthesis inhibitors fungicidal , we used tandem assays with an antifungal susceptibility test followed by spotting onto rich medium without inhibitor . A strain with wild-type PKC1 levels was able to grow on rich medium following exposure to all drug concentrations tested ( Figure 3B ) . Homozygous deletion of C . albicans PKC1 was cidal in combination with any dose of ergosterol biosynthesis inhibitor tested; no cells were able to grow on rich medium following exposure to the treatments . Thus , Pkc1 regulates crucial cellular responses for surviving the cell membrane stress exerted by antifungal drugs . As an initial approach to assess whether the MAPK cascade was implicated in responses to drugs targeting the cell membrane , we monitored activation of the terminal MAPK in C . albicans . Mkc1 is known to be activated in response to distinct stress conditions including oxidative stress , changes in osmotic pressure , cell wall damage , and cell membrane perturbation [52] . To determine if Mkc1 is activated in response to ergosterol biosynthesis inhibitors we monitored Mkc1 phosphorylation using an antibody that detects dual phosphorylation on conserved threonine and tyrosine residues . Exposure to fluconazole , fenpropimorph , and terbinafine led to Mkc1 activation comparable to exposure to the cell wall damaging antifungal micafungin ( Figure S3A ) . However , activation of signal transducers is not always coupled with functional consequences of their deletion . For example , Mkc1 is activated by exposure to hydrogen peroxide but is not required for survival of this stress [52] . To determine if the role of the MAPK cascade was conserved in C . albicans , we constructed homozygous deletion mutants lacking either the MAPKKK Bck1 or the terminal MAPK Mkc1 ( homolog of S . cerevisiae Slt2 ) . Homozygous deletion of either BCK1 or MKC1 rendered strains hypersensitive to fluconazole , fenpropimorph , and terbinafine ( Figure 3A ) but had negligible effect at elevated temperatures ( Figure S3B ) . This stands in contrast to our results with S . cerevisiae that demonstrated an equivalent role of the MAPK cascade at all temperatures tested ( Figure 2 , Figure S1 and S4 ) . While deletion of C . albicans PKC1 rendered the ergosterol biosynthesis inhibitors fungicidal , deletion of BCK1 or MKC1 did not ( Figure 3B ) . These results not only implicate the MAPK cascade in C . albicans but also suggest that alternate effectors downstream of Pkc1 are more important at elevated temperature and enable survival in the presence of ergosterol biosynthesis inhibitors . Effectors downstream of the terminal MAPK of the PKC signaling cascade have been well studied in S . cerevisiae and include both nuclear and cytoplasmic proteins . Slt2 is known to regulate activation of two transcription factors Rlm1 and SBF , which is comprised of Swi4 and Swi6 [30] . Rlm1 mediates the majority of the transcriptional output of cell wall integrity signaling , largely genes involved in cell wall biogenesis [53] . SBF drives cell cycle-specific transcription and is also regulated by Slt2 in response to cell wall stress ( reviewed in [30] ) . Swi4 interacts directly with Slt2 and has additional roles in transcriptional regulation independent of the regulatory subunit Swi6 [54] . Slt2 translocates from the nucleus to the cytoplasm in response to cell wall stress [55] . Cytoplasmic Slt2 is required for activation of a high-affinity Ca2+ influx system in the plasma membrane that is comprised of two subunits , Cch1 and Mid1 , in response to endoplasmic reticulum stress [56] . Activation of the Cch1-Mid1 channel leads to the accumulation of intracellular Ca2+ and activation of the protein phosphatase calcineurin [57] . To dissect the role of downstream effectors of Slt2 in ergosterol biosynthesis inhibitor tolerance , we tested the impact of their deletion individually and in combination on antifungal susceptibility . For reference , we included a strain lacking the regulatory subunit of calcineurin , CNB1 , which is hypersensitive to ergosterol biosynthesis inhibitors [24] . For fluconazole , deletion of RLM1 , CCH1 , or MID1 had negligible impact on tolerance while deletion of SWI4 or SWI6 rendered strains almost as sensitive as the slt2Δ mutant ( Figure 4 ) . To determine if there was redundancy among the downstream effectors , we constructed strains harboring deletion of multiple effectors . Deletion of CCH1 phenocopies deletion of the entire channel and deletion of SWI4 abolishes SBF function as well as Swi4-dependent transcription independent of SBF . Thus , combined deletion of CCH1 , SWI4 , and RLM1 should eliminate the four known targets of Slt2 phosphorylation . No additional increase in sensitivity was observed in double or triple mutants . This suggests that the SBF transcription factor is of central importance for enabling responses to fluconazole . For fenpropimorph , deletion of RLM1 , CCH1 , or MID1 had no impact on tolerance individually while deletion of SWI4 or SWI6 caused a partial increase in sensitivity ( Figure 4 ) . Deletion of RLM1 in the context of the swi4Δ or swi6Δ mutants further increased fenpropimorph sensitivity . Deletion of CCH1 in the mutant backgrounds had little additional impact . This suggests that SBF is the major determinant of fenpropimorph tolerance with RLM1 enabling additional responses important in the absence of SBF . For tolerance to terbinafine , deletion of RLM1 had no impact while deletion of SWI4 caused a partial increase in sensitivity ( Figure 4 ) . Unlike tolerance to fluconazole and fenpropimorph , deletion of SWI6 had negligible impact on terbinafine tolerance while deletion of CCH1 or MID1 caused a partial increase in sensitivity . Deletion of both RLM1 and CCH1 in the swi4Δ mutant caused an incremental increase in sensitivity ( Figure 4 ) . These results suggest that Swi4 enables terbinafine tolerance independent of the SBF complex and that Rlm1 and Cch1 mediate responses that are important in the absence of Swi4 . Thus , distinct downstream effectors are important for tolerance of S . cerevisiae to different ergosterol biosynthesis inhibitors . Next , we tested a set of C . albicans mutants to determine if the role of the effectors downstream of the terminal MAPK of the PKC signaling cascade was conserved . As was the case with S . cerevisiae , deletion of RLM1 on its own had no impact on tolerance to the ergosterol biosynthesis inhibitors ( Figure 5 ) , consistent with recent findings [58] . Deletion of SWI4 rendered strains hypersensitive to all three ergosterol biosynthesis inhibitors tested ( Figure 5 ) . Deletion of SWI6 or combined deletion of both SWI4 and SWI6 conferred a comparable increase in sensitivity ( data not shown; unpublished strains generously provided by Catherine Bachewich ) , implicating the SBF complex in responses to drug-induced membrane stress . Deletion of CCH1 or MID1 individually or in combination had a comparable effect to deletion of SWI4 rendering the strain hypersensitive to all three ergosterol biosynthesis inhibitors tested ( Figure 5 ) . Notably , C . albicans cch1Δ/cch1Δ and mid1Δ/mid1Δ mutants share some but not all phenotypes of a calcineurin mutant [59] . In terms of ergosterol biosynthesis inhibitor sensitivity , deletion of the gene encoding the catalytic subunit of calcineurin , CNA1 , caused hypersensitivity akin to that of the cch1Δ/cch1Δ and mid1Δ/mid1Δ mutants for fluconazole and fenpropimorph but caused slightly greater sensitivity to terbinafine ( Figure 5 ) . Thus , in C . albicans both the SBF complex and the Cch1-Mid1 channel play critical roles in tolerance to drugs that target the cell membrane . Given calcineurin's established role in mediating drug-induced membrane stress responses [24] , [25] , [57] and that Slt2 has been shown to enable calcineurin activation by phosphorylating Cch1 [56] , we tested whether calcineurin was activated in response to ergosterol biosynthesis inhibitors and whether deletion of Slt2 blocked this activation . To monitor calcineurin activation , we used a well-established reporter system that exploits the downstream effector Crz1 , a transcription factor that is dephosphorylated upon calcineurin activation [60] , [61] . Dephosphorylated Crz1 translocates to the nucleus , driving expression of genes with calcineurin-dependent response elements ( CDREs ) in their promoters . We used a reporter containing four tandem copies of CDRE and a CYC1 minimal promoter driving lacZ [61] . We confirmed previous findings that fluconazole activates calcineurin ( [27] , [62] and Figure 6A ) . We also found that the other ergosterol biosynthesis inhibitors terbinafine and fenpropimorph activate calcineurin ( P<0 . 001 , ANOVA , Bonferroni's Multiple Comparison Test , Figure 6A ) . Deletion of SLT2 completely blocked calcineurin activation in response to ergosterol biosynthesis inhibitors as did deletion of the regulatory subunit of calcineurin required for its activation , encoded by CNB1 ( P<0 . 001 ) . Pharmacological inhibition of PKC signaling with staurosporine also blocked calcineurin activation ( P<0 . 001 , Figure 6B ) . The block in calcineurin activation was not an artifact of compromised viability as treatment conditions were optimized such that all cultures underwent comparable growth with equivalent protein yields . Given that the slt2Δ mutant is slightly more sensitive to ergosterol biosynthesis inhibitors than the mutant lacking calcineurin function , it is likely that Slt2 regulates responses to ergosterol biosynthesis inhibitors through additional targets . The swi4Δ mutant is less sensitive than the calcineurin mutant , suggesting that Slt2 regulates calcineurin function independently of Swi4 and that Swi4 regulates ergosterol biosynthesis inhibitor tolerance through additional targets ( Figure 6C ) . Given that deletion of CCH1 and MID1 had negligible effect on tolerance to fluconazole or fenpropimorph and only an intermediate effect on tolerance to terbinafine , it is likely that compromise of PKC signaling blocked calcineurin activation by a mechanism that is largely distinct from the Cch1-Mid1 channel . One possible mechanism is that the effects are transcriptional and mediated through a nuclear target of Slt2 such that inhibition of PKC signaling compromises the expression of calcineurin or CRZ1 . However , deletion of SLT2 did not reduce the expression of genes encoding any of the calcineurin subunits ( CNA1 , CNA2 or CNB1 ) or CRZ1 as measured by quantitative RT-PCR in the presence or absence of ergosterol biosynthesis inhibitor ( P>0 . 05 , ANOVA , Bonferroni's Multiple Comparison Test , Figure S5 ) . Thus , PKC signaling enables calcineurin activation in response to ergosterol biosynthesis inhibitors by a mechanism that is largely distinct from the Cch1-Mid1 channel or transcriptional control of calcineurin . In contrast to the minor impact of deletion of the S . cerevisiae Cch1-Mid1 channel , deletion of the C . albicans Cch1-Mid1 channel had nearly as great an effect as deletion of the catalytic subunit of calcineurin in response to fluconazole and fenpropimorph; for terbinafine the effect was partial ( Figure 5 ) . To test if the ergosterol biosynthesis inhibitors activate calcineurin and if inhibition of PKC signaling blocks this activation , we monitored transcript levels of two calcineurin-dependent genes , PLC3 and UTR2 [63] . In a wild-type strain , fluconazole activated calcineurin as measured by an increase in PLC3 and UTR2 transcript levels ( P<0 . 05 , ANOVA , Bonferroni's Multiple Comparison Test , Figure 7A ) . As expected , deletion of the catalytic subunit of calcineurin , CNA1 , blocked the induction of PLC3 and UTR2 transcripts ( P<0 . 01 ) . Deletion of PKC1 did not block induction of PLC3 or UTR2 indicating that impairment of PKC signaling does not block calcineurin activation ( Figure 7A ) . At 35°C , conditions under which Pkc1 downstream effectors other than the MAPK cascade are more important in tolerance to ergosterol biosynthesis inhibitors , deletion of PKC1 increased the magnitude of induction of PLC3 and UTR2 ( P<0 . 001 , Figure 7A ) . At 30°C , conditions under which the MAPK cascade mediates tolerance to ergosterol biosynthesis inhibitors , deletion of PKC1 had no significant impact on calcineurin-dependent transcription ( data not shown ) . Thus , drugs that inhibit ergosterol biosynthesis induce calcineurin activation in a manner that is independent of PKC signaling . Next , we addressed alternative models that could explain the relationship between PKC signaling and calcineurin in C . albicans tolerance to ergosterol biosynthesis inhibitors . One possible model is that Pkc1 and calcineurin regulate tolerance through parallel but non-redundant pathways . This model leads to two predictions for ergosterol biosynthesis inhibitor tolerance: 1 ) there should be a synergistic effect of inhibiting both pathways simultaneously and 2 ) compromise of one pathway should confer increased sensitivity to inhibition of the other . To test the first prediction , we performed checkerboard assays in which a wild-type strain was exposed to a uniform concentration of fluconazole and a concentration gradient of both the calcineurin inhibitor cyclosporin A and the Pkc1 inhibitor staurosporine . There was no obvious synergy detected upon inhibition of both pathways in combination with fluconazole ( Figure 7B ) . To assess this quantitatively we calculated the standard index of drug synergy , the fractional inhibitory concentration ( FIC ) . The FIC value was 0 . 75 confirming that there was no synergy . To test the second prediction , we measured the impact of Pkc1 inhibition on fluconazole tolerance of a mutant lacking the catalytic subunit of calcineurin , CNA1 . Growth of the cna1Δ/cna1Δ mutant was assessed in the absence or presence of the highest concentration of fluconazole that it could tolerate and with a gradient of the Pkc1 inhibitor cercosporamide . Fluconazole-sensitivity of the cna1Δ/cna1Δ mutant was not affected by cercosporamide ( Figure 7C ) . The reciprocal was also true , such that the pkc1Δ/pkc1Δ mutant was not rendered hypersensitive to fluconazole by the calcineurin inhibitor cyclosporin A ( data not shown ) . Thus , our results do not support either prediction of the model in which Pkc1 and calcineurin regulate ergosterol biosynthesis inhibitor tolerance through parallel pathways . We also ruled out the possibility that inhibition of calcineurin blocks PKC signaling as measured by levels of activated Mkc1 ( data not shown ) . Taken together , these findings support a model in which Pkc1 and calcineurin independently regulate crucial responses to ergosterol biosynthesis inhibitors through a common target ( Figure 7D ) . This target is not Crz1 , the only well-characterized effector downstream of C . albicans calcineurin , given that transcription of the calcineurin-dependent genes PLC3 and UTR2 is mediated through the transcription factor Crz1 [63] . To determine if the role of PKC signaling in tolerance to drugs targeting the cell membrane was conserved in the context of bona fide drug resistance , we turned to C . albicans clinical isolates and resistant mutants ( Figure 8 ) . We tested the impact of two structurally unrelated Pkc1 inhibitors , cercosporamide and staurosporine , on azole susceptibility of a series of C . albicans isolates that evolved fluconazole resistance in a human host [40] . The isolates shown begin with the second isolate in the series , which is the first with elevated resistance and increased expression of the multidrug transporter Mdr1 [39] , [40] , [41] . Azole resistance of this series is known to have evolved from a state of dependence on calcineurin and Hsp90 to a state of independence and this change is associated with the accumulation of additional mutations [25] . The third isolate from the bottom ( Figure 8A ) has mutation ( R467K ) and increased expression of the azole target Erg11; the last two isolates , which show the least of effect of Hsp90 inhibition on resistance , also have increased expression of the multidrug transporter Cdr1 [39] , [40] , [41] . Inhibition of Pkc1 had a strikingly similar impact on azole resistance to inhibition of Hsp90 or calcineurin , reducing resistance of isolates recovered early during treatment to a greater extent than those recovered late during treatment ( Figure 8A ) . To further explore the relationship between stress response signaling and classic resistance mechanisms such as mutation of the drug target and overexpression of multidrug transporters , we characterized additional clinical isolates and laboratory-derived mutants . We tested an additional five sets of clinical isolates for which we had one isolate recovered early during azole treatment and one recovered later . In all cases , inhibition of Pkc1 phenocopied inhibition of Hsp90 , with the least effect on azole resistance of isolates that overexpressed the multidrug transporter Cdr1 ( Figure S6 ) . Since the clinical isolates often harbor multiple mechanisms of resistance , we also tested specific laboratory-derived resistant mutants . Inhibition of Pkc1 abolished resistance of laboratory-derived C . albicans and S . cerevisiae erg3 loss-of-function mutants ( Figure 8B ) , as does inhibition of Hsp90 or calcineurin [24] , [25] . In contrast , inhibition of Pkc1 did not affect S . cerevisiae resistance due to an activating mutation in the transcription factor Pdr1 that causes overexpression of multidrug transporters including Pdr5 ( Figure 8B ) , as was the case with inhibition of Hsp90 [25] . We previously confirmed that genetic compromise of Hsp90 function does not affect resistance due to overexpression of Pdr5 , confirming that the stability of this resistance phenotype is not due to Hsp90 inhibitors being pumped out of the cell [25] . Given the equivalent impact on azole resistance of Pkc1 inhibitors and Hsp90 inhibitors with diverse mutants , this strongly suggests that the stability of resistance of the Pdr1 mutant cannot be attributed to Pkc1 inhibitors being pumped out of the cell . Thus , inhibition of PKC signaling phenocopies inhibition of Hsp90 or its client protein calcineurin , reducing resistance of clinical isolates and specific resistant mutants . These results are consistent with the circuitry connecting PKC signaling and calcineurin delineated above and may also suggest an additional functional connection between Hsp90 and PKC signaling in regulating responses to ergosterol biosynthesis inhibitors . While our findings already establish a link between PKC signaling and calcineurin-mediated stress responses , we next explored the possibility of yet another functional connection between Hsp90 and PKC signaling . In S . cerevisiae , Hsp90 binds exclusively to the activated form of Slt2 and enables Slt2-mediated activation of the downstream target Rlm1 [64] . To determine if the connection between Hsp90 and the terminal MAPK is conserved in C . albicans , we tested the impact of genetic depletion of C . albicans HSP90 on Mkc1 levels and activation status . To deplete Hsp90 , we used a strain with its only HSP90 allele under the control of a doxycycline repressible promoter [65] . To monitor total Mkc1 levels , this kinase was tagged at the C-terminus using a 6x-histidine and FLAG epitope tag . The Mkc1-6xHis-FLAG protein was functional and sufficient to confer wild-type tolerance to ergosterol biosynthesis inhibitors ( Figure S7 ) . To determine whether Hsp90 stabilized only the activated form of Mkc1 , we used a strain lacking the upstream MAPKKK required for Mkc1 activation , Bck1 . All strains were grown in the presence of terbinafine to induce Mkc1 activation . In the absence of doxycycline ( Figure 9A , left panel ) , all strains had comparable levels of Hsp90 as measured relative to a histone H3 loading control . All strains also had comparable levels of activated dually-phosphorylated Mkc1 , with the exception of the strain lacking Bck1 in which Mkc1 activation was blocked . Total Mkc1 levels monitored by a 6X-histidine antibody were comparable for the three strains harboring the tagged protein . In the presence of doxycycline ( Figure 9A , right panel ) , Hsp90 levels were depleted only in the strains with the repressible promoter . Depletion of Hsp90 resulted in a corresponding depletion of total Mkc1 levels , even in the strain lacking Bck1 in which Mkc1 remains in the inactivate state . Depletion of Hsp90 did not affect MKC1 transcript levels as measured by quantitative RT-PCR ( P>0 . 05 , ANOVA , Bonferroni's Multiple Comparison Test , Figure S8 ) , confirming that the chaperone influences Mkc1 stability at the protein level . Thus , Hsp90 stabilizes Mkc1 independent of its activation status and thereby regulates PKC signaling , providing a new mechanism through which Hsp90 regulates drug-induced membrane stress responses ( Figure 9B ) . Given that deletion of PKC1 enhances the efficacy of antifungal drugs , we next explored the therapeutic efficacy in a well-established murine model in which fungal inoculum is delivered by tail vein injection and progresses from the bloodstream to deep-seated infection of major organs , most notably the kidney [27] , [65] , [66] . We compared kidney fungal burden of mice infected with either a wild-type strain or a pkc1Δ/pkc1Δ mutant . The average kidney fungal burden in mice infected with 1×105 CFUs of the wild-type parental strain was 4 . 34+/−0 . 54 log CFU per gram of kidney ( Figure 10A ) . In stark contrast , the kidneys of mice infected with 1×105 CFUs of the pkc1Δ/pkc1Δ mutant were sterile ( Figure 10A ) . To determine if infection with higher inocula of the pkc1Δ/pkc1Δ mutant would lead to sufficient kidney fungal burden to enable assessment of antifungal efficacy in vivo , we tested the impact of infection with 10-fold and 100-fold higher inocula . Mice infected with 1×106 or 1×107 CFUs of the pkc1Δ/pkc1Δ mutant demonstrated significantly reduced fungal burden relative to those infected with only 1×105 CFUs of the wild-type strain ( P<0 . 001 , ANOVA , Bonferroni's Multiple Comparison Test ) . The average kidney fungal burden in mice infected with 1×106 or 1×107 cells of the pkc1Δ/pkc1Δ mutant was 0 . 19+/−0 . 66 and 0 . 23+/−0 . 67 log CFU per gram of kidney , respectively . Thus , while C . albicans PKC1 is dispensable for growth under standard conditions in vitro it is required for proliferation and infection in a murine model , providing evidence for a key role of this stress response regulator in virulence . While the attenuated virulence of the pkc1Δ/pkc1Δ mutant precluded straightforward studies to determine if compromising Pkc1 enhances the efficacy of antifungal drugs in vivo , it provides compelling support for therapeutic potential of compromising fungal Pkc1 . Given our findings that Pkc1 and calcineurin affect drug resistance via a common target in C . albicans ( Figure 7 ) , it is possible that Pkc1-mediated signaling may influence virulence by a target in common with calcineurin , which is known to be required for C . albicans virulence [67] , [68] , [69] . Calcineurin mutants are hypersensitive to calcium present in serum and are unable to survive transit through the bloodstream [68] . However , while a mutant lacking the catalytic subunit of calcineurin was unable to survive on medium containing 50% serum , the pkc1Δ/pkc1Δ mutant exhibited only an intermediate reduction in viability and the mkc1Δ/mkc1Δ and bck1Δ/bck1Δ mutants grew as well as the wild type ( Figure 10B ) . Further , the pkc1Δ/pkc1Δ mutant grew as well as the wild-type strain in liquid serum while the calcineurin mutant was inviable ( data not shown ) . These results suggest that Pkc1 exerts powerful control over C . albicans virulence by means of targets distinct from calcineurin .
Our results establish a new role for the PKC signal transduction cascade in resistance to drugs targeting the cell membrane in the model yeast S . cerevisiae and the fungal pathogen C . albicans . Three out of seven hits from our screen of 1 , 280 pharmacologically active compounds for those that abrogate azole resistance are classified as inhibitors of PKC , suggesting a central role for this cellular regulator in azole resistance ( Figure 1 ) . Pharmacological inhibition of Pkc1 with two additional structurally distinct PKC inhibitors whose mode of action has been validated in fungi or genetic compromise of Pkc1 function enhances sensitivity to azoles as well as other drugs targeting ergosterol biosynthesis , including allylamines and morpholines ( Figures 1 , 2 and 3 ) . Pkc1 regulates responses to ergosterol biosynthesis inhibitors at least in part through the MAPK cascade in both species ( Figures 2 and 3 ) . In S . cerevisiae , signaling through the MAPK cascade is required for calcineurin activation suggesting that PKC signaling regulates crucial responses to ergosterol biosynthesis inhibitors through calcineurin in this species ( Figure 6 ) . In C . albicans , Pkc1 and calcineurin independently regulate responses to ergosterol biosynthesis inhibitors via a common target ( Figure 7 ) . Inhibition of Pkc1 phenocopies inhibition of calcineurin or Hsp90 , reducing drug resistance of clinical isolates of C . albicans ( Figure 8 and S6 ) . We establish an additional level of regulatory complexity in the cellular circuitry linking PKC signaling , Hsp90 , and calcineurin in that genetic reduction of C . albicans Hsp90 results in destabilization of the terminal MAPK , Mkc1 , thereby blocking PKC signaling ( Figure 9 ) . This suggests that Hsp90 regulates basal tolerance and resistance to ergosterol biosynthesis inhibitors through Mkc1 in addition to the established connection with calcineurin . Our findings that compromising Pkc1 renders fungistatic drugs fungicidal ( Figures 2 and 3 ) and attenuates virulence of C . albicans ( Figure 10 ) suggest broad therapeutic potential . The role of PKC signaling in basal tolerance and resistance to drugs targeting the cell membrane expands the repertoire of stress responses that depend upon this signal transduction cascade . In S . cerevisiae , it was previously appreciated that PKC signaling is required for basal tolerance to echinocandins , which target cell wall synthesis [35] , [36] . This tolerance requires activation of the terminal MAPK Slt2 to drive Rlm1-dependent transcription of cell wall genes [36] . In C . albicans , the PKC pathway is activated by diverse stresses [52] and works in concert with calcineurin and the high osmolarity glycerol pathway to regulate chitin synthesis , which can enhance tolerance to echinocandins [37] , [70] . As is the case with drugs compromising cell wall integrity , drugs targeting the cell membrane activate the terminal MAPK in the PKC cascade ( Figure S3A ) . The role of PKC signaling in tolerance to drugs targeting the cell wall and the cell membrane raises the possibility that induction of cell membrane stress by ergosterol biosynthesis inhibitors could induce cell wall stress indirectly . This is consistent with the thought that the sensors involved in PKC cell wall integrity signaling are receptors that respond to changes in the structure of the cell membrane [71] . Despite the commonalities , the downstream regulation mediating responses to these different stresses diverge . Response to cell wall stress is largely dependent on the transcription factor Rlm1 [36] , while regulation of cell membrane stress responses is largely independent of Rlm1 . In S . cerevisiae , distinct downstream effectors contribute to tolerance to different ergosterol biosynthesis inhibitors ( Figure 4 ) . For fluconazole , the SBF transcription factor ( Swi4/Swi6 ) is of central importance . For fenpropimorph , the SBF complex again is a major determinant with Rlm1 enabling responses important in the absence of SBF . For terbinafine , Swi4 enables tolerance largely independent of SBF and Rlm1 and Cch1-Mid1 mediate responses important in the absence of Swi4 . These differences may be due to ergosterol depletion combined with the specific sterol that accumulates when ergosterol biosynthesis is inhibited at different points . In C . albicans , SBF and Cch1-Mid1 confer tolerance to all three ergosterol biosynthesis inhibitors tested suggesting that the point of inhibition of ergosterol biosynthesis has less impact than for S . cerevisiae . The circuitry downstream of Pkc1 mediating membrane stress responses has been rewired considerably between S . cerevisiae and C . albicans . For S . cerevisiae , deletion of components of the MAPK cascade confers hypersensitivity to ergosterol biosynthesis inhibitors at all temperatures tested ( Figure 2 and Figures S1 and S4 ) . For C . albicans , deletion of components of the MAPK cascade confers hypersensitivity to ergosterol biosynthesis inhibitors at 30°C ( Figure 3 ) but not at 35°C ( Figure S3B ) , suggesting that the MAPK cascade is a key mediator of Pkc1-dependent cell membrane stress responses but that alternate downstream effectors play a dominant role in C . albicans at elevated temperature . The importance of alternate downstream effectors of Pkc1 in C . albicans is further emphasized as deletion of PKC1 renders fungistatic ergosterol biosynthesis inhibitors fungicidal , while deletion of MAPK components does not ( Figure 3B ) . Our findings highlight another divergence between the two species . While inhibition of PKC signaling blocks calcineurin activation in response to ergosterol biosynthesis inhibitors in S . cerevisiae ( Figure 6 ) , this is not the case in C . albicans . Rather , our results suggest that Pkc1 and calcineurin independently regulate responses to ergosterol biosynthesis inhibitors via a common target in C . albicans ( Figure 7 ) . As with PKC signaling , calcineurin and Hsp90 regulate resistance to drugs targeting the cell membrane in both C . albicans and S . cerevisiae , however , they regulate responses to echinocandins in C . albicans but not S . cerevisiae [27] , [66] , suggesting both conservation and divergence in circuitry governing fungal drug resistance . The cellular circuitry linking PKC signaling , Hsp90 , and calcineurin is complex with multiple levels of regulatory control . On one level is the connection between PKC signaling and calcineurin , which is divergent between the two species . In S . cerevisiae , inhibition of Pkc1 blocks calcineurin activation . The terminal MAPK Slt2 has been found to activate the Cch1-Mid1 high-affinity Ca2+ channel in response to endoplasmic reticulum stress , thereby enabling calcineurin activation [56] . However , we found that deletion of this channel had little impact on drug tolerance ( Figure 4 ) , implicating calcineurin regulation via a distinct mechanism . Since inhibition of PKC signaling does not affect calcineurin expression ( Figure S5 ) , Slt2 likely regulates calcineurin activation by an alternative mechanism such as through a distinct calcium channel . In C . albicans , cch1Δ/cch1Δ and mid1Δ/mid1Δ mutants share some but not all phenotypes with a calcineurin mutant [59] . Consistent with this , we found that the cch1Δ/cch1Δ and mid1Δ/mid1Δ mutants are almost as sensitive to fluconazole and fenpropimorph as a calcineurin mutant but only show an intermediate sensitivity to terbinafine ( Figure 5 ) . In C . albicans , inhibition of PKC signaling did not block calcineurin function ( Figure 7 ) . Our findings support a model in which Pkc1 and calcineurin independently regulate responses to ergosterol biosynthesis inhibitors in C . albicans via a common target that remains to be identified . On another level is the connection between Hsp90 and the terminal MAPK . In S . cerevisiae , Hsp90 interacts with activated Slt2 and enables activation of Slt2 targets including Rlm1 [64] . In C . albicans , Hsp90 stabilizes Mkc1 independent of its activation status ( Figure 9 ) . Notably , in S . cerevisiae Hsp90 also chaperones Pkc1 [72] , though this has yet to be investigated in C . albicans . In contrast to the extensive Hsp90 network in S . cerevisiae [73] , [74] , we identify Mkc1 as the second Hsp90 client protein in C . albicans . Our work suggests that Hsp90 regulates responses crucial for survival of drug-induced membrane stress through PKC signaling in addition to the established role through calcineurin [24] , [25] , [27] . These stress responses are less important for resistance due to overexpression of multidrug transporters but are critical for basal tolerance as well as resistance acquired by other diverse mutations . Future experiments will address the relative contribution of calcineurin and PKC signaling via the MAPK cascade in Hsp90-mediated resistance acquired by diverse mechanisms . Our results highlight the central importance of fungal stress response pathways in enabling survival in the hostile host environment . We demonstrate that while deletion of PKC1 has little impact on growth in vitro , it drastically attenuates the capacity of C . albicans to proliferate in vivo and cause disease ( Figure 10 ) . While the attenuated virulence precludes studies to determine if compromising Pkc1 enhances the efficacy of antifungals in vivo , it provides compelling support for targeting fungal Pkc1 as a strategy to control fungal infections . The specific mechanism by which Pkc1 enables virulence has yet to be determined , however , it may operate in part via the downstream MAPK cascade given that C . albicans Mkc1 also contributes to virulence in a murine model [75] . While Mkc1 has little impact on susceptibility to oxidative-mediated killing by phagocytes [76] , it is activated by physical contact and is required for invasive hyphal growth and normal biofilm development [77] . The mechanism by which Pkc1 influences virulence is distinct from calcineurin , which is required for C . albicans virulence and survival in the bloodstream [67] , [68] , [69] . While the calcineurin mutant is unable to survive in serum , the pkc1Δ/pkc1Δ mutant only exhibits an intermediate reduction in viability ( Figure 10B ) , suggesting that Pkc1 regulates virulence via alternate targets . Notably , Pkc1 controls the expression of numerous virulence determinants in the fungal pathogen Cryptococcus neoformans [78] , suggesting that Pkc1 governs virulence in phylogenetically diverse fungal species . Our results suggest that targeting Pkc1 may provide a powerful strategy for the treatment of fungal infectious disease . In vitro , compromising PKC signaling renders laboratory strains and clinical isolates hypersensitive to drugs targeting ergosterol biosynthesis ( Figures 1 , 2 , 3 , and 8 ) . These findings coupled with those established by others linking PKC signaling to tolerance of drugs targeting the cell wall [34] , [35] , [36] , [37] , suggest that compromising Pkc1 could have therapeutic benefits by enhancing the efficacy of the two most widely deployed classes of antifungals , the azoles and echinocandins . In a murine model of disseminated candidiasis , deletion of PKC1 attenuates C . albicans virulence ( Figure 10A ) , suggesting therapeutic benefit of simply compromising fungal Pkc1 in addition to the benefits of combinatorial therapeutic strategies . Notably , in mammalian cells disruption of PKC signaling impairs tumor progression and drug resistance such that PKC inhibitors have entered clinical trials for the treatment of several human cancers as single or combination therapy agents [79] , [80] . The complexity of functions and interactions of mammalian PKC isoforms poses a challenge for the development of anti-cancer therapeutics and current efforts focus on enhancing specificity of action to target specific isoforms . While C . albicans and other fungal pathogens only have one PKC isoform , the therapeutic challenge will lie in achieving fungal selectivity . The successful development of Hsp90 and calcineurin as therapeutic targets for fungal disease faces similar challenges due to complications of inhibiting the function of these key cellular regulators in the host [66] , [81] , [82] . As a complement to identifying fungal selective pharmacological agents , elucidating the architecture of cellular circuitry governing stress responses , drug resistance , and virulence is poised to reveal promising therapeutic targets as key points of regulatory control that diverged between pathogen and host .
All procedures were approved by the Institutional Animal Care and Use Committee ( IACUC ) at Duke University according to the guidelines of the Animal Welfare Act , The Institute of Laboratory Animal Resources Guide for the Care and Use of Laboratory Animals , and Public Health Service Policy . Archives of C . albicans and S . cerevisiae strains were maintained at −80°C in 25% glycerol . Strains were grown in either yeast peptone dextrose ( YPD , 1% yeast extract , 2% bactopeptone , 2% glucose ) or in synthetic defined medium ( SD , 0 . 67% yeast nitrogen base , 2% glucose ) and supplemented with amino acids or in RPMI medium 1640 ( Gibco SKU#318000-089 , 3 . 5% MOPS , 2% glucose , pH 7 . 0 ) supplemented with amino acids . 2% agar was added for solid media . Strains were transformed following standard protocols . Strains used in this study are listed in Table S1 . Strain construction is described in Text S1 . Recombinant DNA procedures were performed according to standard protocols . Plasmids used in this study are listed in Table S2 . Plasmid construction is described in Text S1 . Plasmids were sequenced to verify the absence of any nonsense mutations . Primers used in this study are listed in Table S3 . A seed culture of the fungus , Mycosphaerella ( Cercosporidium ) henningsii ( IMI 176827 ) grown on potato dextrose agar ( PDA ) for two weeks was used for inoculation . Mycelia were scraped out and mixed with 20 mL sterile water and filtered through a 100 µm filter . Absorbance of the spore suspension was measured and adjusted to 0 . 4 . A 2 L Erlenmeyer flask containing 1 L of M-1-D medium [83] was inoculated with 10 mL of the spore suspension and incubated at 160 rpm and 28°C for four weeks . Mycelia were then separated from the supernatant by filtration through Whatman No . 1 filter paper and the filtrate was extracted with EtOAc ( 6×500 mL ) . The combined EtOAc extracts were washed with H2O ( 3×500 mL ) , dried over anhydrous Na2SO4 and evaporated under reduced pressure to yield a dark brown semi-solid ( 51 . 2 mg ) . A portion ( 50 . 0 mg ) of the EtOAc extract was separated on preparative TLC ( Merck , TLC silica gel 60 F254 precoated Aluminum sheets ) using MeOH/Et2O ( 3∶97 ) as eluant affording crude cercosporamide ( 8 . 1 mg , Rf 0 . 4 ) . This was further purified by reversed-phase preparative TLC ( Merck , TLC Silica gel 60 RP-18 F254 precoated Aluminium sheets ) using H2O/CH3CN ( 3∶7 ) as eluant to give pure cercosporamide ( 4 . 5 mg , Rf 0 . 5 ) . Cercosporamide: red crystals; mp 187–188°C ( lit . [46] 188–189°C ) ; APCIMS ( + ) -ve mode , m/z 331 [M+1]+; 1H and 13C NMR spectroscopic data were consistent with those reported in the literature [46] . The structure of cercosporamide is shown in Figure S9 . Antifungal tolerance and resistance were determined in flat bottom , 96-well microtiter plates ( Sarstedt ) using a modified broth microdilution protocol as described [25] , [27] . Dimethyl sulfoxide ( DMSO , Sigma Aldrich Co . ) was the solvent for fenpropimorph ( FN , Sigma Aldrich Co ) and terbinafine ( TB , Sigma Aldrich Co . ) ; fluconazole ( FL , Sequoia Research Products ) and micafungin ( MF , generously provided by Julia R . Köhler ) were dissolved in sterile ddH2O . Geldanamycin ( GdA , A . G . Scientific , Inc . ) was used to inhibit Hsp90 at the indicated concentrations . Cyclosporin A ( CsA , Calbiochem ) was used to inhibit calcineurin at the indicated concentrations . Cercosporamide and staurosporine ( STS , A . G . Scientific , Inc . ) were used to inhibit protein kinase C at the indicated concentrations . DMSO was the solvent for GdA , CsA , STS , and cercosporamide . Minimum inhibitory concentration ( MIC ) tests were set up in a total volume of 0 . 2 ml/well with 2-fold dilutions of FL , FN , TB and cercosporamide . FL gradients were from 256 µg/ml down to 0 with the following concentration steps in µg/ml: 256 , 128 , 64 , 32 , 16 , 8 , 4 , 2 , 1 , 0 . 5 , 0 . 25 . FN gradients were from 25 µg/ml down to 0 with the following concentration steps in µg/ml: 25 , 12 . 5 , 6 . 25 , 3 . 125 , 1 . 5625 , 0 . 78125 , 0 . 390625 , 0 . 1953125 , 0 . 09765625 , 0 . 04882813 , 0 . 02441406 . TB gradients were from 250 µg/ml with the following concentration steps in µg/ml: 250 , 125 , 62 . 5 , 31 . 25 , 15 . 625 , 7 . 8125 , 3 . 90625 , 1 . 953125 , 0 . 9765625 , 0 . 48828125 , 0 . 24414063 . Cercosporamide gradients were from 100 µg/ml with the following concentration steps in µg/ml: 100 , 50 , 25 , 12 . 5 , 6 . 25 , 3 . 125 , 1 . 5625 , 0 . 78125 , 0 . 390625 , 0 . 1953125 , 0 . 09765625 . Cell densities of overnight cultures were determined and dilutions were prepared such that ∼103 cells were inoculated into each well . Plates were incubated in the dark at 30°C or 35°C for the period of time indicated in the figure legend , at which point plates were sealed with tape and re-suspended by agitation . Absorbance was determined at 600 nm using a spectrophotometer ( Molecular Devices ) and corrected for background from the corresponding medium . Each strain was tested in duplicate on at least 3 occasions . MIC data was quantitatively displayed with color using the program Java TreeView 1 . 1 . 1 ( http://jtreeview . sourceforge . net ) . Checkerboard assays were set up in a total volume of 0 . 2 ml/well with 2-fold dilutions of cyclosporin A across the x-axis of the plate and 2-fold dilutions of STS across the y-axis of the plate . STS gradients were from 0 . 5 µg/ml to 0 in the following concentrations steps in µg/ml: 0 . 5 , 0 . 25 , 0 . 125 , 0 . 0625 , 0 . 03125 , 0 . 015625 , 0 . 0078125 . CsA gradients were from 48 µg/ml down to 0 in the following concentration steps in µM: 48 , 24 , 12 , 6 , 3 , 1 . 5 , 0 . 75 , 0 . 375 , 0 . 1875 , 0 . 09375 , 0 . 046875 . Plates were inoculated and growth was measured as with MIC tests . To test for synergy , the fractional inhibitory concentration ( FIC ) was calculated as follows: [ ( MIC80 of drug A in combination ) / ( MIC80 of drug A alone ) ] + [ ( MIC80 of drug B in combination ) / ( MIC80 of drug B alone ) ] . Values of ≤0 . 5 indicate synergy , those of >0 . 5 but <2 indicate no interaction and those ≥2 indicate antagonism . Strains were grown overnight to saturation in indicated media and cell concentrations were determined based on cell counts using a hemacytometer ( Hausser Scientific ) . Five-fold serial dilutions of cell suspensions starting at indicated concentrations ( 105 or 107cells/ml ) were performed in sterile ddH2O or sterile phosphate buffered saline . Cell suspensions were spotted onto indicated media using a spotter ( Frogger , V&P Scientific , Inc ) . Plates were photographed after 3 days in the dark at indicated temperature . For S . cerevisiae , MIC assays with two-fold dilutions of FL , FN , or TB were performed in SD as described above . For FL the gradients were from 256 µg/ml down to 0 with the following concentration steps in µg/ml: 256 , 128 , 64 , 32 , 16 . FN gradients were from 100 µg/ml down to 0 with the following concentration steps in µg/ml: 100 , 50 , 25 , 12 . 5 , 6 . 25 . TB gradients were from 250 µg/ml with the following concentration steps in µg/ml: 250 , 125 , 62 . 5 , 31 . 25 , 15 . 62 . Plates were incubated for two days at 35°C . Cells from the MIC assay were spotted onto solid YPD medium and incubated at 30°C for two days before they were photographed . For C . albicans , MIC assays with FL , FN , or TB were performed in YPD as described above with the following modification; four-fold dilutions of each drug were tested . For FL the gradients were from 256 µg/ml down to 0 with the following concentration steps in µg/ml: 256 , 64 , 16 , 4 , 1 . FN gradients were from 25 µg/ml down to 0 with the following concentration steps in µg/ml: 25 , 6 . 25 , 1 . 5625 , 0 . 390625 , 0 . 09765625 . TB gradients were from 250 µg/ml with the following concentration steps in µg/ml: 250 , 62 . 5 , 15 . 625 , 3 . 90625 , 0 . 9765625 . Plates were incubated for two days at 35°C . Cells from the MIC assay were spotted onto solid YPD medium and incubated at 30°C for two days before they were photographed . S . cerevisiae cultures were grown overnight at 25°C in SD medium supplemented for auxotrophies . Cells were diluted to OD600 of 0 . 05 and were either left untreated or were treated with FL ( 16 µg/ml ) , FN ( 1 µg/ml ) , or TB ( 25 µg/ml ) for 24 hours at 25°C . When STS was used as an inhibitor in the assay , cultures were grown overnight in SD at 25°C and diluted to OD600 of 0 . 05 in SD with or without STS ( 2 . 5 µg/ml ) for 24 hours at 25°C . Cells were then diluted to OD600 of 0 . 05 in SD with or without STS and with or without FL ( 32 µg/ml ) for an additional 24 hours at 25°C . Cells were harvested , washed , protein was extracted , and protein concentrations were determined by Bradford analysis as described [27] . Protein samples were diluted to the same concentration and β-galactosidase activity was measured using the substrate ONPG ( O-nitrophenyl-β-D-galactopyranosidase , Sigma Aldrich Co . ) as described [27] . β-galactosidase activity is given in units of nanomoles ONPG converted per minute per milligram of protein . Statistical significance was evaluated using GraphPad Prism 4 . 0 . For the Mkc1 activation assay , yeast cultures were grown overnight in YPD at 30°C . In the morning , cells were diluted to OD600 of 0 . 2 in 50 mL YPD and were grown to mid-log ( ∼3 hours ) at 30°C and then cultures were split into 5×10 mL cultures and were either left untreated or were treated with FL ( 8 µg/mL ) , FN ( 1 µg/mL ) , MF ( 30 ng/mL ) , or TB ( 25 µg/ml ) for 2 hours at 30°C . Cells were harvested by centrifugation at 1308×g for 10 minutes at 4°C and were washed with sterile cold phosphate buffered saline ( PBS ) . Cell pellets were resuspended in lysis buffer containing 50 mM HEPES pH 7 . 4 , 150 mM NaCl , 5 mM EDTA , 1%Triton ×100 , 50 mM NaF , 10 mM Na3VO4 , 1 mM PMSF , and protease inhibitor cocktail ( complete , EDTA-free tablet , Roche Diagnostics ) . For the Mkc1 destabilization assay , cultures were grown overnight in YPD at 30°C . In the morning , cells were diluted to OD600 of 0 . 2 in 10 mL YPD with or without doxycycline ( 20 µg/mL; BD Biosciences ) and left at 30°C for 24 hours . Cells were diluted once again to OD600 of 0 . 2 in the same treatment conditions as overnight and were grown at 30°C until mid-log phase ( ∼4 hours ) . Doxycycline reduces the growth rate of strains with the repressible promoter driving expression of the only HSP90 allele but does not affect stationary phase cell density [65] . Cells were then treated with 50 µg/mL TB for 3 hours at 30°C to elicit phosphorylation of Mkc1 . Cells were harvested after TB treatment at 1308×g at 4°C and washed with ice-cold ddH2O . Cell pellets were flash frozen in liquid N2 , resuspended in lysis buffer ( 50 mM HEPES pH 7 . 4 , 150 mM NaCl , 5 mM EDTA , 1% Triton ×100 , 100 mM NaF , 20 mM Na3VO4 , 1 mM PMSF and protease inhibitor cocktail complete , EDTA-free tablet , Roche Diagnostics ) . Cells suspended in lysis buffer were mechanically disrupted by adding acid-washed glass beads and bead beating for 1 minute for six cycles with 1 minute on ice between each cycle . Protein concentrations were determined by Bradford analysis . Protein samples were mixed with one-sixth volume of 6× sample buffer containing 0 . 35 M Tris-HCl , 10% ( w/w ) SDS , 36% glycerol , 5% β-mercaptoethanol , and 0 . 012% bromophenol blue for SDS-PAGE . Samples were boiled for 5 minutes and then separated by 10% SDS-PAGE . Protein was electrotransferred to PVDF membrane ( Bio-Rad Laboratories , Inc . ) and blocked with 5% skimmed milk in PBS with 0 . 1% tween or 5% bovine serum albumin in phosphate buffered saline with 0 . 1% tween . Blots were hybridized with antibodies against CaHsp90 ( 1∶10000 dilution , generously provided by Brian Larsen , [84] ) , histone H3 ( 1∶3000 dilution; Abcam ab1791 ) , His6 ( 1∶10 , P5A11 , generously provided by Elizabeth Wayner ) and phospho-p44/42 MAPK ( Thr202/Tyr204 ) ( 1∶2000 , Cell Signaling ) . To monitor gene expression changes in response to FL treatment in S . cerevisiae , cells were grown overnight in SD supplemented for auxotrophies at 30°C . Cells were diluted to OD600 of 0 . 1 in SD and grown for 2 hours in duplicate at 25°C . After 2 hours of growth 16 µg/mL FL was added to one of the two duplicate cultures and left to grow for an additional 4 hours at 25°C . Cell pellets were frozen at −80°C immediately . To monitor gene expression changes in response to FL treatment in C . albicans , cells were grown overnight in YPD at 30°C . Cells were diluted to OD600 of 0 . 1 in YPD and grown for 2 hours in duplicate at 35°C . After 2 hours of growth 16 µg/mL FL was added to one of the two duplicate cultures and left to grow for an additional 4 hours at 35°C . Cell pellets were frozen at −80°C immediately . To monitor MKC1 transcript levels in response to decreased levels of Hsp90 , cultures were grown overnight in YPD at 30°C . In the morning , cells were diluted to OD600 of 0 . 2 in 10 mL YPD with or without 20 µg/mL doxycycline ( BD Biosciences ) and left at 30°C for 24 hours . The next morning , cells were diluted once again to OD600 of 0 . 2 in the same treatment conditions and were grown at 30°C until mid-log phase ( ∼4 hours ) . Cell pellets were collected and immediately frozen at −80°C . RNA was isolated using the QIAGEN RNeasy kit and RNAase-free DNase ( QIAGEN ) , and cDNA synthesis was performed using the AffinityScript cDNA synthesis kit ( Stratagene ) . PCR was performed using SYBR Green JumpStart Taq ReadyMix ( Sigma-Aldrich Co . ) with the following cycling conditions: 94°C for 2 minutes , 94°C for 15 seconds , 60°C for 1 minute , 72°C for 1 minute , for 30 or 40 cycles . All reactions were performed in triplicate , using primers for the following genes: CaGPD1 ( oLC752/753 ) , CaHSP90 ( oLC754/755 ) , ScACT1 ( oLC1015/1016 ) , ScCNA1 ( oLC1286/1287 ) , ScCNA2 ( oLC1288/1289 ) , ScCNB1 ( oLC1290/1291 ) , CaCNB1 ( oLC1292/1293 ) , CaCNA1 ( oLC1294/1295 ) , ScCRZ1 ( oLC1328/1329 ) , CaCRZ1 ( oLC1330/1331 ) , CaMKC1 ( oLC1332/1333 ) , CaPLC3 ( oLC1432/1433 ) , and CaUTR2 ( oLC1434/1435 ) . Data were analyzed using iQ5 Optical System Software Version 2 . 0 ( Bio-Rad Laboratories , Inc . ) . Statistical significance was evaluated using GraphPad Prism 4 . 0 . Inoculum was prepared as described [20] , [27] , [65] . Cultures were started from frozen stocks onto Sabouraud dextrose agar plates and incubated at 35°C for 48 hours . Colonies were suspended in sterile pH 7 . 4 PBS , centrifuged at 324×g for 5 minutes , washed with sterile PBS one time and diluted to the desired concentration as verified by counting on a Neubauer hematocytometer as well as by serial dilution and culture . Male CD1 mice ( Charles River Laboratories , Wilmington , MA ) age 8 weeks ( weight 30–34 g ) were infected via the tail vein with 100 µL of a 1×106 CFU/mL suspension of the wild type strain ( CaLC239 , 1×105 CFU per mouse , n = 9 mice ) , an inoculum previously determined to produce morbidity but not mortality when using C . albicans strain SC5314 at 4 days following tail vein injection ( Zaas et al . unpublished data ) . We observed discordance between cell counts and CFU measurements for the pk1cΔ/pkc1Δ mutant , such that CFU values were ∼50% lower than expected; thus , inocula for the pk1cΔ/pkc1Δ mutant were prepared at higher concentrations based on cell counts and the effective concentrations in CFUs were confirmed by dilution plating . For infection with the pk1cΔ/pkc1Δ mutant , we used an inoculum equivalent to that for the wild type ( 1×105 CFU , n = 8 mice ) as well a 10-fold and 100-fold increase in inoculum ( 1×106 , n = 11 mice and 1×107 CFU , n = 8 mice ) . Mice were observed three times daily for signs of illness and weighed daily . At day 4 following injection , mice were sacrificed using CO2 asphyxiation and the left kidney was removed aseptically , placed in sterile PBS , homogenized using a FastPrep 120 ( QBiogene ) using 0 . 5 mm zirconium beads ( Biospec , Inc . ) for 1 minute and serial dilutions plated for determination of kidney fungal burden . The CFU values in kidneys were expressed as CFU/g of tissue and log-transformed . Statistical significance was evaluated using GraphPad Prism 4 . 0 . S . cerevisiae: PKC1 ( 852169 ) ; HSC82 ( 855224 ) ; HSP82 ( 855836 ) ; CNA1 ( 851153 ) ; CNA2 ( 854946 ) ; CNB1 ( 853644 ) ; ERG11 ( 856398 ) ; ERG3 ( 850745 ) ; RHO1 ( 856294 ) ; BCK1 ( 853350 ) ; MKK1 ( 854406 ) ; MKK2 ( 855963 ) ; SLT2 ( 856425 ) ; SWI4 ( 856847 ) ; SWI6 ( 850879 ) ; ERG2 ( 855242 ) ; ERG24 ( 855441 ) ; ERG1 ( 853086 ) ; RLM1 ( 856016 ) ; CCH1 ( 853131 ) ; MID1 ( 855425 ) ; CRZ1 ( 855704 ) ; PDR5 ( 854324 ) ; PDR1 ( 852871 ) ; PDR3 ( 852278 ) ; ACT1 ( 850504 ) . C . albicans PKC1 ( 3635298 ) ; HSP90 ( 3637507 ) ; CNA1 ( 3639406 ) ; CNB1 ( 3636463 ) ; MKC1 ( 3639710 ) ; ERG11 ( 3641571 ) ; ERG3 ( 3644776 ) ; RHO1 ( 3642564 ) ; BCK1 ( 3641434 ) ; MKK2 ( 3645580 ) ; MDR1 ( 3639260 ) ; ERG2 ( 3639416 ) ; ERG24 ( 3648198 ) ; ERG1 ( 3646509 ) ; CEK1 ( 3642789 ) ; CEK2 ( 3642459 ) ; RLM1 ( 3635703 ) ; SWI4 ( 3645507 ) ; SWI6 ( 3634957 ) ; CCH1 ( 3639950 ) ; MID1 ( 3647441 ) ; CRZ1 ( 3641722 ) ; PLC3 ( 3635941 ) ; UTR2 ( 3636747 ) ; CDR1 ( 3635385 ) ; GPD1 ( 3643986 ) .
|
Treating fungal infections is challenging due to the emergence of drug resistance and the limited number of clinically useful antifungal drugs . We screened a library of 1 , 280 pharmacologically active compounds to identify those that reverse resistance of the leading human fungal pathogen , Candida albicans , to the most widely used antifungals , the azoles . This revealed a new role for protein kinase C ( PKC ) signaling in resistance to drugs targeting the cell membrane , including azoles , allylamines , and morpholines . We dissected mechanisms through which PKC regulates resistance in C . albicans and the model yeast Saccharomyces cerevisiae . PKC enabled survival of cell membrane stress at least in part through the mitogen-activated protein kinase ( MAPK ) cascade in both species . In S . cerevisiae , inhibition of PKC signaling blocked activation of a key regulator of membrane stress responses , calcineurin . In C . albicans , Pkc1 and calcineurin independently regulate resistance via a common target . Deletion of C . albicans PKC1 rendered fungistatic drugs fungicidal and reduced virulence in a mouse model . The molecular chaperone Hsp90 , which stabilizes client proteins including calcineurin , also stabilized the terminal C . albicans MAPK , Mkc1 . We establish new circuitry connecting PKC with Hsp90 and calcineurin and suggest a promising strategy for treating life-threatening fungal infections .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/fungal",
"infections",
"infectious",
"diseases/antimicrobials",
"and",
"drug",
"resistance"
] |
2010
|
PKC Signaling Regulates Drug Resistance of the Fungal Pathogen Candida albicans via Circuitry Comprised of Mkc1, Calcineurin, and Hsp90
|
Although leptospirosis is a zoonosis of major concern on tropical islands , the molecular epidemiology of the disease aiming at linking human cases to specific animal reservoirs has been rarely explored within these peculiar ecosystems . Five species of wild small mammals ( n = 995 ) as well as domestic animals ( n = 101 ) were screened for Leptospira infection on Reunion Island; positive samples were subsequently genotyped and compared to Leptospira from clinical cases diagnosed in 2012–2013 ( n = 66 ) , using MLST analysis . We identified two pathogenic species in human cases , namely Leptospira interrogans and Leptospira borgpetersenii . Leptospira interrogans was by far dominant both in clinical samples ( 96 . 6% ) and in infected animal samples ( 95 . 8% ) , with Rattus spp and dogs being its exclusive carriers . The genetic diversity within L . interrogans was apparently limited to two sequence types ( STs ) : ST02 , identified among most clinical samples and in all rats with complete MLST , and ST34 , identified in six humans , but not in rats . Noteworthy , L . interrogans detected in two stray dogs partially matched with ST02 and ST34 . Leptospira borgpetersenii was identified in two clinical samples only ( 3 . 4% ) , as well as in cows and mice; four haplotypes were identified , of which two seemingly identical in clinical and animal samples . Leptospira borgpetersenii haplotypes detected in human cases were clearly distinct from the lineage detected so far in the endemic bat species Mormopterus francoismoutoui , thus excluding a role for this volant mammal in the local human epidemiology of the disease . Our data confirm rats as a major reservoir of Leptospira on Reunion Island , but also pinpoint a possible role of dogs , cows and mice in the local epidemiology of human leptospirosis . This study shows that a comprehensive molecular characterization of pathogenic Leptospira in both clinical and animal samples helps to gaining insight into leptospirosis epidemiology within a specific environmental setting .
Leptospirosis is a bacterial systemic infection , occasionally fatal , caused by the pathogenic spirochetes of the genus Leptospira . Though claimed as the most widespread zoonosis in the world [1 , 2] , the disease is considered as emerging in many parts of the world . Leptospirosis is most prevalent in tropical and subtropical countries [3 , 4] , presumably because the survival of the bacterium outside the host requires humid and warm conditions that are typical of tropical areas [5] . Rodents are recognized as the main reservoir of pathogenic Leptospira although several animal species are also capable of sustaining biofilm colonization of the renal tubules and shedding the bacteria in their urine [6] . Humans usually get infected through indirect exposure with water or soil contaminated with urine , but direct transmission has also been suggested as important for some species [7] . Traditionally , Leptospira have been classified serologically into 25 serogroups and over 300 serovars using Microscopic Agglutination Test ( MAT ) and Cross agglutination absorption test ( CAAT ) analysis , respectively [1 , 8] . More recently , a genetic classification based on DNA-DNA hybridization complemented by molecular methods and experimental studies have confirmed the existence of at least 22 Leptospira species [9] . The congruence between serological and molecular classifications is poor [1]: one serogroup can be linked to several Leptospira species while serovars can vary within a given clone or lineage , an observation considered as probably resulting from horizontal gene transfer [10] . Among the various molecular tools currently used to genotype Leptospira , multilocus sequence typing ( MLST ) has emerged as a method of choice as it provides data produced at a local scale that can be further compared to genotypes obtained all over the world and made freely accessible to the scientific community through public databases . Leptospirosis is endemic in several islands of the southwestern Indian Ocean ( SWIO ) including the two French overseas departments of Mayotte [11 , 12] and Reunion Island [13 , 14] together with the island state of Seychelles [14–16] , the latter recording the highest human incidence reported worldwide [17] . On Madagascar , serological evidence of exposure to leptospirosis has been reported in the human community of Moramanga [18] , while to our knowledge , a single case of acute leptospirosis infection has been PCR-confirmed on a traveller returning from Madagascar [19] . Only serological evidence of human exposure to pathogenic Leptospira has been reported on the three neighboring islands of Union of the Comoros [20] , while no or only scarce data is available for Mauritius [21] and Rodrigues Islands . On Mayotte , improved diagnostic procedures allowed to significantly increase the number of confirmed cases , with about 100 new human cases reported each year since 2009 [12 , 22] . The MLST analysis of Leptospira isolated from human incident cases on the island revealed a large bacterial diversity of clinical isolates , represented by at least 17 different “sequence types” ( STs ) and four distinct Leptospira species , namely Leptospira interrogans , Leptospira borgpetersenii , Leptospira kirschneri , and the newly described Leptospira mayottensis [12 , 23] . Comparatively , information is more abundant on pathogenic Leptospira infecting wild animals from the SWIO . On Mayotte , sequencing of a portion of the 16S rRNA ( rrs ) gene of leptospires infecting twenty black rats Rattus rattus has identified the same four Leptospira species previously reported in clinical cases from the same island , with a strict identity between the sequences of Leptospira infecting R . rattus and humans [11] , designating this rodent species as the probable major reservoir and transmission source of Leptospira to humans on Mayotte [11] . On Comoros and Madagascar , sequencing of the nearly complete 16S rRNA gene of Leptospira infecting kidneys from six bat species ( n = 7 ) has shown the carriage of L . interrogans ( Comoros ) , L . borgpetersenii ( Comoros and Madagascar ) and other thus far unknown Leptospira species [24] . On Madagascar , partial sequencing of rrs locus has identified a single L . interrogans haplotype among 70 samples from introduced small mammals ( rats , shrews and mice ) [25] . Only one full five gene-based MLST analysis has been reported so far from the SWIO region , i . e . on Madagascar , where authors identified L . borgpetersenii , L . mayottensis and L . kirschneri from endemic small mammals and bats [26] . This study revealed distinct clustering associated with host type , with no overlap between Leptospira species infecting endemic small mammals versus those infecting introduced ones , a feature which certainly deserves further investigation . On Reunion Island , human leptospirosis was first reported in 1953 but the first outbreak most likely occurred in 1868 [27] . Since 1953 , several studies have been conducted to assess the burden of leptospirosis on the island . From 2008 to 2012 , the mean annual incidence was estimated at 8 . 2 cases per 100 , 000 inhabitants with a fatality rate around 4% , and Icterohaemorrhagiae identified as the major serogroup in severe forms [13] . Lower prevalences of serogroups Canicola , Grippotyphosa and Australis have also been reported [28] . Pagès and colleagues [13] have highlighted the following population groups as being at highest risk: farmers and green space workers , people under 20 years old and participating in aquatic activities , people between 20 and 30 years old that fish , and people between 50 and 60 year old gardening at home . The seroprevalence of infection and/or Leptospira carriage in potential reservoirs , have been explored in wild small mammals ( rodents , shrews , tenrecs and bats ) and domestic animals ( dogs , cats , cattle , goats , swine , rusa deers and horses ) [14 , 29–32] . Since then , the black rat , Rattus rattus , abundant in most regions and biotopes of the island , has been considered to be the primary reservoir and transmitter of Leptospira spp . , a conclusion mainly based on the observation that Icterohaemorrhagiae is by far the main serogroup found in rats and clinical cases [29] , though the same serogroup has also been identified in dogs , pigs , rusa deer and tenrecs [29] . Recently , the first molecular study from Reunion Island identified that patients with acute leptospirosis ( n = 42 ) were all infected by L . interrogans [33] . Although the typing method , High Resolution Melting ( HRM ) analysis of two VNTR sequences , allows a rapid diagnosis on clinical samples together with the characterization at the serovar and species levels , it is likely not resolutive enough as to trace the source ( s ) leading to human infection and illness . The aim of our study was to determine the wild and/or domestic animal species that not only serve as reservoir hosts for pathogenic Leptospira species , but also are primary sources for human infection on Reunion Island , an area considered as endemic for human leptospirosis .
Sixty-six patients whose blood samples tested positive for Leptospira ( see details hereafter ) were included in the study . Sera were collected for diagnostic purpose from patients originated from all four sanitary regions recognized on the island and admitted for acute febrile syndrome in 2012–2013 to either of the two main University hospitals of La Reunion ( northern and southern University hospitals , hereafter referred to as NUH and SUH respectively ) . All patients included in the study were considered as autochthonous leptospirosis cases as none of them reported travel history during the month preceding the onset of symptoms . We trapped 799 rodents ( 562 Rattus rattus , 170 Rattus norvegicus , 67 Mus musculus domesticus ) , 171 shrews ( Suncus murinus ) and 25 tenrecs ( Tenrec ecaudatus ) , those species representing all terrestrial small mammal diversity occurring on Reunion Island . Trapping occurred in 2012–2014 during dry and rainy seasons on twenty sampling sites , fifteen of them selected along two altitudinal transects on wet windward and dry leeward coasts ( Fig 1 ) . Details on trapping of wild animals and study sites have been published elsewhere ( see [34] ) . Kidneys from cows ( n = 33 ) and pigs ( n = 22 ) were directly collected at the unique slaughterhouse of the island , which was not able to provide information regarding the geographic origin of the samples . Kidneys from stray dogs ( n = 45 ) were collected just after the animals were euthanized at the community pound of Saint André ( eastern coast ) . The only cat sample included in the study was urine collected by a veterinarian . The ethical terms of the research protocol were approved by the CYROI institutional ethical committee ( Comité d’Ethique du CYROI n°114 , IACUC certified by Ministry of Higher Education and Research ) and by the Ministry of higher education and research under accreditation 03387 . 01 ( LeptOI ) . All animal procedures carried out in our study were performed in accordance with the European Union legislation for the protection of animals used for scientific purposes ( Directive 2010/63/EU ) . The stray dogs were euthanized by lethal injection administered by a veterinarian in the frame of population control measures implemented by the local authorities . Residual sera from anonymized clinical samples collected for diagnostic purposes and laboratory-confirmed as leptospirosis cases were obtained from the clinical laboratories of NUH and SUH . Genomic DNA was extracted from human sera at hospital laboratories using NucliSENS easyMAG system ( BioMérieux , Marcy l’Etoile , France ) ( NUH ) or Dneasy Blood and Tissue Kit ( Qiagen , Courtaboeuf , France ) ( SUH ) according to the manufacturer’s instructions . The presence of Leptospira was assessed with a probe-specific real-time PCR targeting either ( i ) the 23S rRNA gene [35] in NUH or ( ii ) LA0322 locus ( LFB1 primers ) [36] in SUH . Extraction of total nucleic acids from animal samples was performed from a pool of kidney , lung and spleen tissues ( for wild animals ) or from kidneys only ( for domestic animals ) using the Biorobot EZ1 and EZ1 Virus Mini Kit version 2 . 0 ( QIAGEN , Les Ulis , France ) . A reverse transcription step was then performed with GoScript reverse transcriptase ( Promega , Charbonnières-les-Bains , France ) ; generated cDNA was used as template for Leptospira detection through a previously described probe-specific real-time PCR targeting the rrs ( 16S rRNA ) locus [24 , 37] . Animal samples leading to a PCR amplification at Ct < 42 were considered as positive . Twenty-four animal samples were randomly selected in order to attempt Leptospira isolation from kidney cultures , i . e . 15 R . rattus , two R . norvegicus , two M . musculus , three S . murinus and two T . ecaudatus . A small piece of the freshly sampled kidney was crushed under sterile conditions and used to inoculate three distinct culture media: ( i ) Ellinghausen-McCullough-Johnson-Harris ( EMJH ) liquid medium ( Difco , Detroit , MI , USA ) supplemented with Albumin Fatty Acid Supplement ( AFAS; Royal Tropical Institute , Amsterdam , Netherlands ) [38 , 39]; ( ii ) EMJH liquid medium supplemented with AFAS , rabbit serum and foetal calf serum ( 1% each ) ; and ( iii ) semisolid Fletcher medium ( Difco , Detroit , MI , USA ) supplemented with rabbit serum ( 8% ) . All media were supplemented with 5-fluorouracil ( 5-FU ) at a final concentration of 200 μg . mL-1 . Cultures were incubated at 28°C , visually checked for the presence of Leptospira using a dark field microscope once a week for four months , and positive cultures were further sub-cultured in fresh EMJH liquid medium deprived of 5-FU . DNA was extracted from 1 mL of each positive culture using the EZ1 Biorobot with Qiagen EZ1 DNA Tissue kits ( Qiagen , Les Ulis , France ) . Three major MLST schemes exist for Leptospira spp . typing worldwide , all supported by the website database http://pubmlst . org/leptospira/ . In our study , scheme #3 [40] was preferred to the other two [10 , 41] as several molecular data from the SWIO have been made freely available using this same scheme , hence allowing comparison of local STs with those from SWIO islands [12 , 26] . MLST was attempted for adk , icdA , lipL32 , rrs2 , secY , and lipL41 loci as previously described [40] and recently optimized [26] . All positive domestic animal samples were first submitted to secY sequencing , revealing a nearly clonal population among infected rats ( see Results ) . Hence , a randomly selected subgroup of positive wild animal samples was selected from eleven sites ( n = 17; Table 1 ) and submitted to full MLST genotyping . For species identification of other positive wild animal samples , including two culture isolates , the secY gene was amplified with the primers secY F/R [40] or alternatively G1/G2 [42] . When secY PCR amplification failed , a 16S rRNA gene sequencing was alternatively used with previously published primers , i . e . LA/LB [43] or LA/R2 [44] . Obtained amplicons were directly sent for direct Sanger sequencing on both strands ( Genoscreen , Lille , France ) , or alternatively cloned into the pGEM-T easy vector ( Promega , Madison , WI , USA ) when the quantity was insufficient for direct sequencing . Obtained clones ( five per sample ) were subsequently sequenced using M13 universal primers for Sanger sequencing . The overall analysis strategy is presented in Fig 2 . Consensus sequence for each sample and multiple alignments between sequences were obtained using Geneious Pro version 5 . 4 [45] . In order to provide a comparison with Leptospira strains occurring worldwide , our dataset included sequences from two human clinical isolates from Mayotte [12] and 118 human and animal isolates from all over the world [2] . Sequence alignments were constructed separately for all six considered loci . A five-loci concatenate was generated using SEQMATRIX v1 . 7 . 8 [46] , excluding icdA locus that was unavailable from the Mayotte clinical dataset and for some of our samples . Bayesian analyses were performed to infer phylogenetic relationships between Leptospira species . The best-fitting model and associated parameters were selected using jModelTest [47] and phylogenies were constructed by Bayesian inference . We performed two independent runs of Metropolis-coupled Markov chain Monte Carlo ( MCMCMC ) analyses in MrBayes v3 . 2 . 1 [48] of all loci independently and of concatenated sequences . Each run included 20 , 000 , 000 generations , and trees were sampled every 100 generations . The initial 20 , 000 trees were discarded as a conservative "burn-in" and the harmonic mean of the likelihood was calculated by combining the two independent runs . The 50% majority-rule consensus tree was then computed from the sampled trees under the best model . Neighbor-joining trees were constructed using Seaview v4 . 3 ( Kimura’s 2-parameter distances , 500 replicates ) . Trees were visualized in FigTree v1 . 3 . 1 ( http://tree . bio . ecd . ac . uk/ ) . GenBank accession numbers of the sequences produced in the frame of the present study are provided as additional table 1 ( S1 Table ) . Unique allele identifiers for all six loci were assigned , and corresponding allelic profiles ( or sequence types STs ) were defined using the established Leptospira MLST website ( http://pubmlst . org ) , focusing on MLST scheme #3 . Leptospira-positive samples with incomplete Leptospira MLST cannot be assigned to a ST . In order to determine the DNA relatedness among Leptospira carried in the human or animal specimens , we drew a minimum-spanning tree ( MST ) based on a 501 bp secY gene fragment , using the goeBURST Full MST algorithm ( goeburst . phyloviz . net/ ) [49] .
For 44 of the 66 clinical samples , we successfully amplified all six MLST loci ( Table 1 ) . For twelve clinical samples , only one to five MLST loci could be successfully amplified , even when using degenerated primers [26] ( see S2 Table for details ) . For the last seven clinical samples , we failed to obtain successful PCR amplification at any of the six MLST loci . When these seven samples were further re-tested through an alternative real time PCR [37] , only two samples tested positive at Ct values exceeding 42 , whereas five sera provided negative amplification , suggesting that these were either false positive samples or positive samples that degraded during transport or conservation . Those seven human negative samples were discarded from the analysis and were no longer considered ( Fig 2 ) . The twelve partially sequenced samples allowed only identification of the infecting Leptospira at the species level; samples for which secY PCR product could be amplified were included in the MST analyses . In fine , of the 59 clinical samples successfully amplified at one or more locus , 57 were assigned to L . interrogans and two to L . borgpetersenii ( Table 1 ) . The secY-based MST ( Fig 3 ) , the neighbor-joining trees based on secY and rrs2 ( Fig 4 ) , and the MLST analysis all showed two L . interrogans clusters . As the two L . borgpetersenii-positive clinical samples allowed PCR amplification on two or three loci only ( i . e . lipL32 and rrs2 for both samples , lipL41 for one sample only ) but not on secY locus , they do not appear in the secY-based MST or the MLST-deduced phylogeny . Out of 995 wild terrestrial small mammals trapped on Reunion Island , 278 animals ( 27 . 9% ) tested positive for Leptospira . Out of 24 wild terrestrial small mammals for which culture was attempted , five were tested positive for Leptospira , among which two were successfully cultured ( both R . rattus ) . Positive samples amplified and sequenced on secY locus provided sequences for 191/278 ( 68 . 7% ) individuals , all of them being Rattus spp . and including the two successfully cultured samples . The remaining 87 positive samples were further tested for each of the other five MLST loci , as well as for small portions of the 16S region . This effort allowed identifying the Leptospira species infecting 61 additional samples from wild animals . However , because of the lack of either secY sequence or full MLST genotyping , these samples were excluded from the diversity analyses at the infra-specific level . Altogether , 249 out of the 252 Leptospira sequences that were obtained from wild mammal samples were identified as L . interrogans ( 98 . 8% ) . The three remaining sequences , all from mouse tissue samples , were identified either as L . borgpetersenii ( n = 1 ) or L . kirschneri ( n = 2 ) ; the L . borgpetersenii sample allowed lipL32 amplification while rrs2 amplification was obtained from one of the two L . kirschneri samples . The last L . kirschneri sample did not lead to any PCR amplification at any of the six MLST loci but was identified at the species level using LA/LB primers . Noteworthy , all T . ecaudatus tested PCR negative . These results are summarized in Table 1 . Interestingly , when aligning the 191 available secY sequences , all from rat tissue samples , it appeared that a single L . interrogans secY allele ( secY-1 ) was present in all rats but two . The two exceptions ( both R . norvegicus ) had different single substitutions at the secY sequence . The amino acid translation showed an internal stop codon in one of the two sequences , suggesting a sequencing error , and this sequence was later excluded . Thus , just one undescribed allele remained ( KU183598; see Fig 3 ) . The 101 domestic animals screened in our study provided the following positivity rates for Leptospira: 42 . 4% in cows , 15 . 6% in dogs and 0% in pigs ( Table 1 ) . The single sample from a cat was urine tested positive for Leptospira . Sequences on different loci were obtained for ten animals ( S1 Table ) . Leptospira infecting eight cows were identified as L . borgpetersenii , whereas Leptospira infecting two dogs were identified as L . interrogans , each dog showing a different secY allele , i . e . secY-6 ( also identified in six clinical samples ) and secY-1 ( highly dominant in both clinical and rat samples ) ( Fig 3 ) . Leptospira samples that were identified as L . borgpetersenii ( see Table 1 ) did not allow PCR amplification at most MLST loci , even when using recently described degenerated primers [26] . For cows ( n = 8 ) , we could amplify either adk ( n = 4 ) , lipL32 ( n = 1 ) , rrs2 ( n = 3 ) , or secY locus ( n = 6 ) , depending on samples ( see S1 Table ) , whereas only lipL32 locus was successfully amplified for the single L . borgpetersenii positive mouse . As for clinical samples ( n = 2 ) , we could amplify either lipL32 ( n = 2 ) , lipL41 ( n = 1 ) or rrs2 ( n = 2 ) . Thus , the only locus that allowed a comparison between human and animal samples was a 450 bp lipL32 PCR fragment amplified from two clinical samples , one mouse and one cow ( S1 Table ) . Alignments showed 100% nucleotide identity between L . borgpetersenii infecting the first clinical sample and the only mouse positive for L . borgpetersenii , and between L . borgpetersenii infecting the second clinical sample and one cow sample ( based on a partial 434 bp amplification for the cow ) . The secY locus amplification and sequencing provided sequences for six cow samples , revealing alleles secY-47 ( n = 4 ) , secY-49 ( n = 1 ) , and a third not yet described allele in one sample ( KU183602 , see Fig 3 ) . The allele secY-47 is closely related to secY-48 identified in clinical samples from Mayotte , whereas secY-49 has been identified in Leptospira-positive Tanzanian rodent samples and in a clinical sample from China ( comparison with sequences from [2] ) . The third undescribed secY allele was identified in the cow sample showing a lipL32 sequence common to those from one clinical sample . Noteworthy , twelve secY Leptospira sequences from the insectivorous bat Mormopterus francoismoutoui endemic to Reunion Island ( GenBank accession numbers KJ607946 to KJ607957 ) [50] were not related to any of these alleles ( Fig 3 ) . For 44 clinical samples and eleven rats , five to six loci of the MLST scheme [40] were successfully amplified , sequenced and concatenated for subsequent analyses ( Fig 2 , Table 1 and S1 Table ) . The 44 clinical samples which allowed successful amplification at all six MLST loci were identified as L . interrogans and fell into two clusters corresponding to previously described sequence types ( STs ) : one predominant , ST02 ( 38/44 = 86 . 4% ) and one minor , ST34 ( 6/44 = 13 . 6% ) . Including samples with successful amplification at 5 MLST loci only , we identified two clonal complexes , CC02 and CC34 , including human and rats for the former , and human and dogs for the later ( see S1 Fig ) . Among the 17 randomly selected Leptospira-positive wild animal samples , six rat samples were successfully amplified on five loci only ( excluding icdA , which PCR failed ) and five rat samples were successfully amplified on all six loci; they were all identified as L . interrogans CC02 or ST02 . Of note , we failed to amplify DNA at icdA , lipL32 and lipL41 loci for the rat sample that showed an alternative secY allele . Among the 22 Leptospira-positive domestic animals , none were successfully amplified on all six MLST loci . However , produced sequences revealed a perfect match between L . interrogans genotyped from two dogs and from clinical samples . As for the first dog , the six-loci concatenated sequence showed close identity with ST34 but a 14 nucleotides long sequence was missing on LipL41 locus , while for the second dog , the sequencing of four loci revealed relatedness to ST02 ( four alleles in common; adk and LipL41 non sequenced , see S2 Table ) . As highlighted above , the L . borgpetersenii-infected samples were hardly detectable by PCR and could not be fully genotyped , likely because of low bacterial loads; thus the phylogenetic tree based on the concatenated sequences of the MLST loci included L . interrogans samples only ( S1 Fig ) .
Although the mean prevalence of renal carriage among rats on Reunion Island ( 36 . 3% ) was close to that reported in R . rattus on Mayotte ( 29 . 8% ) [11] , the genetic data reported in the present study reveal a striking contrast between the two islands: the rich Leptospira genetic diversity ( at least 4 species with rather balanced representation ) reported in rats from Mayotte , also found in human cases , contrasts with the low Leptospira genetic diversity reported herein in humans , rats and shrews on Reunion Island: though we identified three Leptospira species , L . interrogans represents 96 . 6% of clinical samples , 95 . 8% of positive animal samples and 100% of rat samples . Other studies have reported a low genetic diversity among locally circulating Leptospira infecting local rat populations [25 , 51–53] . In contrast to earlier studies in New Caledonia [54] , New Zealand [55] or Argentina [56] , showing the carriage of L . interrogans or L . borgpetersenii in black rats ( R . rattus ) , and the absence of L . borgpetersenii in Norway rats ( R . norvegicus ) , L . borgpetersenii was absent in rats or shrews in our large sample of wild small mammals but was identified in a single mouse . The role of mice as maintenance hosts for L . borgpetersenii serogroup Ballum is largely recognized [1 , 3] . The perfect identity between lipL32 fragments from a single mouse and one of the two clinical samples positive for L . borgpetersenii highlights mice as a possible reservoir of pathogenic Leptospira at risk for humans on Reunion Island . However , this result should be interpreted cautiously as a single locus might not be resolutive enough as to infer Leptospira species; indeed , LipL32 sequences of Leptospira borgpetersenii and Leptospira weilii have been shown to be indistinguishable [57] . Further investigations targeting other loci with higher nucleotide polymorphism are needed to ascertain this point . A different L . borgpetersenii infecting the second clinical case suggested the existence of a second potential reservoir host . A bat species endemic to Reunion Island , Mormopterus francoismoutoui , has been demonstrated as a L . borgpetersenii carrier in urine [50] . However , the comparison of rrs2 sequences from our two clinical samples infected with L . borgpetersenii to GenBank sequences obtained from this endemic bat species ( n = 12 ) showed a low degree of genetic relatedness ( Fig 3 ) , thus excluding bats as a source of contamination leading to overt clinical leptospirosis in humans . A serological survey conducted in 2009 showed that up to 32% of cattle were seropositive for leptospirosis in Reunion Island , with Sejroe reported as the main circulating serogroup [30] . Our investigation finally identified cattle as a potential reservoir of L . borgpetersenii at risk for humans , as one L . borgpetersenii infecting a cow was found related to the L . borgpetersenii infecting a clinical case ( perfect identity between 434 bp lipL32 fragments ) . As for mice , further investigation is needed in order to verify a potential transmission between cattle and human , possibly through an environmental maintenance of pathogenic leptospires [58] . Apart from the two L . borgpetersenii alleles common to human cases of leptospirosis and one cow or one mouse , at least two additional L . borgpetersenii haplotypes were identified in cows only , indicating a higher Leptospira diversity in cattle than in any other animal investigated on the island , which was already evidenced in other settings [59] . Our experience and previous published work [60 , 61] have highlighted recurrent failure to detect and amplify specifically L . borgpetersenii strains . Consequently , acute clinical cases related to this species might be underestimated . The L . kirschneri found in two mice was not identified in any of our clinical samples on Reunion Island , whereas this species has been reported in clinical infection in Mayotte [11 , 12] as well as in other countries [2] . The reasons that may account for this difference are unknown . As mild infections not requiring hospitalisation or spontaneously healing cases are probably being underdiagnosed and underreported , it would be worth exploring whether these cases are associated with infection with other Leptospira species . At an infraspecific level , we identified two clusters of L . interrogans , referring to sequence types ST02 and ST34 ( http://www . pubmlst . org/leptospira/ ) . International isolates previously published on the pubmlst . org database showed ST02 to be isolated from Belgium and Brazil , while ST34 was isolated from Jamaica and India . The second ( minor ) cluster , ST34 , identified in six clinical samples , could not be detected in any of the five wild small mammal species targeted by our study . Additional sampling of domestic animals interestingly highlighted stray dogs as possible carriers of both L . interrogans STs found in human acute cases . Dogs have been repeatedly pinpointed as involved directly or indirectly in human epidemiology ( contamination from dogs to humans , or from a common environmental source ) [62–66] . For instance , a serosurvey conducted on Reunion Island in 2009 identified Canicola and Icterohaemorrhagiae as the major serogroups ( 43 . 5% and 21 . 7% respectively ) in seropositive stray dogs , while Sejroe , Panama , Tarassovi and Ballum serogroups were also identified [29] . On the other hand , up to 17% of clinical cases were found to be infected with serogroup Canicola [28] , suggesting that dogs , as the reservoir for this serogroup , are the source of contamination on Reunion Island . The importance of dogs in the epidemiology of the disease , if confirmed on Reunion Island , might be related to the lack of mammals diversity locally , as suggested on the island of Barbados [67] . An epidemiological survey conducted by CIRE-OI ( Regional Office of French Institute for Public Health Surveillance in Indian Ocean , see [13] ) indicates that , regarding the six patients infected with L . interrogans ST34 , ( i ) five were actually living in the eastern , windward humid coast of the island and ( ii ) for those five cases , the disease was associated to leisure activities in the eastern side rivers and waterfalls . Noteworthy , of 56 R . rattus positive for Leptospira trapped in sites near rivers in the eastern transect ( Fig 1 ) , none of the 53 available sequences corresponded to L . interrogans ST34 . Hence , it seems very unlikely that R . rattus acts as the reservoir of L . interrogans ST34 . In the end and as expected , our analyses stress black and Norway rats as reservoirs of Leptospira lineages of medical importance on the island . However , we were able to identify other animal species likely playing a role in the local epidemiology of human leptospirosis: mice , cattle and stray dogs . Noteworthy , the very low infra-specific diversity of L . interrogans infecting our infected samples did not enable us to assess the respective contribution of rats versus dogs to acute human infection with L . interrogans . Leptospira isolation from animals and humans would help to better assess the epidemiological links between humans and animals . Our conclusions are weakened by PCR and/or sequencing failures , likely resulting from different primer annealing efficiencies depending on the Leptospira species used as template ( supported by our incomplete results on cattle and mice samples infected with L . borgpetersenii ) , or from low renal carriage , as suggested by high Ct values ( e . g . Ct >30 in rats or >37 in dogs with undetermined infecting Leptospira , see S3 Table ) . Although time-consuming and fastidious , Leptospira culture needs to be carried out systematically , whenever possible , as previously suggested by other authors [68] . The high infection prevalence of cows with L . borgpetersenii raises another issue: the available human vaccine against leptospirosis currently proposed on the island target only L . interrogans species serogroup Icterohaemorrhagiae , while Sejroe was reported to be the main serogroup circulating in cattle in Reunion Island in 2009 [29] . This means that the vaccination of the ranchers’ population , at risk for leptospirosis for being in close contact with cows , might be inefficient against any other serogroup likely carried by livestock , in particular the one associated with L . borgpetersenii . Pigs are another potential reservoir that deserves attention . Leptospirosis is a common disease of swine throughout the world and can be a significant cause of reproductive loss [69] . None of the 22 pig samples included in the study tested positive to Leptospira , while a serological survey conducted in 2009 on Reunion Island showed that up to 47 . 2% of pigs were seropositive for leptospirosis [29] . These two observations are not contradictory . First , the acute infection of an animal with leptospires does not prejudge its ability to develop a bacterial biofilm in the renal tubules and to shed the bacteria into the environment . Hence , serological data and PCR data on kidneys may be discrepant . Second , the kidney samples used in our study were collected from growing-finishing pigs under one year old , a potential age bias that might have affected our findings . Indeed , a proper assessment of pigs as reservoirs of leptospires at risk for human health requires the screening of older animals sampled from different farms as well as from family breedings throughout the entire island . It is hypothesized that Leptospira were introduced onto islands with their animal hosts , and that a variable number of introduced strains have adapted to the new local environment and available hosts [14] . On Reunion Island , the fact that L . interrogans is almost clonal in rats is in favour of a recent leptospiral introduction on the island concomitant with the introduction of rats , back to the 16th century . Leptospires detected in dogs , mice and cows show higher genetic diversity , which could result from multiple introduction events of infected animals of different species ( for cows , there has been multiple importations from France of both Holstein Frisonne and Montbéliarde until 2008 ) or from distinct geographical origin , carrying different Leptospira lineages . We suggest that in contrast to Mayotte where the high diversity of Leptospira in humans and animal reservoirs is most likely related to ( older ) introduction events of infected hosts , potentially from countries with high host endemicity associated with Leptospira diversity [26] , on Reunion Island , the narrow diversity of Leptospira might reflect more recent and/or fewer host introduction events . This is partly supported by the fact that fifteen of the seventeen STs from Mayotte strains have not been previously identified [12] , contrary to STs from Reunion Island clinical cases that have already been identified worldwide . Previous studies have suggested that Leptospira from Mayotte and Madagascar are genetically closely related , and that Leptospira identified in Mayotte are probably mainly derived from Leptospira from Madagascar , possibly introduced in Mayotte via their hosts [14] . Although no genetic information is available so far on pathogenic Leptospira prevalent on Comoros , the serological profiles in humans from Comoros are comparable to those depicted in Mayotte , and noteworthy , antibodies from the serogroup Icterohaemorrhagiae are not detectable [20] . These findings contrast with those from human leptospirosis in Reunion Island and the Seychelles , where the Icterohaemorrhagiae serogroup is most common [70] . Hence there is compelling evidence that leptospirosis epidemiology might be similar along the Comoros , Mayotte and Madagascar axis and associated to indigenous or endemic Leptospira lineages , but different from that prevalent on Reunion Island and potentially Seychelles and Mauritius , where Icterohaemorrhagiae serogroup is dominant in clinical cases [15 , 16 , 21] . The near clonality reported herein in rat-borne Leptospira on Reunion Island is highly suggestive of recent pathogen introduction . This has to be addressed in the future by similar molecular analyses of Leptospira isolates from the other SWIO islands for which almost no molecular data are available so far , i . e . Mauritius and Seychelles .
|
Leptospirosis is a zoonosis caused by infection with pathogenic Leptospira species . A broad range of animals , including rodents , pets and livestock , are maintenance hosts for leptospires . However , assessing the relative importance of each host in the contamination of the environment and , in fine , in the infection of humans , has rarely been performed . In this study , we surveyed various wild and domestic animal species and their Leptospira carriage in Reunion Island , where human leptospirosis is endemic . We determined and compared the Leptospira genetic diversity at the species and infra-species levels in laboratory-confirmed human cases and in infected animals . The two Leptospira species infecting humans , Leptospira interrogans and Leptospira borgpetersenii , could be traced back to different animal species: rats and dogs for the former species , cows and mice for the latter . The Leptospira infecting the single bat species endemic to the island was not found to be involved in human leptospirosis . Aside from rats , which were expected to play a role in the local epidemiology of the disease , the putative role of dogs , cattle and mice in human epidemiology on Reunion Island , pinpointed by our data , deserves a specific investigation . These results have strong implications in terms of local control actions aimed at reducing the burden of human leptospirosis .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"tropical",
"diseases",
"microbiology",
"vertebrates",
"animals",
"mammals",
"dogs",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"zoonoses",
"artificial",
"gene",
"amplification",
"and",
"extension",
"medical",
"microbiology",
"microbial",
"pathogens",
"molecular",
"biology",
"genetic",
"loci",
"leptospirosis",
"polymerase",
"chain",
"reaction",
"genetics",
"biology",
"and",
"life",
"sciences",
"leptospira",
"interrogans",
"amniotes",
"organisms"
] |
2016
|
Human Leptospirosis on Reunion Island, Indian Ocean: Are Rodents the (Only) Ones to Blame?
|
It is currently believed that type I and III interferons ( IFNs ) have redundant functions . However , the preferential distribution of type III IFN receptor on epithelial cells suggests functional differences at epithelial surfaces . Here , using human intestinal epithelial cells we could show that although both type I and type III IFNs confer an antiviral state to the cells , they do so with distinct kinetics . Type I IFN signaling is characterized by an acute strong induction of interferon stimulated genes ( ISGs ) and confers fast antiviral protection . On the contrary , the slow acting type III IFN mediated antiviral protection is characterized by a weaker induction of ISGs in a delayed manner compared to type I IFN . Moreover , while transcript profiling revealed that both IFNs induced a similar set of ISGs , their temporal expression strictly depended on the IFNs , thereby leading to unique antiviral environments . Using a combination of data-driven mathematical modeling and experimental validation , we addressed the molecular reason for this differential kinetic of ISG expression . We could demonstrate that these kinetic differences are intrinsic to each signaling pathway and not due to different expression levels of the corresponding IFN receptors . We report that type III IFN is specifically tailored to act in specific cell types not only due to the restriction of its receptor but also by providing target cells with a distinct antiviral environment compared to type I IFN . We propose that this specific environment is key at surfaces that are often challenged with the extracellular environment .
During viral infection interferons ( IFNs ) are the predominant cytokines made to combat viral replication and spread . Following binding to specific receptors , IFNs induce a JAK/STAT signaling cascade which results in the production of interferon stimulated genes ( ISGs ) . These ISGs will then establish an antiviral state within the cell and will also alert surrounding cells and immune cells to assist in viral clearance [1] . There are three classes of IFNs . Type I IFNs are produced by all cell types and are recognized by the ubiquitously expressed heterodimeric receptor IFNAR1/IFNAR2 . Type II IFNs are only produced by immune cells [2 , 3] . Type III IFNs are made by all cell types but the IFNLR1 ( or IL28Ra ) subunit of the heterodimeric receptor IFNLR1/IL10Rβ is restricted to epithelial and barrier surfaces and to a subset of immune cells [4–9] . Despite the fact that type I and type III IFNs are structurally unrelated and engage different receptors , signaling downstream of both receptors exhibits a remarkable overlap and leads to the induction of a similar pool of ISGs . These observations originally led to the hypothesis that type I and III IFNs were functionally redundant . This model has been challenged more and more in recent studies which highlight that the cell type specific compartmentalization of IFNLR1 provides type III IFNs a unique potential for targeting local infections at mucosal surfaces . For example , in vivo data on enteric virus infection of the murine gastrointestinal tract showed that responsiveness to type III IFN is necessary and sufficient to protect murine intestinal epithelial cells ( IECs ) against rotavirus and reovirus infection [10–12] . On the contrary , type I IFN was necessary to protect against viral infection of cells in the lamina propria and against systemic spread [10 , 11] . Likewise , it was demonstrated that fecal shedding of norovirus was increased in IFNLR1-deficient , but not IFNAR1-deficient , mice , showing that type III IFN uniquely controls local norovirus infection in the gut [13 , 14] . Similarly , in the respiratory tract , type III IFNs are predominately produced upon infection with influenza A virus [15–19] . However , as infection progresses type I IFN comes into play to reinforce viral inhibition by inducing a pro-inflammatory response [20] . Differences in the antiviral activity conferred by both cytokines appear to be not only driven by the spatial restriction of their receptors but also by intrinsic subtle differences in signal transduction . It was demonstrated , in human hepatocytes and lung epithelial cells , that type I IFN confers a more potent antiviral protection compared to type III IFNs [5 , 21–23] . Additionally , it was shown in human IECs that type III IFN partially requires MAP kinase activation to promote an antiviral state while type I IFN was independent of it [24] . Although it has been reported in many studies that very similar ISGs are induced upon type I or type III IFN stimulation of cells , work mostly performed in hepatocytes revealed that both cytokines induce these ISGs with different kinetics [21 , 25–27] . Type III IFN mediated signaling was found to be associated with a delayed and reduced induction of ISGs compared to type I IFNs [25 , 26] . Similar differences in the magnitude and/or kinetics of ISGs induction upon type I versus type III IFN treatment were observed in human primary keratinocytes , airway epithelial cells and in Burkitt's lymphoma derived B ( Raji ) cells , as well as in murine intestinal and lung epithelial cells and immune cells [20 , 28–31] . The molecular mechanisms leading to this delayed and reduced induction of ISGs upon type III IFN treatment remains unclear . As these differences in kinetics of ISG expression between both IFNs could not be directly explained by their signaling cascades an alternative explanation was proposed where type III IFN receptor is expressed at lower levels at the cell surface . This lower receptor expression level could provide a biochemical explanation for the observed differences in delayed kinetics and weaker amplitude of ISG expression compared to type I IFN . However , to date , there is no direct experimental evidence for this model . Similarly , whether the observed differences between both IFNs is intrinsic to both specific signal transduction pathways and whether it is restricted to some cell types ( e . g . hepatocytes ) or represents a global signaling signature in all cells expressing both IFN receptors has not been fully addressed . In this study , we have investigated how type I and III IFNs establish their antiviral program in human mini-gut organoids and human IEC lines . We found that type I IFN can protect human IECs against viral infection faster than its type III IFN counterpart . Correspondingly , we determined that type I IFN displays both a greater magnitude and faster kinetics of ISG induction compared to the milder , slower type III IFN . By developing mathematical models describing both type I and type III IFN mediated production of ISGs and by using functional receptor overexpression approaches , we demonstrated that the observed lower magnitude of ISG expression for type III IFNs was partially the result of its lower receptor expression level compared to the type I IFN receptor . Inversely , the observed delayed kinetics of type III IFN cannot be explained by receptor expression level indicating that this property is specific to type III IFN and inherent to its signaling pathway . Our results highlight important differences existing between both type I and type III IFN-mediated antiviral activity and ISG expression which are not only the result of receptor compartmentalization but also through intrinsic fundamental differences in each IFN-mediated signaling pathway .
We have previously reported that both type I and III IFNs mediate antiviral protection in human IECs [24] . To address whether type I and type III IFN have a different profile of antiviral activity in primary non-transformed human IECs , as reported in human lung cells [22] , we compared the antiviral potency of both IFNs in human mini-gut organoids . Colon organoids were pre-treated with increasing concentrations of either type I or III IFNs for 2 . 5 hours and subsequently infected with vesicular stomatitis virus expressing luciferase ( VSV-Luc ) . Viral infection was assayed by bioluminescence and results showed that both IFNs induced an antiviral state in a dose-dependent manner . We observed that type I IFN was slightly more potent in protecting against viral infection at higher concentration compare to type III IFNs . Type I IFN could almost fully inhibit viral infection while type III IFN was only able to reduce infection to around 80% ( Fig 1A ) . Interestingly , the concentration of type I IFN necessary to provide 90% of relative antiviral protection ( EC90 ) was significantly lower than the one for type III IFN ( Fig 1B ) . To determine whether type III IFN requires a prolonged treatment to achieve similar antiviral protection as observed with type I IFN , we performed a time course experiment in which human colon organoids were pre-treated for different times with either IFN prior infection with VSV-Luc ( Fig 1C ) . We found that approximately 2 hours pre-treatment with type I IFN was sufficient to reduce VSV infection by 90% ( 10% remaining infection ) , while type III IFN required around 5 hours to achieve a 90% reduction of infectivity ( Fig 1C and 1D ) . Interestingly , 24 hours of pretreatment was necessary for type III IFN to almost completely prevent VSV infection ( Fig 1C ) . These results strongly suggest that both type I and type III IFN could have similar potency but that type III IFN requires more time to establish an antiviral state . We next addressed how long after infection IFN treatment is still able to promote antiviral protection . Colon organoids were infected with VSV-Luc and treated at different times post-infection with either type I or III IFNs . Interestingly , type I IFN could inhibit viral replication even when added several hours post-infection . In contrast , type III IFN appeared to require a much longer time to establish its antiviral activity and was unable to efficiently protect the organoids after VSV infection has initiated ( Fig 1E and 1F ) . Importantly , these differences in the kinetics of antiviral activity of type I versus type III IFNs were neither donor nor colon specific as similar results were observed in intestinal ileum-derived organoids derived from different donors ( S1 Fig ) . In addition , the human colon carcinoma-derived cell line T84 ( S2 Fig ) fully phenocopy the difference in type I versus type III IFN antiviral activity generated by primary mini-gut organoids . Taken together these results demonstrate that while both type I and III IFNs can promote similar antiviral states into target cells , they do so with distinct kinetics . The cytokine-induced antiviral state is promoted faster by type I IFN compared to type III IFN . To understand how type I and type III IFNs promote an antiviral state in primary IECs but with different kinetics , we analyzed the magnitude of ISG expression over time upon IFN treatment . Colon organoids were treated with increasing concentrations of either type I or type III IFN and the expression levels of two representative ISGs ( IFIT1 and Viperin ) were assayed at different times post-IFN treatment . Results revealed that type I IFN ultimately leads to a significantly higher induction of both IFIT1 and Viperin compared to type III IFN ( Fig 2A and 2B ) . This difference in the magnitude of ISG stimulation was independent of the duration of IFN treatment ( Fig 2A and 2B ) . To determine if this pattern of expression applies to other ISGs , we treated colon organoids with either type I or type III IFN over a 24-hour time course , and analyzed the mRNA levels of 132 different ISGs and transcription factors involved in IFN signaling ( see complete list of genes and corresponding primers in S1 and S2 Tables ) ( Fig 2C and 2D ) . Differential expression analysis revealed that both type I and type III IFNs induce almost the same set of ISGs and that most of the genes significantly induced by type III IFN were also induced by type I IFN ( Fig 2C ) . However , similar to IFIT1 and Viperin ( Fig 2A and 2B ) , we found that the magnitude of ISG expression was greater for type I IFN compared to type III IFN ( Fig 2D ) . Similar results were found in the immortalized colon carcinoma-derived T84 cells ( S3A–S3C Fig ) . To address whether there is any correlation between the different antiviral protection kinetics conferred by type I and III IFNs ( Fig 1 ) and the kinetics of ISG expression , we analyzed the temporal expression of ISGs upon IFN treatment of human colon organoids . Hierarchical clustering analysis of all ISGs up-regulated upon type I or type III IFN treatment defined four distinct expression profiles based on the time of their maximum induction ( Fig 3A–3C ) . Group 1 are ISGs whose expression peaks 3 hours post-IFN treatment . The expression of ISGs in group 2 and 3 peaks at 6 and 12 hours post-treatment , respectively . Group 4 corresponds to ISGs with a continuous increase in expression over time ( Fig 3A and 3B ) . Under type I IFN treatment , ISGs were nearly equally distributed in all four expression groups ( Fig 3A , 3C and 3D ) . By contrast , although the same ISGs were induced by type III IFN , they almost all belong to the expression group 4 , being expressed later after IFN treatment ( Fig 3B–3D ) . In line with the primary mini-gut organoids , T84 cells presented similar differences in the kinetics of ISGs expression ( S3D Fig ) . We next wanted to control that our observed differences in the kinetics of ISGs expression induced by both cytokines were independent of IFN concentration . Colon organoids were treated with increasing amounts of type I or type III IFNs and the transcriptional up-regulation of representative ISGs belonging to each of the expression profile groups ( group 1–4 ) was measured over time ( Fig 4 ) . Consistent with our previous results , the temporal expression patterns of each representative ISGs were independent of the IFN concentration and the ISG expression kinetic signature was specific to each IFN ( Fig 4 ) . Complementarily , to address whether the observed differences between type I and type III IFNs were not due to the lower affinity of type III IFN for its receptor compared to type I IFN , we employed the high affinity variant of type III IFN ( H11-IFNλ3 ) [32] to monitor the kinetics of ISG expression . Results show that cells treated with the higher affinity H11-IFNλ3 display a higher magnitude of ISG expression but their kinetics of expression were unchanged ( S4 Fig ) . Altogether , our results strongly suggest that although both type I and type III IFNs induce a similar set of ISGs in hIECs , type III IFN induces globally a lower amplitude and a delayed ISG expression compared to type I IFN . Our data show remarkable differences in the magnitude and kinetics of ISGs induced by type I versus type III IFN ( Figs 2 and 3 and S3 Fig ) , and in the subsequent induction of a differential antiviral state ( Fig 1 and S1 and S2 Figs ) . To investigate the mechanisms underlying these differences , we used data-driven mathematical modeling and model selection . We considered three mechanistic causes for the observed differential signaling: ( 1 ) Receptor abundance ( different number of IFNLR compared to IFNAR complexes ) ; ( 2 ) Receptor regulation ( different rates of activation and/or inactivation of IFNLR compared to IFNAR complexes ) ; ( 3 ) STAT activation ( different rates of STAT activation by type I and type III IFNs ) . We devised corresponding mathematical models describing the dynamics of receptor activation and inactivation , STAT1/2 phosphorylation and STAT-dependent activation of ISG expression ( Fig 5A ) . The models were implemented as systems of ordinary differential equations ( S3 Table ) and fitted to the time-resolved data for the prototypical ISG , Viperin , measured with different doses of type I or type III IFNs and with the high affinity H11-IFNλ3 . We ranked the models according to the Akaike information criterion corrected for small sample size ( AICc ) , which evaluates the goodness of fit and , at the same time , penalizes the number of fit parameters ( for more details see Materials and Methods ) . Throughout , we allowed different receptor abundance , but this difference alone could not account for the different signaling dynamics ( Fig 5B; model M1 has negligible support by the data , as quantified by the small AICc weight , which is a weight of evidence for the respective model ) . Interestingly , in addition to receptor abundance , the best-fitting model ( M3 ) has also different rates of activation and inactivation of IFNLR and IFNAR complexes . However , alternative models with different rates of STAT activation and/or ISG expression have good performance ( M2 and M4 , respectively ) . Therefore , the modeling indicates that differential ISG activation by type I and type III IFNs is likely due to different abundance of the respective receptors and cell-intrinsic differences in how the signals from bound receptors are processed . The best-fitting model ( M3 ) accounted for the dose-response and the different Viperin expression kinetics triggered by type I , type III and the high affinity H11-IFNλ3 in T84 cells , group 3 and group 4 expression kinetics , respectively ( Fig 5C and 5D ) . The different kinetics of the IFN responses–fast and transient for type I IFN vs slower and sustained for type III IFN–are predicted to be largely due to receptor inactivation , which is faster for IFNAR than for IFNLR complex ( S5A–S5C Fig ) . Interestingly , the model shows that at low IFN concentrations , Viperin is induced almost equally by both IFNs whereas at higher concentrations , type I IFN induces Viperin more strongly ( Fig 5E ) . These dose-dependent features agree with our experimental data ( S3B Fig , right panel ) . Next , we tested the pivotal impact of receptor expression on ISG induction that was indicated by our model . Specifically , the model predicts that an increase in IFNAR1 or IFNLR1 level will increase the amplitude of ISG induction while preserving the specific kinetic profiles elicited by the two types of IFNs ( S5D and S5E Fig ) . To experimentally validate the model predictions , IFNAR1 and IFNLR1 were overexpressed in T84 cells . Overexpression of the respective IFN receptor chain was confirmed by reverse quantitative PCR ( S6 Fig ) . To ensure the functionality of both IFN receptors , IFNAR1 or IFNLR1 were expressed in our previously characterized knockout T84 cell lines deficient for either the IFN alpha receptor 1 ( IFNAR1-/- ) or the IFN lambda receptor 1 ( IFNLR1-/- ) ( S7A and S7E Fig ) [24] . Our results show that overexpression of IFNAR1 in our IFNAR1-/- T84 cells ( IFNAR1-/-+rIFNAR1 ) restores their antiviral activity , their ability to phosphorylate STAT1 and induce the production of the ISGs IFIT1 and Viperin in the presence of type I IFN ( S7B–S7D Fig ) . Similarly , although IFNLR1-/- cells were insensitive to type III IFN treatment , overexpression of IFNLR1 ( IFNLR1-/-+rIFNLR1 ) restored their antiviral activity , pSTAT1 and ISG induction after addition of type III IFN ( S7F–S7H Fig ) . These results demonstrate the functionality of both IFN receptors and validate our overexpression approach as a means to increase IFNAR1 and IFNLR1 levels at the cell surface . Wild-type T84 cells overexpressing type I IFN receptor ( WT+rIFNAR1 ) were treated with increasing concentrations of type I IFN . Our results showed elevated levels of STAT1 phosphorylation and ISG induction in response to stimulation with type I IFN compared to wild-type cells ( Fig 6A–6D ) . Importantly , the response of T84 cells overexpressing type I IFN receptor to type III IFN remained unchanged , indicating a selective enhancement of the type I IFN signaling pathway . Similarly , overexpression of type III IFN receptor ( WT+rIFNLR1 ) shows a significant increase in phosphorylated STAT1 and ISG expression compared to wild-type cells upon type III IFN stimulation , while no difference was observed upon type I IFN treatment ( Fig 6E–6H ) . Altogether , our experimental data are consistent with the modeling predictions and confirm the crucial impact of surface receptor levels for regulating the magnitude of type I and III IFN response . We next addressed whether this increase of ISG expression in cells overexpressing either the type I or type III IFN receptor was associated with an improved antiviral activity . Wild-type T84 cells overexpressing type I IFN receptor ( WT+rIFNAR1 ) were treated with type I IFN at different time points prior to infection with VSV-Luc virus and their antiviral activity was compared to wild-type T84 cells . Our results showed that the potency and the kinetics of the antiviral activity of cells overexpressing type I IFN receptor does not present any significant change upon type I IFN treatment ( S8A Fig ) . Similarly , there is no difference in the antiviral activity when cells overexpressing type I IFN receptor were treated with type I IFN at different time points post-infection ( S8B Fig ) . However , overexpression of type III IFN receptor ( WT+rIFNLR1 ) shows a modest but significant enhancement in type III IFN antiviral potency in the earlier time points of pre-treatment ( between 30 minutes and 2 hours ) compared to wild-type cells upon type III IFN stimulation ( S8G Fig ) , while they responded similarly to wild-type cells upon type I IFN treatment ( S8E Fig ) . Consistent with this , cells overexpressing type III IFN receptor are more protected than wild-type cells when type III IFN was added post-infection for the early time points ( S8H Fig ) . Finally , to experimentally validate the limited impact of the IFN receptors abundance on the kinetic profile of ISG expression , as predicted by the model ( S5D and S5E Fig ) , wild-type cells overexpressing either of the IFN receptors were treated with increasing doses of type I or type III IFNs and the expression of a representative ISG belonging to each of the expression profile groups ( group 1–4 , Fig 3 ) was analyzed over time ( Fig 7A–7D ) . The experimental data show that the amplitude of ISG expression was dependent on both the dose of IFNs used to stimulate the cells and on the expression levels of the IFN receptors ( Fig 7A–7D ) . Importantly , the kinetic profile of ISG expression was similar between WT cells and cells overexpressing the IFNAR1 ( WT+rIFNAR1 ) , independent of the applied IFN type I dose ( Fig 7A–7D left panel ) . Similarly , wild-type cells overexpressing the IFNLR1 ( WT+rIFNLR1 ) showed no change in the kinetic profile of ISG induction upon type III IFN stimulation ( Fig 7A–7D right panel ) . Moreover , we found that the model reproduced the kinetic dose-response data when the IFNAR1 and IFNLR1 expression levels were increased ~2 . 6 and ~1 . 5 times , respectively , while all other parameters were held constant ( S9 Fig ) . Indeed , we found that IFNAR1 overexpression was stronger than IFNLR1 overexpression , as judged by the transcript levels ( S6B and S6C Fig ) , with the ratio being consistent with the model prediction ( S9D Fig and S6B and S6C Fig ) . To directly correlate ISG expression kinetics and amplitude with the expression level of the type III IFN receptor , we thought of overexpressing an IFNLR1 tagged with the GFP fluorescent protein ( IFNLR-GFP ) in human IECs . To control the functionality of the GFP tagged receptor , the IFNLR1-GFP construct was overexpressed in the human embryonic kidney cell line 293 HEK , which normally elicit a very limited response upon type III IFN treatment . Quantitative RT-PCR revealed that 293 HEK cells overexpressing IFNLR1-GFP produced significantly more ISGs upon type III IFN treatment compared to WT 293 HEK cells or 293 HEK cells expression GFP alone ( S10 Fig ) . Wild-type T84 cells overexpressing the IFNLR1-GFP ( WT+rIFNLR1-GFP ) were treated with type III IFN over time and cells were sorted by flow cytometry based on their level of IFNLR1-GFP expression ( no GFP expressing ( neg ) , or low and high GFP expressing cells ) ( Fig 8A ) . The induction of a representative ISG belonging to each of the expression profile groups ( group 1–4 , Fig 3 ) was measured over time in each sorted population ( negative , low and high , Fig 8B ) . As anticipated , WT cells overexpressing the IFNLR1-GFP chain show stronger ISG expression compared to WT cells and the magnitude of the ISG induction correlates with the relative levels of IFNLR1 expression ( Fig 8B ) . However , the kinetic profiles of the ISGs upon type III IFN stimulation were not affected by the differential expression levels of the IFNLR1 chain ( Fig 8B ) . Altogether , our results demonstrate that type I and type III IFNs both induce an antiviral state in hIECs but with different kinetics . We could show that although both cytokines induce similar ISGs , type III IFN does it with slower kinetics and lower amplitude of individual ISG expression compared to type I IFN . Importantly , coupling mathematical modeling of both type I and type III IFN-mediated signaling and overexpression of functional IFN receptors approaches allowed us to demonstrate that these kinetic differences in type I and type III IFN ISG expression are not due to different expression level of the respective IFN receptors but are intrinsic to type I and type III IFN signaling pathways .
In this work , we have for the first time , performed a parallel study of the role of type I and III IFN in human mini-gut organoids and IEC lines . Our results demonstrate that type I and III IFNs are unique in their magnitude and kinetics of ISG induction . Type I IFN signaling is characterized by relatively strong expression of ISGs and confers to cells a fast-antiviral protection . On the contrary , the slow acting type III IFN mediated antiviral protection is characterized by a weak induction of ISGs in a delayed manner compared to type I IFN . Our results are in line with previous studies which also demonstrated that type III IFN is less potent than its type I IFN counterpart [5 , 21 , 23 , 33 , 34] . Additionally , we have confirmed that the delayed ISG induction seen upon type III IFN treatment of hepatocytes [21 , 23 , 25 , 26] is not tissue specific but likely represents a global pattern of action of this cytokine in cells expressing the type III IFN receptor ( i . e . human epithelial cells ) . In other words , the different kinetics of ISG expression induced by type I and type III IFNs are specific to each IFN signaling pathways . In the current work , we have employed , a data-driven mathematical modeling approach to explain signal transduction kinetic differences downstream type I and type III IFN receptors . While type I IFN-mediated signaling has been previously modeled [35 , 36] , type III IFN has not . Our model predicted that the receptor levels directly influence the magnitude of ISG expression however , the kinetics of ISG expression appear to be intrinsic to each IFN-signaling pathway and is largely preserved under receptor overexpression . This prediction was experimentally validated by studying the response of wild-type and IFN receptor overexpressing cells to different doses of IFN ( Fig 7A–7C and Fig 8 ) . This suggests that the kinetic differences in the ISG induction are intrinsic to each IFN signaling pathway . We propose that these phenotypic differences reflect functional differences , which are important for mounting a well-tailored antiviral innate immune response at mucosal surfaces where type III IFN receptors are expressed . Both type I and III IFNs have unique and independent receptors which are structurally unrelated . These receptors are likely expressed at different levels on individual cells and their relative expression to each other might also be cell type specific . To address whether the unique ISG and antiviral expression kinetics shown by each IFN were not due to differences in their expression levels , we overexpressed into cells functional type I ( rIFNAR1 ) and type III IFN ( rIFNLR1 ) receptors . Our results from IFNAR1 overexpressing cells ( Figs 6 and 7 ) are in line with previous studies showing a direct relationship between the surface levels of type I IFN receptors and the magnitude of ISG induction [37 , 38] . Interestingly , we could demonstrate a similar relationship when overexpressing IFNLR1 ( Figs 6 and 7 ) which was also associated with an increase of type III IFN antiviral potency ( S8 Fig ) . These findings are in agreement with previous experiments which show that overexpression of IFNLR1 in cells which normally do not express this IFN receptor rescues both type III IFN-mediated signaling and IFN-mediated antiviral protection [5 , 28] . Our IFN receptor overexpression approach demonstrates that the observed differences in ISG expression kinetics are not the results of different levels of receptors at the cell surface but is likely specific to each signal transduction pathway . Apart from the expression levels of IFN receptors , lower binding affinity towards their respective receptors could be an alternative explanation for the differential potencies of both type I and type III IFNs . Multiple studies have tried to affect the binding affinity of type I IFNs with their receptors however , results suggest that wild-type IFNs exert their antiviral activities already at maximum potency . Modifications leading to an increased affinity for their receptors do not lead to improvement of antiviral potency [32 , 38–41] . To address whether the weaker activity of type III IFN could be the result of its weaker affinity for its receptor , Mendoza et al , engineered a variant of type III IFN with higher-affinity for its receptor ( H11-IFNλ3 ) . They showed increased IFN signaling and antiviral activity in comparison with wild-type IFNλ3 . However , the engineered variant of IFNλ3 was still acting with weaker efficacy compared to type I IFNs [32] . By exploiting the high affinity variant H11-IFNλ3 , we could also show a significant increase of the amplitude of ISG expression but importantly , the kinetics of ISG expressions were not altered ( S4 Fig ) . Our results indicate a model were inherent temporal differences exist between type I and type III IFNs signaling . These differences are not the result of differential surface expression of the receptors but is the result of distinct signaling cascades from the receptors to the nucleus or within regulatory mechanisms of gene expressions . While few studies have focused on the endocytosis and inactivation of IFNAR1 , there is no information about how these processes occur for IFNLR1 . It has been shown that the ternary IFNAR complex is internalized by clathrin mediated endocytosis [42] and that upon type I IFN stimulation , IFNAR1 is rapidly endocytosed and routed for lysosomal degradation [43 , 44] , whereas IFNAR2 can be recycled back to the cell surface or degraded [45] . Our data-driven mathematical modeling approach suggests a different kinetics of receptor activation/inactivation between both IFNs ( Fig 5B and S5A Fig ) . Therefore , further studies investigating trafficking of IFNLR1 will be important and may show that subtle changes in the time course of receptors internalization , recycling or degradation can have profound effect on kinetics of IFN activity . Apart from receptor internalization and degradation , several molecular mechanisms leading to IFN receptor inactivation have been described , such as de-phosphorylation [46 , 47] , or by negatively targeting the interaction of IFNAR1 with downstream signaling elements of the JAK/STAT signaling , for instance ubiquitin-specific protease USP18 , and members of the suppressor of cytokine signaling protein ( SOCS ) family . In particular , the inhibitory role of SOCS1 in type I IFN signaling has been demonstrated in a number of previous studies , where they have shown that SOCS1 associates with TyK2 and blocks its interaction with IFNAR1 [48] . USP18 has also been shown as an important negative regulator of type I IFN signaling with a dual role acting as isopeptidase which removes the ubiquitin like-ISG15 from target proteins [49] and as a competitor of JAK1 for binding to IFNAR2 [50] . Although , limited information is available for negative regulators of the IFNLR receptor complex , the specific contribution of USP18 or SOCS in inhibition of type I versus type III IFN mediated signaling has been addressed in recent studies . In particular , it has been showed that both type I and III IFNs ( IFNα , IFNβ and IFNλ1 , λ2 , λ3 and λ4 ) induced the expression of USP18 , SOCS1 and SOCS3 [51–57] and overexpression of all these negative regulators inhibited both IFNα and IFNλ1 mediated JAK-STAT signaling [54 , 56] suggesting that at ‘‘supraphysiological” expression levels all the inhibitors can block type I and type III mediated JAK-STAT signaling [56] . Additionally , it has been shown that USP18 is induced later and that its level increased over time , correlating with the long lasting refractories of IFNα signaling [51 , 52 , 56] . In our study we observed also a later peak of induction of USP18 at 12h or 24h upon type I or type III IFNs , respectively . In line with the above-mentioned studies we also observed rapid and transient induction of SOCS1 upon type I IFN treatment and sustained induction upon type III IFN stimulation . However , further investigation is required to determine the correlation of the kinetics of induction of these negative regulators with the ISGs induction in type I versus type III IFN treatment in human IECs . In the canonical type I and III IFN signaling pathway the next downstream players from the IFN receptors are the JAKs , STAT1 , STAT2 and IRF9 , which are all regulated on the level of expression and activation . Our own observations and previous studies could not explain the major differences in the kinetics of type I versus type III IFNs activity by focusing on the time course of phosphorylation of STATs [21 , 25] . However , given that alternative modifications of STATs ( e . g . phosphorylation on alternative residues , acetylation , methylation and sumoylation patterns ) have been proposed to contribute to the activity of type I IFNs [26 , 58–60] it might be possible that new modifiers of STAT activity may determine the kinetic pattern of action of type I versus type III IFNs . In addition , apart from the JAK/STAT axis , there is accumulating evidence which correlates ISG transcription upon IFN treatment with a plethora of JAK-STAT independent pathways , such as members of the CRK [61–63] and MAPKinase family [24 , 28 , 64–66] , which might also temporally coordinate IFNs kinetic profile of action . Apart from the differences in the signaling cascade of type I versus type III IFNs , an explanation for their differential kinetics of action might stem from the physiology of the different cell types . For example , in a recent study Bhushal et al . reported that polarization of mouse intestinal epithelial cells eliminates the kinetic differences between type I and type III IFNs , by accelerating type III IFN responses [33 , 67] . Several studies describing the transcriptional activities of both type I and type III IFNs have reported that very similar sets of ISGs are produced upon both type I and III IFN stimulation [12 , 17 , 21 , 22 , 25 , 28] while only few ISGs appear to be predominantly expressed upon type III IFN treatment in murine IECs [67] . We believe that there are several functional advantages for adopting a lower and slower activity , like the profile of action of type III IFN , in the antiviral protection of epithelial tissues . The differences in the temporal expression of ISGs could create unique antiviral environments for each IFN . Many ISGs function as pro-inflammatory factors [30 , 68] . By stimulating ISGs production in high magnitude , an excessive amount of antiviral and pro-inflammatory signals could be produced which on the one hand will eliminate efficiently viral spreading but on the other hand may cause local exacerbated inflammation and irreversible tissue damage , leading to chronic inflammation in mucosal surfaces . In addition , the expression of different functional groups of ISGs at early and at late time points ( Fig 3 ) might allow cells to create two distinct phases within the antiviral response . At early time points , minimum levels of ISGs may act to protect the host against viral infection . Antiviral ISGs will be responsible for fighting the pathogens and pro-inflammatory ISGs will stimulate members of the adaptive immune system to assist the antiviral protection . At later time points the produced ISGs , may be involved in anti-inflammatory processes , such as resolving of inflammation and tissue healing and repair [66 , 69] . To exert this anti-inflammatory role , ISGs may need to be produced in higher levels , as they might act more paracrine and spread through the tissue to balance again the tissue homeostasis after the viral attack . In conclusion , we propose that type III IFN-mediated signaling is not only set to act predominantly at epithelium surfaces due to the restriction of its receptor but also is specifically tailored to mount a distinct immune state compared to other IFNs which is critical for mucosal surfaces that face the challenge .
Commercially available primary antibodies were mouse monoclonal antibodies recognizing beta-Actin ( Sigma #A5441 ) , phospho STAT1 and STAT1 ( BD Transductions #612233 and #610115 , respectively ) . Anti-mouse ( GE Healthcare #NA934V ) , coupled with horseradish peroxidase was used as secondary antibody for Western blot at a 1:5000 dilution . Human recombinant IFN-beta1a ( IFNβ ) was obtained from Biomol ( #86421 ) . Recombinant human IFNλ1 ( IL-29 ) ( #300-02L ) and IFNλ2 ( IL28A ) ( #300-2K ) were purchased from Peprotech and IFNλ3 ( IL-28B ) from Cell signaling ( #8796 ) . High affinity engineered IFNλ3 variant ( H11 ) and wild type IFNλ3 were produced as described in [32] . The IFN concentrations used to treat the cells are stated in the main text and in the figure legends . T84 human colon carcinoma cells ( ATCC CCL-248 ) were maintained in a 50:50 mixture of Dulbecco’s modified Eagle’s medium ( DMEM ) and F12 ( GibCo ) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin ( GibCo ) . SKCO15 cells were maintained in DMEM with 10% fetal bovine serum , 1% penicillin/streptomycin , 15mM HEPES and 1% NEAA ( Non-Essential Amino Acids ) . Mini-gut organoids were harvested and maintained as described earlier [24] . VSV-Luc was used as previously described [24] . Human colon tissue was received from colon and small intestine resection from the University Hospital Heidelberg . This study was carried out in accordance with the recommendations of “University Hospital Heidelberg” with written informed consent from all subjects . All subjects gave written informed consent in accordance with the Declaration of Helsinki . All samples were received and maintained in an anonymized manner . The protocol was approved by the “Ethic commission of University Hospital Heidelberg” under the approved study protocol S-443/2017 . RNA was harvested from cells using NuceloSpin RNA extraction kit ( Macherey-Nagel ) as per manufacturer’s instructions . cDNA was made using iSCRIPT reverse transcriptase ( BioRad ) from 200ng of total RNA as per manufacturer’s instructions . qRT-PCR was performed using SsoAdvanced SYBR green ( BioRad ) as per manufacturer’s instructions , TBP and HPRT1 were used as normalizing genes . Colon organoids and T84 cells were treated with 2000 RU/ml of type I IFN ( β ) or 100 ng/ml of each type III IFN ( λ1 , 2 and 3 ) . Total RNA was isolated at 3 , 6 , 12 and 24h post-treatment as described above . For the gene expression analysis of interferon stimulated genes ( ISGs ) , qRT-PCR was performed using the predesigned 384-well assay of type I IFN response for use with SYBR Green assaying the expression of 87 ISGs ( Biorad # 10034592 ) . The expression of 45 additional ISGs and transcriptional factors was analyzed by qRT-PCR with primer sets obtained as previously described [27] . The complete gene list monitored in this study and the primers used to amplify each gene is available in S1 and S2 Tables . Differential expression analysis of each treatment was performed by comparing the baseline expression of genes in an untreated control at each time point . Only genes which were either induced or reduced more than 2-fold in any of the samples were considered to be significantly regulated . These genes were either analyzed using scatterplots or visualized by a heatmap after sorting the fold change of expression in response to type I IFN ( β ) in decreasing order . For the T84 cells all fold change values above 20 and below 0 . 05 were replaced with 20 and 0 . 05 respectively . For the organoids , the fold change values above 800 and below 1/800 were replaced with 800 and 1/800 . This data adaptation was done to center the heatmap around 0 ( white ) and to avoid errors in logarithmic calculations . When visualizing the expression peaks , only the highest value is shown per time point for each gene . All analyses were performed using R version 3 . 3 . 0 and 3 . 3 . 3 including the packages gplots and ggplot2 . At time of harvest , media was removed , cells were rinsed one time with 1X PBS and lysed with 1X RIPA buffer ( 150 mM sodium chloride , 1 . 0% Triton X-100 , 0 . 5% sodium deoxycholate , 0 . 1% sodium dodecyl sulphate ( SDS ) , 50 mM Tris , pH 8 . 0 with phosphatase and protease inhibitors ( Sigma-Aldrich ) ) for 20mins at 4°C . Lysates were collected and equal protein amounts were separated by SDS-PAGE and blotted onto a PVDF membrane by wet-blotting . Membranes were blocked with 5% milk or 5% BSA , when the phospho STAT1 antibody is used , in TBS containing 0 . 1% Tween 20 ( TBS-T ) for one hour at room temperature . Primary antibodies were diluted in blocking buffer and incubated overnight at 4°C . Membranes were washed 4X in TBS-T for 15mins at RT . Secondary antibodies were diluted in blocking buffer and incubated at RT for 1h with rocking . Membranes were washed 4X in TBS-T for 15mins at RT . HRP detection reagent ( GE Healthcare ) was mixed 1:1 and incubated at RT for 5mins . Membranes were exposed to film and developed . Colon organoids and T84 cells were seeded in a white F-bottom 96-well plate . Samples were pre-treated prior to infection or treated post-infection as indicated with increasing concentrations of type I or type III IFNs . VSV-Luc was added to the wells and the infection was allowed to proceed for 8hrs . At the end of the infection , media was removed , samples were washed 1X with PBS and lysed with Cell Lysis Buffer ( Promega ) at RT for 20 mins . A 1:1 dilution of Steady Glo ( Promega ) and Lysis Buffer were added to the samples and incubated at RT for 15 mins . Luminescence was read using an Omega Luminometer . Fluorescence-activated cell sorting ( FACS ) was performed on FACSMelody Cell Sorter ( BD Biosciences ) . DAPI was added for nuclear staining . Data were processed using FlowJo 10 . 0 . 5 . Knockout of IFNAR1 and IFNLR1 in T84 cells were achieved by using the CRISPR/Cas9 system as described earlier [24] . For back-compensation of the IFN receptor KO cell lines and for generation of wild-type T84 cells overexpressing the IFNAR1 and IFNLR1 , plasmids containing the cDNA of IFNAR1 and IFNLR1 were obtained from a gateway compatible ORF bank ( pENTRY221-IFNAR1 ) and from GE Healthcare ( pCR_XL_TOPO_IFNLR1 , #MHS6278-213246004 ) , respectively . The IFNLR1-GFP construct ( pC1-HsIFNLR1-GFP ) was generated using the following cloning strategy . A mammalian expression plasmid producing a N-terminal EGFP-tagged extracellular domain of IFNLR1 ( EGFP-IFNLR1 ) was generated as follows: cDNA corresponding to this open reading from was generated synthetically ( GeneArt , Life Technologies ) and subsequently sub-cloned directly into the pC1 expression plasmid ( Promega ) backbone . Specifically , monomeric EGFP was introduced between the signal peptide sequence and the remaining glycoprotein flanked by three alanine residues at its amino terminus and a short glycine-serine linker sequence of N-AAASGSGS-C at its carboxyl terminus . Tri-alanine flanking allowed facile incorporation of restriction enzyme sites ( Not1 and SacII ) allowing removal or swapping of EGFP tag . Sequences available on request . Caspase-cleavage resistant IFNAR1 and IFNLR1 were generated using the Quick Change II XL site directed mutagenesis kit ( Agilent Technologies , Germany ) , following manufacturer’s instructions . Point mutations were controlled by plasmid sequencing . The expression vectors were generated by inserting the respective constructs into the lentiviral vector pDest GW35 by using the Gateway cloning technology ( Life Technologies , Germany ) according to manufacturer’s instructions . Lentiviruses were produced as previously described [24] , and T84 cells were transduced two times using concentrated stocks of lentiviral particles encoding the cleavage resistant IFNAR1 and IFNLR1 . 36 hours post-transduction , transduced cells were selected for using blasticidin . The mathematical model was implemented in terms of ordinary differential equations ( ODEs ) in MATLAB 2016b ( S3 Table ) . The numerical simulations were conducted using the CVODES , a module from SUNDIALS numerical simulation package , in the MATLB environment . The model was initially set to a steady state condition and most of the initial conditions were set ( S4 Table ) . Only , the IFNLR efficacy factor was estimated using time-resolved ISG expression data that we measured with different doses of type I IFN ( β ) or III IFN ( λ1−3 ) . All of the ISG expression data for the IFNAR1 and IFNLR1 overexpression experiments were reproduced only by fitting new initial values of IFNAR1 and IFNLR1 ( S5 Table ) . Parameter estimation was conducted by minimizing the weighted nonlinear least squares , wSSR=∑i=1N ( 1σi2 ) ∑j=1M ( ysimulation_i , j−yobserved_i , j ) 2 , of model simulations versus data points , j = 1 , … , M , of different experiments , i = 1 , … , N . The variance , σi2 , of every time-resolved experimental data was used as a weighting factor for fitting the corresponding data . The variance was calculated by multiplying the respective mean value with the average coefficient of variation of the experimental data . To assess the uncertainty in the estimated parameter values , we used the profile-likelihood method [70] . In this method , the parameter confidence bounds are calculated based on their contribution to the likelihoods , or in another word , the objective function ( wSSR ) . This computational approach is conducted in a stepwise manner . In every step , the respective parameter is fixed at a new value distant from the optimum estimated one . Then , the new maximum likelihood is calculated ( wSSRmin ( θ ) ) . Using this approach , we can calculate the profile of the maximum likelihoods over different values of the considered parameter . Then a threshold , Δα , Δχ2=wSSRmin ( θ ) −wSSRmin ( θoptimum ) , {θ|Δχ2<Δα} , is used to define the confidence bounds for the respective parameter . The threshold , Δα , is the α quantile of the chi-squared distribution . To investigate the effect of the parameter uncertainty on model predictions we calculated approximate 95% confidence bands , as explained in Seber and Wild [71] . Approx95%confidencebands=ysimulated±tinvN−Pα∙MSE∙ ( 1+S∙ ( S∙S ) −1∙S ) 12 where “tinvN−Pα” is the α quantile of student's t distribution , “N” is the number of data points and “P” is the number of estimated model parameters , “MSE” is the mean standard error and “S" is the sensitivity matrix of the respective simulated observable . To select the most parsimonious model , the simplest model with good predictive power , from the ensemble of the four alternative models of the ISG response to type I versus type III interferon , we used the Akaike information criterion corrected for small sample size ( AICc ) . After fitting the models to the experimental data , we calculate the AICc score for every model . AICc is calculated as: AICc=n ( ln ( 2π∙wSSRn ) +1 ) +2k+2k ( k+1 ) n−k−1 , where n is the number of data points used to fit the model , k is the number of estimated parameters of the respective model , and wSSR is the minimum weighted sum of squared residuals for the respective model . The model with the minimum AICc value is selected as the most parsimonious model from the ensemble of alternative models . In order to compare the selected model with other models , we calculate both ΔAICc , the difference between the AICc value of the models with the minimum AICc value from the ensemble of the models , and the AICc weight ( w i ) . The Akaike weight is a weight of evidence for the respective model and is calculated as: wi=exp ( −12ΔAICci ) ∑r=1Mexp ( −12ΔAICcr ) .
|
The human intestinal tract plays two important roles in the body: first it is responsible for nutrient absorption and second it is the primary barrier which protects the human body from the outside environment . This complex tissue is constantly exposed to commensal bacteria and is often exposed to both bacterial and viral pathogens . To protect itself , the gut produces , among others , secreted agents called interferons which help to fight against pathogen attacks . There are several varieties ( type I , II , and III ) of interferons and our work aims at understanding how type I and III interferon act to protect human intestinal epithelial cells ( hIECs ) during viral infection . In this study , we confirmed that both interferons can protect hIECs against viral infection but with different kinetics . We determined that type I confer an antiviral state to hIECs faster than type III interferons . We uncovered that these differences were intrinsic to each pathway and not the result of differential abundance of the respective interferon receptors . The results of this study suggest that type III interferon may provide a different antiviral environment to the epithelium target cells which is likely critical for maintaining gut homeostasis . Our findings will also help us to design therapies to aid in controlling and eliminating viral infections of the gut .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"organoids",
"medicine",
"and",
"health",
"sciences",
"luciferase",
"vesicular",
"stomatitis",
"virus",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"pathogens",
"biological",
"cultures",
"enzymology",
"microbiology",
"viruses",
"rna",
"viruses",
"organ",
"cultures",
"digestive",
"system",
"research",
"and",
"analysis",
"methods",
"proteins",
"medical",
"microbiology",
"oxidoreductases",
"microbial",
"pathogens",
"gene",
"expression",
"viral",
"replication",
"gastrointestinal",
"tract",
"biochemistry",
"rhabdoviruses",
"anatomy",
"virology",
"viral",
"pathogens",
"interferons",
"genetics",
"biology",
"and",
"life",
"sciences",
"colon",
"organisms"
] |
2018
|
Differential induction of interferon stimulated genes between type I and type III interferons is independent of interferon receptor abundance
|
Dengue hemorrhagic fever and dengue shock syndrome ( DHF/DSS ) are life-threatening complications following infection with one of the four serotypes of dengue virus ( DENV ) . At present , no vaccine or antiviral therapies are available against dengue . Here , we characterized a panel of eight human or mouse-human chimeric monoclonal antibodies ( MAbs ) and their modified variants lacking effector function and dissected the mechanism by which some protect against antibody-enhanced lethal DENV infection . We found that neutralizing modified MAbs that recognize the fusion loop or the A strand epitopes on domains II and III of the envelope protein , respectively , act therapeutically by competing with and/or displacing enhancing antibodies . By analyzing these relationships , we developed a novel in vitro suppression-of-enhancement assay that predicts the ability of modified MAbs to act therapeutically against antibody-enhanced disease in vivo . These studies provide new insight into the biology of DENV pathogenesis and the requirements for antibodies to treat lethal DENV disease .
The four serotypes of dengue virus ( DENV ) are transmitted by Aedes aegypti and Ae . albopictus mosquitoes and are endemic predominantly in tropical and sub-tropical regions of the world [1] , [2] . Syndromes associated with DENV infection range from inapparent infection to classic dengue fever ( DF ) , a debilitating self-limited disease , to life-threatening dengue hemorrhagic fever/dengue shock syndrome ( DHF/DSS ) , characterized by vascular permeability and hypotensive shock [3] . Due to several factors , including geographic expansion of the DENV mosquito vectors and increased global urbanization , trade , and travel [4] , [5] , there has been a substantial increase in both the incidence of dengue epidemics and co-circulation of the four DENV serotypes in the same region [6] . This has resulted in an increased number of severe cases in dengue-endemic regions previously known for epidemics of only mild disease [1] , [7]–[10] . While several tetravalent dengue vaccines are currently in various stages of clinical evaluation [11]–[14] , no vaccine or therapy has been licensed to prevent or treat DENV-induced disease . DENV is a member of the Flavivirus genus and is closely related to other medically important arboviruses including West Nile ( WNV ) , Japanese encephalitis , tick-borne encephalitis , and yellow fever viruses [15] , [16] . DENV has a 10 . 7-kb , positive-sense RNA genome with 5′ and 3′ untranslated regions flanking a polyprotein that encodes three structural and seven non-structural proteins [17] . Among the three structural proteins , the pre-membrane ( prM/M ) and envelope ( E ) proteins are the primary antigenic targets of the humoral immune response in humans [18]–[20] . The E protein is comprised of three domains ( I ( EDI ) , II ( EDII ) and III ( EDIII ) [21]–[24] ) , with EDII and EDIII containing the fusion peptide [25] and putative viral receptor binding site ( s ) [26] , [27] , respectively . For DENV , the most potently neutralizing antibodies generated in mice thus far target two sites on EDIII , corresponding to epitopes on the lateral ridge and A-strand [26] , [28]–[31] . However , in human dengue-immune serum after primary DENV infection , highly neutralizing type-specific antibodies appear to be directed to quaternary epitopes on adjacent E proteins present only on virons [32] . A large proportion of human anti-DENV antibodies appear to be cross-reactive and to target the fusion loop or prM [18] , [19] . Epidemiological analysis has established that a previous DENV infection is the greatest risk factor for the development of severe disease [33]–[37] . Infection with one serotype is believed to provide life-long immunity against re-infection with the same serotype but does not provide sustained protection against re-infection with a different serotype [38] , [39] . Indeed , adaptive B and T cell responses may be poorly inhibitory against re-infection with a second serotype , and in a small percentage ( ∼1% ) of cases , even exacerbate disease . One hypothesis , termed antibody-dependent enhancement , is that antibodies from a previous infection facilitate virus entry into Fcγ-receptor ( FcγR ) -bearing target cells , thereby increasing viral load and ultimately disease severity [40] . Experimental evidence in cell culture and in animal models supports this concept [41]–[44] . In a mouse model of ADE , passive transfer of monoclonal antibodies ( MAb ) or polyvalent serotype-cross-reactive serum , when administered at sub-neutralizing concentrations , was sufficient to enhance infection and cause lethal disease with DENV2 strain D2S10 in interferon α/β and γ-receptor deficient ( AG129 ) mice [42] , [43] . Recently , we showed that passive transfer of genetically engineered MAbs lacking binding to FcγR and C1q was sufficient to reduce viral load and TNF-α levels and to prevent lethal disease in vivo , even when administered one or two days after infection . Here , we evaluated the therapeutic activity of a larger panel of MAbs targeting different epitopes on the E protein following both a virus-only as well as an antibody-enhanced lethal infection . We determined that the two most potent therapeutic MAbs acted by competitively displacing either fusion-loop specific MAbs or enhancing polyclonal serum antibodies targeting a proximal epitope . Using this information , we designed a novel suppression-of-enhancement assay in human FcγRIIA-expressing K562 cells that predicts the ability of modified MAbs to act therapeutically against antibody-enhanced disease in vivo . Our observations provide new insight into the mechanism by which therapeutic MAbs prevent an antibody-enhanced lethal DENV infection .
Severe forms of DENV infection , including DHF/DSS , can be fatal , as no specific antiviral therapy is currently available . As such , we extended previous observations of the prophylactic and therapeutic efficacy of the EDII fusion loop-specific MAb E60 [42] by studying a larger panel of neutralizing MAbs targeting additional E protein epitopes , including the dimer interface ( E44 ) on EDII and the C-C′ loop ( E87 ) and A-strand ( E76 and 87 . 1 ) on EDIII ( Figure 1A ) . Although secondary infection with a different DENV serotype is the greatest risk factor for severe DENV disease , DHF/DSS also has been reported following primary infection [45] . Thus , for a genetically-modified MAb to be a viable therapeutic option , it must protect following both a virus-only and an antibody-enhanced lethal DENV infection . To assess the ability of MAbs to protect in a direct model of lethal DENV infection , AG129 mice were infected with a lethal dose ( 4×106 PFU ) of DENV2 D2S10 and 24 hours later , administered 20 µg of individual genetically modified MAbs lacking the ability to bind FcγR or C1q . Notably , all of the modified MAbs tested prevented development of overt disease and protected against death in this model ( P<0 . 05 for all MAbs as compared to untreated mice , Figure 1B , Table S1 ) . We subsequently assessed whether these MAbs also protected against antibody-enhanced lethal DENV infection . Anti-DENV1 serum was administered 24 hours prior to a sub-lethal infection ( 105 PFU ) of DENV2 D2S10 , and animals were treated 24 hours post-infection with 20 µg of genetically-modified MAbs lacking effector functions , where the N297Q variant MAbs are fully aglycosylated and the LALA variant MAbs remain glycosylated but incapable of binding either FcγR or C1q [42] , [46] . Of those tested , only the EDII fusion loop-specific E60 N297Q and EDIII A strand-specific 87 . 1 LALA MAbs completely prevented mortality ( P<0 . 01 , Figure 1C and Table S1 ) . In comparison , the EDIII-A-strand-specific MAb E76 N297Q showed partial protection ( P<0 . 05 ) , whereas MAbs E44 N297Q ( EDII dimer interface ) and E87 N297Q ( EDIII C-C′ loop ) provided no protection against lethal disease ( Figure 1C , Table S1 ) . To determine why some MAbs had therapeutic activity in the virus-only lethal infection model but not in the context of antibody-enhanced infection , we examined several properties including epitope specificity , neutralization potency , and avidity . We first assessed whether neutralization potency correlated with in vivo therapeutic potential . The neutralizing activity against DENV2 D2S10 of each of the MAbs was assessed using a flow cytometry-based assay with human monocytic U937 cells expressing DC-SIGN , a known attachment factor for DENV [47] . The potency of each intact and modified MAb was assessed and expressed as the 50% neutralization titer ( NT50 in ng/ml of MAb ) . No significant difference was observed between each intact MAb and its modified variant . The NT50 of therapeutically effective MAbs E60 N297Q ( EDII fusion loop ) and 87 . 1 LALA ( EDIII A strand ) were 72 ng/mL and 24 ng/mL , respectively . In comparison , the NT50 of MAbs E44 N297Q ( EDIII C-C′ loop ) and E87 N297Q ( EDII dimer interface ) , which bound other epitopes in EDII and EDIII and lacked therapeutic activity , were similar ( 68 ng/mL and 95 ng/mL , respectively ) ( Table 1 ) . Thus , NT50 values among MAbs targeting different epitopes failed to demonstrate a clear relationship between neutralizing potency and in vivo therapeutic efficacy in the context of antibody-enhanced lethal infections ( Spearman ρ 0 . 47 , P = 0 . 45 ) . We hypothesized that MAb avidity , the strength of binding between a bivalent antibody and two ligands on a single virion or across virions , might correlate better with therapeutic efficacy following an antibody-enhanced lethal infection . To test this hypothesis , we measured the avidity ( Kdapp ) of binding , using a direct , virion-coated ELISA [30] . While we noted a correlation between MAb neutralization titer and avidity ( Spearman ρ 0 . 9 , P<0 . 083 ) , analogous to our neutralization data , we did not observe a relationship between MAb avidity and therapeutic efficacy ( Spearman ρ 0 . 32 , P<0 . 68 ) by MAbs targeting non-fusion loop epitopes ( Table 1 ) . As neutralization potency and avidity failed to correlate directly with the therapeutic efficacy of our modified MAbs across different epitopes , we investigated whether epitope specificity had greater predictive potential . As MAb E60 N297Q was highly protective even 48 hours following antibody-enhanced DENV infection [42] , we hypothesized that the fusion loop epitope might be an important target for therapeutic MAbs . Therefore , we tested the therapeutic activity following either a virus-only or an antibody-enhanced lethal DENV infection of three additional modified MAbs that also target the EDII fusion loop but displayed between 2- to 8-fold reduced neutralization potency compared to MAb E60 N297Q: 82 . 11 LALA ( NT50 131 ng/ml ) , E18 N297Q ( NT50 363 ng/mL ) and E28 N297Q ( NT50 544 ng/mL ) ( Table 1 ) . Whereas all of the animals treated with MAb 82 . 11 LALA , E18 N297Q or E28 N297Q survived infection after a virus-only lethal challenge ( P<0 . 01 compared to PBS-treated mice , Figure 2A and Table S2 ) , MAbs E18 N297Q and E28 N297Q failed to confer a therapeutic benefit following an antibody-enhanced lethal infection ( Figure 2B , Table S2 ) . MAb 82 . 11 LALA protected 50% ( 3/6 ) of animals following an antibody-enhanced infection , though this difference trended but did not attain statistical significance compared to non-treated control animals ( Figure 2B , Table S2 ) . In contrast to the experiments with non-fusion loop-specific MAbs , studies with MAbs targeting the same fusion loop epitope suggest that neutralization potency can predict therapeutic efficacy following an antibody-enhanced infection ( Spearman ρ 0 . 9487 , P<0 . 083 ) . To explain the correlation between in vitro neutralizing potency and in vivo therapeutic efficacy within fusion loop-specific MAbs , we generated a model of competitive displacement . Recent work has suggested that a significant fraction of the human anti-flavivirus E protein antibody response is directed against the fusion loop epitope in EDII [20] , [48]–[50] . We hypothesized that these cross-reactive antibodies found in DENV-immune serum are of intermediate or low affinity and bind to the heterologous virus at a stoichiometry insufficient for neutralization but adequate for enhancement of infection [51] . However , after administration of a therapeutic , high-affinity , genetically-modified fusion loop-specific MAb , natural dissociation of the enhancing antibody occurs , and the more avid therapeutic MAb binds to the fusion loop epitope , effectively preventing enhancing antibodies from binding again to the virion . Additionally , highly avid modified MAbs would compete favorably with enhancing antibodies for binding to nascently-produced virions . In this scenario , modified MAbs lacking effector functions either coat the virion allowing for direct neutralization or compete against cross-reactive fusion-loop enhancing antibodies in serum , such that the stoichiometry required for enhancement [51] is not reached . To test this hypothesis , we used 4G2 , a weakly neutralizing ( NT50 of 393 ng/mL ) mouse MAb that binds to the fusion loop epitope [49] to enhance an otherwise sub-lethal DENV2 D2S10 infection and administered 20 µg of the modified MAbs 24 hours post-infection . E60 N297Q , the most therapeutic fusion loop-specific MAb in the context of a polyvalent serum-enhanced infection , again achieved 100% protection ( P<0 . 01 ) when administered after a 4G2-enhanced infection , whereas MAb 82 . 11 LALA was less therapeutic ( P<0 . 05 ) , protecting 4 of 6 treated animals ( Figure 2C , Table S3 ) . None of the animals treated with MAb E18 N297Q succumbed to infection ( P<0 . 05 ) , although all demonstrated signs of illness ( P<0 . 05 as compared to E60 N297Q-treated mice , Table S3 ) . However , mice treated with MAb E28 N297Q all succumbed to 4G2-enhanced DENV2 D2S10 infection . These data support a model in which modified fusion loop-specific MAbs of sufficient avidity and neutralizing potency compete effectively for binding sites in the context of enhancing polyvalent DENV-immune serum or other fusion loop-specific MAbs to prevent disease . We next evaluated directly whether MAb E60 ( EDII fusion loop-specific ) could effectively compete for binding with less potent fusion loop-specific MAbs , as compared to either therapeutic MAb 87 . 1 ( EDIII A strand-specific ) or non-therapeutic MAb E87 ( EDII C-C′ loop-specific ) that both target distinct epitopes . After directly coating DENV2 virions on microtiter plates , we added the moderately neutralizing mouse MAb 4G2 at 1 µg/mL mixed with increasing concentrations ( 0 . 1 , 1 and 10 µg/mL ) of modified human MAbs followed by an anti-mouse , Fc-specific secondary MAb . Binding of mouse MAb 4G2 was not affected by the amount of bound E87 ( non-therapeutic , C-C′ loop-specific ) ( P = 0 . 64 by Friedman's analysis of data combined from seven experiments ) . In contrast , both MAb E60 ( therapeutic , fusion loop-specific ) and , surprisingly , MAb 87 . 1 ( therapeutic , A strand-specific ) altered binding of MAb 4G2; higher concentrations of MAb E60 and MAb 87 . 1 resulted in lower amounts of MAb 4G2 bound ( P≤0 . 001 for both E60 and 87 . 1 by Friedman's analysis of data combined from seven experiments , Figure 3A ) . Less potently neutralizing and non-therapeutic fusion loop-specific MAbs competed as or less effectively against MAb 4G2 for binding to the fusion loop epitope ( Figure S1A ) . We next tested whether modified MAbs targeting the same or different epitopes with respect to the enhancing MAb 4G2 ( fusion loop-specific MAb ) could suppress enhancement in vitro . We mixed serial dilutions of both 4G2 and the modified MAbs E60 N297Q ( therapeutic EDII fusion loop-specific , Figure S2A ) , 87 . 1 LALA ( EDIII , A strand-specific , Figure S2B ) and E87 N297Q ( EDII dimer interface-specific , Figure S2C ) in the following ratios: 100% 4G2; 95% 4G2 and 5% modified MAb; 85% 4G2 and 15% modified MAb; 75% 4G2 and 25% modified MAb . Each MAb combination was incubated with DENV2 D2S10 virus and used to infect K562 cells , a human erythroleukemic cell line that expresses FcγRIIA ( CD32A ) and is non-permissive in the absence of enhancing anti-DENV antibodies . Infection was monitored after 48 hours by intracellular DENV antigen staining and quantified by flow cytometry . Peak enhancement with MAb 4G2 occurred at 466 ng/mL and resulted in ∼7 to 15% of the cells becoming infected ( Figure S2 ) . However , when E60 N297Q or 87 . 1 LALA , MAbs that were effective therapeutically and lack the ability to engage Fcγ receptors , comprised only 5% of the MAb population , peak enhancement by 4G2 was reduced by 33% and 65% , respectively ( Figure 3B , Figure S2 ) . Remarkably , when 4G2 and the modified MAbs were mixed in a 75%: 25% ratio , E60 N297Q and 87 . 1 LALA both reduced infection by 85% and 86% , respectively ( Figure 3D , Figure S2 ) . In comparison , mixture of 5% of the non-therapeutic modified MAb E87 N297Q , reduced enhancement by only 21% ( P<0 . 04 , compared to E60 N297Q and P<0 . 02 compared to 87 . 1 LALA ) , and by 57% ( P<0 . 01 , compared to E60 N297Q and P<0 . 02 compared to 87 . 1 LALA ) when a 75%: 25% mixture was used ( Figure 3B–3D ) . Similarly to E87 N297Q , MAbs E18 N297Q and E28 N297Q , both fusion loop-specific but non-therapeutic MAbs , reduced enhancement by 19% and 6% when E18 N297Q and E28 N297Q were 5% of the MAb population , respectively , and by 46% when either E18 N297Q or E28 N297Q comprised 25% of the MAb population ( Figure S1B–D ) . Thus , highly avid , fusion-loop specific MAb E60 N297Q and A-strand-specific MAb 87 . 1 LALA minimized in vitro enhancement , presumably by preventing binding of the enhancing fusion loop-specific MAb 4G2 . Furthermore , the ability of the modified MAbs to prevent 4G2-mediated enhancement in vitro correlated with in vivo therapeutic activity in the context of anti-DENV polyvalent serum-enhanced infection . Given the results in K562 cells with mixtures of modified MAbs and decreased 4G2-mediated enhancement , we wanted to evaluate further this relationship in vivo . We administered 20 µg of MAb 4G2 24 hours prior to infection with DENV2 D2S10 and then treated mice one day post-infection with 20 µg of MAb E60 N297Q ( EDII fusion loop-specific ) ) , 87 . 1 LALA ( EDIII A strand-specific ) , or E87 N297Q ( EDIII C-C′-loop-specific ) . While MAb E87 N297Q was not therapeutically protective against a 4G2-enhanced lethal infection , E60 N297Q and 87 . 1 LALA protected against mortality in 6 of 6 and 5 of 6 animals , respectively ( P<0 . 05 , Figure 3E , Table S3 ) . Thus , potently neutralizing , fusion loop-specific ( E60 N297Q ) and A strand-specific ( 87 . 1 LALA ) MAbs both prevent antibody-enhanced disease , likely by displacing binding of enhancing MAbs that target the fusion loop epitope . We next assessed how different ratios of intact and genetically modified variants affected enhancement in K562 cells in vitro . We selected the two most therapeutically effective , modified MAbs ( E60 N297Q ( EDII fusion loop ) and 87 . 1 LALA ( EDIII A strand ) ) , and mixed them with the intact parent MAbs in the following proportions: 100% intact MAb , 90% intact and 10% modified MAb , 75% intact and 25% modified MAb , 50% of each MAb , 25% intact and 75% modified MAb , 10% intact and 90% modified MAb , and 100% modified MAb . While several mixtures of E60 parent:E60 N297Q showed reduced enhancement , only the combination of 10% intact:90% modified was non-enhancing in vitro , suggesting that the majority of the antibody mixture must not bind to Fcγ receptors in order to abolish enhancement of DENV infection when MAb pairs are of comparable neutralizing potency and avidity ( Figure 4A ) . The combination of intact 87 . 1 and 87 . 1 LALA also demonstrated a complete reduction in enhancement , but this occurred under conditions where a lower ratio of intact to modified mAb was required ( ratios of 25% intact:75% aglycosylated ( Figure 4B ) ) . The relative differences in enhancement profiles observed between the E60:E60 N297Q and 87 . 1:87 . 1 LALA MAb pairs could be due to the small difference in the avidity and neutralization potency of the intact and modified MAb . Similar relationships between modified and intact MAb pairs were observed when studying MAbs that were moderately ( E76/E76 N297Q ) and poorly ( E18/E18 N297Q ) therapeutic ( data not shown ) . Using combinations of intact E60 and modified E60 N297Q ( EDII fusion loop-specific MAb ) , we evaluated whether the requirement for 90% of the MAb mixture to lack FcγR binding for suppression of enhancement in vitro translated into therapeutic efficacy in vivo . The same ratios were mixed in a total of 20 µg and administered therapeutically 24 hours after serum-enhanced DENV2 infection of AG129 mice . Notably , and consistent with our data in K562 cells , complete therapeutic protection in vivo required 90% of the E60 mixture to be present in the modified form ( P<0 . 02 , Figure 4C ) . Mixtures that were combined in a ratio of less than 9∶1 showed reduced or no therapeutic efficacy ( Figure 4C ) . This in vivo data suggests that when the same MAb is used for enhancement and therapy ( intact versus modified ) , the majority of the mixture must lack the capacity for binding FcγR to avoid enhancement . Thus , a low stoichiometric threshold of binding is likely sufficient for enhancement of infection and disease . Given the results with intact and modified MAbs , we evaluated whether we could use this in vitro relationship to predict the ability of modified MAbs to be therapeutically effective in vivo in the context of immune serum-enhanced DENV infection . Initially , using DENV1-immune mouse serum , we identified the serum dilution ( 1∶180 ) responsible for peak enhancement of DENV2 D2S10 in K562 cells ( Figure 5A ) . We then tested the ability of modified MAbs to reduce enhancement in K562 cells by pre-incubating D2S10 with the peak enhancing dilution of DENV1-immune serum for 30 minutes , then adding increasing amounts of modified MAbs for 30 minutes , followed by incubation with K562 cells for 48 hours . Importantly , the concentrations of DENV-immune serum and virus used in the in vitro assay were comparable to those used in the in vivo infections . At concentrations of 2 , 000 ng/mL and 1 , 000 ng/mL , modified MAbs with moderate to strong ( >60% protection ) therapeutic activity in vivo were more efficient ( P<0 . 05 ) at suppressing ADE in K562 cells than MAbs that were less therapeutically active ( Figure 5B and 5C ) . The three most therapeutically effective MAbs ( 87 . 1 LALA , E60 N297Q and E76 N297Q ) reduced enhancement on average by 88% , 70% and 65% , respectively , when added at 1 , 000 ng/mL while less protective MAbs ( 82 . 11 LALA , E44 N297Q , E87 N297Q , E18 N297Q and E28 N297Q ) reduced enhancement less efficiently ( Figure 5C ) . This trend also was observed when DENV1-immune serum was added at a different enhancing concentration ( 1∶540 dilution ) ( data not shown ) . Based on these data that differentiate in vitro therapeutic from non-therapeutic MAbs , we established ∼50% reduction at 1 , 000 ng/mL as the criterion for predicting therapeutic efficacy using the suppression-of-enhancement assay . As the K562 cell-based assay with mouse polyclonal anti-DENV1 serum and modified MAbs appeared to predict in vivo outcomes , we repeated the experiments with DENV-immune human serum; this was important as humans and mice produce overlapping yet distinct antibody repertoires against flavivirus epitopes [19] , [48] , [52] , [53] . We evaluated whether modified MAbs reduced enhancement in K562 cells using human DENV-immune serum collected years after a primary DENV4 infection . The peak serum enhancement dilution again was identified as between 1∶180 and 1∶540 ( Figure 6A ) . In contrast to the limited protection provided by the modified MAbs following a mouse DENV1-serum enhanced infection , most modified MAbs suppressed enhancement by DENV4 human immune serum below the 50% cut-off at the higher ( P<0 . 05; E18 N297Q and E28 N297Q , P<0 . 08 ) , yet physiologically relevant concentrations ( 1 and 2 µg/mL ) of modified MAb ( Figure 6B , Figure S3 ) , while non-binding , DENV4-specific MAb 22 . 3 LALA did not ( Figure 6B ) . Similar results were obtained when primary DENV1 or DENV3 human immune serum was tested ( Figure S3 ) . To determine whether the enhancement data with human serum predicted protection in vivo , we administered normal human serum ( NHS ) or enhancing amounts of anti-DENV4 human immune serum 24 hours prior to a sub-lethal infection with DENV2 D2S10 , and tested the therapeutic efficacy of the modified MAbs . As expected , all mice pre-treated with NHS survived infection without any signs of morbidity . All mice receiving enhancing anti-DENV4 human immune serum and treated with a modified MAb ( E60 N297Q , 87 . 1 LALA , 82 . 11 LALA , E87 N297Q , and E28 N297Q ) survived lethal enhanced infection with minimal signs of disease ( P<0 . 05 for all modified MAbs compared to PBS-treated controls , Figure 6C ) , whereas mice treated with modified , DENV4-specific MAb 22 . 3 LALA did not ( Figure 6C ) . Thus , the suppression-of-enhancement assay in K562 cells correlated with the therapeutic efficacy of modified MAbs in vivo in an antibody-enhanced lethal DENV model in the context of both mouse and human DENV immune serum . Moreover , and for reasons that likely relate to the distinct repertoire of cross-reactive enhancing antibodies in human serum , modified MAbs against the EDIII A-strand , EDIII C-C′ loop , and EDII fusion loop all efficiently suppressed antibody enhancement in cell culture and in vivo .
In this report , we analyzed a panel of eight MAbs that bind to several epitopes on the dengue virion , including the fusion loop and dimer interface on EDII and the A strand and C-C′ loop on EDIII . We determined that differences exist between the ability of modified MAbs lacking the capacity to engage FcγR and C1q to act therapeutically following a virus-only lethal infection and an antibody-enhanced lethal infection . Analysis of MAb characteristics such as binding avidity and neutralization potency did not clearly define an in vitro correlate of in vivo efficacy across different epitopes , but were more predictive when studying MAbs targeting a specific class , such as those binding the fusion loop epitope . Further analysis suggested that modified , fusion loop- and A-strand-specific MAbs act therapeutically by competing against enhancing antibodies in polyvalent serum that recognize the same or proximal epitopes . By studying these relationships , for the first time , we established a novel in vitro suppression-of-enhancement assay with polyclonal mouse and human anti-DENV immune serum that appears to predict the ability of modified MAbs to act therapeutically against ADE in vivo . Thus , we provide in vivo data that support in vitro observations about the mechanism of ADE as well as a means to suppress ADE in vivo . Multiple parameters , including neutralization potency , avidity and epitope specificity , affect whether a modified MAb is therapeutic against an antibody-enhanced DENV infection . In our panel , in addition to binding to either the fusion loop or A-strand epitope , a therapeutic MAb needed to be strongly neutralizing ( NT50<100 ng/mL ) , which itself is a function of epitope accessibility on the virion , mechanism of inhibition , and avidity of binding [51] . Four of the MAbs tested ( E60 , 82 . 11 , E18 , and E28 ) recognize similar residues within the EDII fusion loop ( [54] and S . Sukupolvi-Petty and M . S . Diamond , unpublished data ) , but two ( E18 and E28 ) had lower neutralizing potency and avidity of binding to the virion , and , correspondingly , showed less or no therapeutic activity in vivo following DENV enhancement by polyvalent mouse serum . While the avidity of binding to solid-phase DENV2 for 82 . 11 LALA and E60 N297Q was comparable , E60 N297Q is ∼2 . 5 fold more neutralizing , suggesting that the two MAbs might bind overlapping yet slightly distinct epitopes , or that the ensemble of viral conformations in solution [55] allows for enhanced recognition of E60 relative to 82 . 11 . Analogously , when comparing two modified MAbs targeting the A-strand in EDIII , MAb 87 . 1 LALA showed higher avidity of binding and therapeutic efficacy in vivo compared to MAb E76 N297Q . Although further study is warranted , our data suggest that within an epitope class , there is a direct relationship between MAb avidity and neutralization potential in vitro and therapeutic efficacy in vivo . Studies comparing therapeutic efficacy following virus-only and mouse antibody-enhanced lethal DENV2 infections revealed that all modified MAbs tested were therapeutic following a virus-only infection , but only two ( E60 N297Q and 87 . 1 LALA ) were completely protective following antibody-enhanced infection with DENV1-immune mouse serum . This observation suggests a direct interplay between the enhancing antibodies in polyvalent serum and the neutralizing therapeutic MAbs that determines outcome . Thus , a second parameter affecting therapeutic efficacy is the ability of a modified MAb to out-compete the enhancing antibodies in polyvalent immune serum for binding to the virion . This concept is supported by functional data in vitro and in vivo using the weakly neutralizing , fusion loop-specific MAb 4G2 and a panel of modified fusion loop-specific MAbs . In cellular assays , the more avid and strongly neutralizing MAb E60 N297Q was more effective at suppressing 4G2-enhanced infection in K562 cells than the less potent E18 N297Q and E28 N297Q MAbs . Consistent with this , E60 N297Q but not E28 N297Q prevented mortality as a therapeutic when MAb 4G2 was used to enhance a sub-lethal DENV2 D2S10 infection . Our data support a model in which therapeutic activity occurs when high-affinity , modified MAbs can bind to virions and neutralize infection by competing with and/or displacing enhancing antibodies for binding to similar epitopes . An additional parameter that likely affects therapeutic efficacy of a MAb against antibody-enhanced DENV infection is its mechanism of action: whether the MAb binds prior to or following attachment of the virion to the target cell . However , the interpretation is not straightforward , as mechanism of neutralization of a given MAb may be affected by several variables: ( a ) stoichiometry and relative fractional occupancy at a given concentration [51]; ( b ) cell type and repertoire of attachment ligands or receptors [51] , [54]; ( c ) virus particle maturation [56]; and ( d ) dynamic state of the virion [55] . From in vitro ADE experiments using K562 cells and non-modified MAbs , we can conclude that all MAbs ( excluding E44 as it was not available for these studies ) have the capacity to neutralize infection via a post-attachment mechanism at saturating concentrations ( Figure 4 and Figure S2 ) . Following uptake via FcγR , antibody-enhanced DENV may still be neutralized by MAbs that block post-attachment – this phenomenon of trans-dominant neutralization of ADE by MAbs was described previously with the anti-WNV MAb E16 [57] , a MAb that neutralizes WNV infection by blocking the structural changes required for viral fusion [58] , [59] . Indeed , at saturating concentrations , all of the non-modified WT MAbs in our panel reduce K562 infection to background levels ( Figures 4A and B , and Figure S2 , right side of the curve ) , most likely by blocking fusion , a critical step required for release of the DENV genome into the cytosol following receptor-mediated endocytosis . To distinguish this post-attachment neutralization pattern from a MAb that blocks via a pre-attachment mechanism and cannot prevent infection in a K562 assay , we can compare these data to the effects of anti-fusion loop MAb E60 on WNV infection . MAb E60 cannot prevent WNV infection in K562 cells even at saturating concentrations ( E . Mehlhop and M . S . Diamond , unpublished data ) . The inability of MAb E60 to neutralize WNV in K562 cells occurs because WNV virions are present in a mature state to a far greater degree than DENV and therefore have a lower stoichiometry of binding for the fusion loop epitope [56] , [60] . For WNV , MAb E60 fails to achieve a stoichiometry sufficient to block fusion of virus that has entered via FcγR-dependent enhancement . In contrast , in the current study , all MAbs appear to act in a post-attachment mechanism in K562 cells at saturating concentrations . However , in vivo , it is unlikely that the modified MAbs circulate at saturating concentrations , given the large amount of DENV virus and antigen present . Thus , the relevant question becomes which MAbs can reduce enhancement ( generated by either polyvalent DENV-immune sera or MAbs such as fusion loop-specific 4G2 ) most efficiently when the modified MAbs are at sub-saturating concentrations . Under these conditions , therapeutically effective MAbs ( 87 . 1 LALA and E60 N297Q ) show greater efficacy than the other MAbs evaluated , likely due to their ability to compete for binding with enhancing antibodies , which interferes with FcγR crosslinking and limits DENV uptake and infection . Remarkably , MAb 87 . 1 LALA , which maps to the A strand of EDIII , also was effective against anti-fusion loop MAb 4G2-mediated lethal DENV infection . While not as potent as MAb E60 , MAb 87 . 1 appeared to compete with fusion loop-specific MAb 4G2 in the solid-phase DENV2 ELISA and also suppressed 4G2-induced enhancement in K562 cells . One possible explanation is that the A-strand epitope on EDIII is located next to the EDII fusion loop on adjacent DENV E proteins within a dimer [23] , such that on the virion in solution , high avidity binding of 87 . 1 LALA prevents lower-avidity fusion loop-specific enhancing MAbs ( e . g . , 4G2 ) from binding . While all MAbs tested appear to block DENV in a post-attachment mechanism at saturating concentrations , MAb 87 . 1 may be more potent at blocking fusion at sub-saturating concentrations than MAb E60 . This hypothesis may explain why 87 . 1 LALA is more efficient at reducing 4G2-enhanced DENV infection in K562 cells , but less efficient at competing with MAb 4G2 in a fixed-virion ELISA than MAb E60 N297Q . Another possible explanation is that binding of the A-strand MAb 87 . 1 LALA alters the conformation of the mature DENV virion [61] , [62] , enhancing exposure of the fusion loop epitope and increasing binding and neutralization . Even though E87 N297Q is unable to compete for binding with fusion loop-specific MAb 4G2 in the solid phase assay , it can still bind to the virion and contribute to the stoichiometry required to neutralize DENV , thus accounting for the ∼50% reduction in in vitro enhancement when E87 N297Q comprises 25% of the antibody mixture . Despite this , E87 N297Q did not have therapeutic activity in vivo when 4G2 was used as the enhancing MAb . In addition , MAb E87 N297Q was less efficient at reducing MAb 4G2-enhanced infection than the therapeutically effective MAbs E60 N297Q and 87 . 1 LALA at any of the three conditions tested ( 5% , 15% or 25% modified MAb ) . Although more study is warranted , we speculate that the MAbs which bind epitopes that do not displace 4G2 enhancing MAbs did not protect in vivo because they failed to reach a stoichiometry that was sufficient for neutralization or do not block a post-attachment step ( e . g . , viral fusion ) . Previous studies have established that the epitope repertoire of anti-flavivirus neutralizing antibody in mouse and human serum is different . Mice were found to generate neutralizing antibody responses against epitopes in EDIII ( ∼30% ) [24] , [26] , [28] , [30] , [63] , [64] that can be serotype-specific [29] , [30] , [65] or cross-reactive [28] , [29] , [31] , [66] . In comparison , DENV-immune human serum preferentially targets the fusion loop epitope in EDII [52] , [53] as well as complex quaternary epitopes near the EDI-DII hinge that span adjacent E proteins within a dimer [32] , with little EDIII-specific neutralizing antibody generated ( 10–15% ) [48] , [64] , [67] , [68] . While it has not been explicitly studied , it seems plausible that the epitope repertoire for enhancing antibodies against DENV in human and mouse serum also vary . In support of this , we observed differences in the ability of modified MAbs to prevent antibody-enhanced lethal DENV infection when DENV-immune mouse or human serum was used . Only E60 N297Q , 87 . 1 LALA and E76 N297Q were therapeutically effective against infection enhanced with anti-DENV1-immune mouse serum . In contrast , all modified DENV2-reactive MAbs were therapeutic following an infection enhanced with DENV4-immune human serum . These data likely imply one of two non-mutually exclusive hypotheses: ( a ) cross-reactive enhancing MAbs present in DENV-immune human serum are weakly avid , such that higher affinity modified MAbs can bind and/or displace the enhancing antibodies , resulting in therapeutic protection in vivo; ( b ) cross-reactive enhancing MAbs present in DENV-immune human serum bind distinct epitopes , which do not interfere with binding and neutralization by modified MAbs targeting the EDII fusion loop , EDII dimer interface , EDIII A strand , or EDIII C-C′ epitopes . In possible support of this , recent studies of the human antibody repertoire against DENV suggest that anti-prM antibodies are a major component of the cross-reactive response and promote enhancement in vivo [18] , [20] . Future studies that test the therapeutic efficacy of modified E protein MAbs in the presence of enhancing concentrations of prM-specific MAbs will be important to perform . One limitation of this study is the passive transfer model used to develop lethal DENV disease; we tested limited concentrations of enhancing polyvalent immune serum to distinguish between therapeutic and non-therapeutic modified MAbs . In the future , a more detailed dose-response study with different enhancing sera or MAbs will be needed to determine the range of efficacy of modified MAbs in mediating protection . In addition , while the passive transfer model of enhancement and protection may be relevant for infant DHF/DSS where potentially enhancing antibodies are received passively in utero , it remains uncertain if similar principles apply during natural secondary DENV infection . In summary , our results suggest a model in which neutralization , avidity , and epitope specificity contribute to the therapeutic efficacy of modified MAbs . Despite the differences between mouse and human polyvalent antibody repertoires , the suppression-of-enhancement assay in K562 cells accurately predicted in vivo therapeutic efficacy in both situations . While further study is needed , this assay could be used to screen additional modified MAbs for potential use as DENV therapeutics . Overall , given these promising results , we suggest that further exploration of the utility of modified MAbs as therapy for DENV infections is warranted .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee at the University of California Berkeley ( R252-1012B ) . All viruses were propagated in Aedes albopictus C6/36 cells ( American Type Culture Collection ) and titered by plaque assay on baby hamster kidney cells ( BHK21 , clone 15 ) [69] . DENV2 D2S10 was generated as previously described [70] . All in vitro neutralization and enhancement assays and sub-lethal in vivo infections were performed with non-concentrated virus . DENV2 D2S10 virus was concentrated by ultra-centrifugation for use in the virion ELISA and by centrifugation using 100 , 000 MWCO Amicon filters ( Millipore ) for lethal , virus-only in vivo infections . U937-DC-SIGN ( gift from A . de Silva , University of North Carolina , Chapel Hill ) and K562 cells were used for flow cytometry-based in vitro neutralization [71] and enhancement assays [64] , respectively . Mouse MAbs E60 , E18 and E28 were generated against WNV E protein , but are cross-reactive with DENV E protein [54] . Anti-DENV2 MAb E44 , E76 , and E87 were generated against DENV2 and described previously [28] . All mouse MAbs were purified by protein A affinity chromatography ( Invitrogen , Carlsbad , CA ) and have been mapped previously [29] , [72] . The generation of a chimeric human-mouse E60 MAb with the human IgG1 constant region and the mouse VH and VL region was performed as described previously [72] . The generation of chimeric E18 , E28 , E44 , E76 , and E87 MAbs was performed similarly . Point mutations in the Fc region ( N297Q ) that abolish Fcγ receptor and C1q binding were introduced by QuikChange mutagenesis ( Stratagene ) . All recombinant MAbs were produced after transfection of HEK-293T cells , harvesting of supernatant , and purification by protein A affinity chromatography . The accession numbers for the sequences of the VH-VL regions of the recombinant MAbs are as follows: E18_VL KC254882 E18_VH KC254883 E28_VL KC254884 E28_VH KC254885 E44_VL KC254886 E44_VH KC254887 E60_VL KC254888 E60_VH KC254889 E76_VL KC254890 E76_VH KC254891 E87_VL KC254892 E87_VH KC254893 . MAbs 87 . 1 and 82 . 11 are fully human MAbs , and their generation has been described previously [20] . Production of the LALA variants was performed according to a previously published protocol [20] . Recombinant MAbs were produced in HEK-293T cells and purified by sequential protein A affinity chromatography and size-exclusion chromatography . The accession numbers for the sequences of the VH-VL regions of the recombinant human MAbs are as follows: DV82VH KC294013 , DV82VL KC294014 , DV87VH KC294015 , DV87VL KC294016 . The single DENV1- and DENV3-immune human sera samples used in the in vitro suppression-of-ADE assay were de-identified and pre-collected as part of the Nicaraguan Pediatric Dengue Hospital-based Study [73] . Both serum samples were collected three months post-symptom onset and were obtained from individuals with a primary DENV infection [74] , [75] . The protocol for the study was reviewed and approved by the Institutional Review Boards ( IRB ) of the University of California ( UC ) , Berkeley , and of the Nicaraguan Ministry of Health . Parents or legal guardians of all subjects provided written informed consent , and subjects 6 years of age and older provided assent . The primary DENV4-immune human serum sample used for in vitro suppression-of-ADE and in vivo ADE experiments was a gift from Dr . Aravinda de Silva ( University of North Carolina ( UNC ) , Chapel Hill ) and were received as de-identified and pre-collected samples , and as such were not considered part of human subjects research by the IRB at UC Berkeley . Convalescent DENV-immune sera were obtained at UNC Chapel Hill from volunteers who had experienced natural DENV infections during travel abroad . The protocol for recruiting and collecting blood samples from returned travelers was approved by the IRB of UNC Chapel Hill . Written informed consent was obtained from all subjects before collecting blood . All procedures were pre-approved and conducted according to UC Berkeley Animal Care and Use Committee guidelines . AG129 mice [76] were bred at UC Berkeley . AG129 mice were infected intraperitoneally ( i . p . ) with 105 PFU of DENV1 strain 448 ( gift from S . Kliks ) . Six to eight weeks post-infection , mice were sacrificed , and whole blood was collected by terminal cardiac puncture . Serum was isolated from whole blood by centrifugation , heat inactivated , and stored at −80°C . DENV2 D2S10 enhanced disease: AG129 mice were administered either 25 µl mouse anti-DENV1-immune serum or 200 µl human anti-DENV-4 immune serum or 20 µg of MAb 4G2 ( in 200 µl final volume ) i . p . 24 hours prior to infection with an intravenous ( i . v . ) sub-lethal , 105 PFU dose of DENV2 D2S10 . DENV2 D2S10 virus-only lethal disease: AG129 mice were infected i . v . with 4×106 PFU of DENV2 D2S10 . Treatment model: Mice were administered 25 µl of DENV1-immune serum on Day −1 , 105 PFU of DENV2 D2S10 on Day 0 , and 20 µg of modified MAbs in a final volume of 100 µl i . v . 24 hours following infection ( Day +1 ) . All animals were monitored carefully for morbidity and mortality for 10 days following infection . The neutralization titer of each parent and modified MAb was measured using the U937-DC-SIGN flow cytometry-based neutralization assay as described [71] . NT50 titers were calculated as described previously [42] . Each NT50 titer is the average of between 3 and 5 individual experiments , with the exception of E44 N297Q ( 1 experiment ) . Polyvalent serum: Enhancement curves were generated as described in [64] . Briefly , eight , three-fold dilutions of DENV-immune serum beginning at 1∶10 were pre-mixed with DENV2 D2S10 prior to addition to K562 cells . A Gaussian distribution was used to fit each enhancement curve , where percent infection was recorded on the y-axis and log-reciprocal serum dilution on the x-axis . The area under the curve ( AUC ) for each enhancement infection was calculated using Prism software . MAb competition: Intact and modified MAbs were pre-mixed in different ratios at a starting concentration of 40 µg/mL . Eight 3-fold dilutions were incubated in either duplicate or triplicate with DENV2 D2S10 at an MOI of 0 . 1 in a 1∶1 ratio before being added to K562 cells for an enhancement of infection assay [64] . DENV-immune serum was diluted to the concentration that resulted in the greatest enhancement of DENV2 infection in K562 cells . DENV2 D2S10 virus at an MOI of 0 . 1 and serum were mixed together in equal volumes for 30 to 45 minutes at 37°C . Modified MAbs were prepared in five 2-fold dilutions beginning at 2 , 000 ng/mL , added to the polyvalent serum/virus mixture , and incubated for an additional 30–45 minutes prior to the addition of 5×104 K562 cells . The cells were washed 2 hours following infection and fixed and stained for viral antigen . Relative infection was expressed as the average percent infection for each duplicate divided by the percent infection measured without modified antibody ( positive control , between 7 and 15% ) . Avidity ELISA: DENV2 D2S10 virus was isolated by ultra-centrifugation at 53 , 000× g for 2 hours at 4°C and resuspended in cold PBS with 20% FBS . Concentrated virus was diluted to 5×104 pfu in carbonate coating buffer , pH 9 . 6 , and 50 µl was added to each well of a 96-well flat-bottomed plate as described previously [30] . The plate was coated overnight at 4°C and washed thoroughly with PBS with 0 . 1% Tween-20 ( PBS-T ) prior to blocking ( 5% milk w/v in PBS-T ) for one hour . Both the non-modified and modified MAbs were diluted to 120 µg/mL in blocking buffer and titrated two-fold for a total of 12 serial dilutions . Each MAb dilution was added in duplicate to the coated plate for one hour . The plates were washed with PBS-T and incubated with an alkaline phosphatase ( AP ) -conjugated goat anti-human secondary antibody ( Meridian ) and AP substrate PNPP ( Sigma ) for one hour each , with additional PBS-T washes in between each step . The reaction was developed for 45 minutes , and the absorbance was read at 405 nm on a UV-plate reader ( Bio-Tek ) using KC Junior software . Competition ELISA: Ninety-six-well flat-bottomed plates were coated with DENV2 as described above . Mouse MAb 4G2 was diluted to 1 µg/mL and mixed with human mAb diluted to 10 , 1 and 0 . 1 µg/mL in a separate 96-well plate , and 100 µl of the mixture was added . After one hour , the plates were washed and incubated with goat anti-mouse Fcγ-specific biotinylated secondary ( Jackson ) followed by Streptavidin-AP ( Invitrogen ) and the plates were developed with PNPP as described above . All graphs were produced using GraphPad Prism 5 software ( La Jolla , CA ) . Statistical analysis was performed using Stata v10 ( College Station , TX ) and Prism 5 software . Comparison between NT50 titers of non-modified and modified MAb pairs was conducted using a Wilcoxon rank-sum analysis . Comparison of survival rates was conducted using a non-parametric log rank test . A Spearman rho ( ρ ) was calculated to assess correlations between modified MAb NT50 titer , avidity , and therapeutic efficacy ( 0–100% survival ) . A Kruskal-Wallis test was used to compare 4G2 binding across increasing concentrations of human MAb , and a Friedman's analysis ( matched pairs Kruskal-Wallis ) was conducted by combining all data for each MAb tested . A Wilcoxon rank-sum test was used to compare differences in the percent reduction of 4G2-enhanced D2S10 infection in K562 cells with different mixtures of modified MAb as well as in the enhancement-suppression assay using mouse DENV1-immune serum to compare relative infection between therapeutic MAbs and non-therapeutic MAbs . A sign rank test was used to determine whether 1 , 000 ng/mL of modified MAb could reduce an infection enhanced with DENV4-immune serum significantly lower than 50% .
|
The four dengue virus serotypes ( DENV1-4 ) cause the most prevalent mosquito-transmitted viral disease globally , infecting 50–100 million people annually in tropical and sub-tropical regions worldwide , yet no vaccine or therapy has been licensed to prevent or treat dengue . The greatest risk factor for severe dengue disease is a previous infection with a different serotype , which is thought to be due in part to a phenomenon known as antibody-dependent enhancement ( ADE ) whereby anti-DENV antibodies from a prior infection augment DENV infection of target Fcg receptor ( FcgR ) -expressing cells . We previously developed a mouse model that demonstrates antibody-enhanced lethal DENV disease and showed that genetically-modified antibodies incapable of interacting with the FcgR eliminate ADE in vitro and in vivo . In this study , we studied a larger panel of modified MAbs that recognize different regions of the DENV envelope protein . While all modified MAbs acted therapeutically to prevent a lethal , virus-only DENV infection , only certain MAbs effectively protected mice following an antibody-enhanced lethal infection . We determined that therapeutically effective MAbs following an ADE infection worked by competing for binding of enhancing antibodies on the DENV virion . Based on this , we designed an in vitro suppression-of-enhancement assay that predicted the ability of modified MAbs to act therapeutically in vivo .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"humoral",
"immunity",
"medicine",
"infectious",
"diseases",
"immunity",
"dengue",
"fever",
"neglected",
"tropical",
"diseases",
"immunology",
"biology",
"viral",
"diseases",
"dengue",
"immune",
"response"
] |
2013
|
Therapeutic Efficacy of Antibodies Lacking FcγR against Lethal Dengue Virus Infection Is Due to Neutralizing Potency and Blocking of Enhancing Antibodies
|
Yersinia pseudotuberculosis forms biofilms on Caenorhabditis elegans which block nematode feeding . This genetically amenable host-pathogen model has important implications for biofilm development on living , motile surfaces . Here we show that Y . pseudotuberculosis biofilm development on C . elegans is governed by N-acylhomoserine lactone ( AHL ) -mediated quorum sensing ( QS ) since ( i ) AHLs are produced in nematode associated biofilms and ( ii ) Y . pseudotuberculosis strains expressing an AHL-degrading enzyme or in which the AHL synthase ( ypsI and ytbI ) or response regulator ( ypsR and ytbR ) genes have been mutated , are attenuated . Although biofilm formation is also attenuated in Y . pseudotuberculosis strains carrying mutations in the QS-controlled motility regulator genes , flhDC and fliA , and the flagellin export gene , flhA , flagella are not required since fliC mutants form normal biofilms . However , in contrast to the parent and fliC mutant , Yop virulon proteins are up-regulated in flhDC , fliA and flhA mutants in a temperature and calcium independent manner . Similar observations were found for the Y . pseudotuberculosis QS mutants , indicating that the Yop virulon is repressed by QS via the master motility regulator , flhDC . By curing the pYV virulence plasmid from the ypsI/ytbI mutant , by growing YpIII under conditions permissive for type III needle formation but not Yop secretion and by mutating the type III secretion apparatus gene , yscJ , we show that biofilm formation can be restored in flhDC and ypsI/ytbI mutants . These data demonstrate that type III secretion blocks biofilm formation and is reciprocally regulated with motility via QS .
The human pathogenic Yersiniae ( Yersinia pseudotuberculosis , Yersinia enterocolitica and Yersinia pestis ) share a high degree of DNA identity , but cause distinct diseases ranging from enterocolitis ( Y . enterocolitica and Y . pseudotuberculosis ) to pneumonic , bubonic or septicaemic plague ( Y . pestis ) . Essential for the virulence of all pathogenic Yersiniae , is the ∼70-kb pYV virulence plasmid , which encodes the Yop virulon . This consists of a type III secretion system which enables Yersinia to inject multiple Yop effector proteins directly into the cytosol of eukaryotic cells and so subvert host cell signalling pathways ( for reviews see [1]–[3] . Yop virulon genes are tightly regulated by environmental conditions and in particular , temperature ( only expressing at 37°C ) and Ca2+ concentration ( reviewed in [4] ) . Y . pestis and Y . pseudotuberculosis are capable of forming biofilms around the anterior and along the surface of the nematode Caenorhabditis elegans [5] , [6] . However , biofilm formation is strain-dependent and a study of over 40 different Y . pseudotuberculosis strains showed that some formed biofilms on C . elegans but not on abiotic polystyrene surfaces and vice versa [6] . No relationship was observed between strains forming biofilms on C . elegans and those that formed biofilms on polystyrene surfaces . These findings suggest that biofilm development on the living surface of C . elegans is different from that on an abiotic surface such as polystyrene . Y . pestis is transferred between mammalian hosts by a flea borne vector that feeds on blood . The hmsHFRS operon is key to the colonisation and blockage of the flea proventriculus which results from the accumulation of biofilm [7]–[9] and hmsHFRS mutants of both Y . pestis and Y . pseudotuberculosis fail to form biofilms on C . elegans . Since C . elegans has been thoroughly studied at the genetic level and orthologous genes frequently studied in human health and disease , the C . elegans/Yersinia model can be used to identify genetic features of both the pathogen and the host that contribute to biofilm-mediated interactions between bacteria and invertebrates . These in turn have interesting implications for both the Yersinia/flea and human biofilm-centred infections . Although there are some limitations , the importance of C . elegans as a model organism for investigating prokaryotic/eukaryote interactions should not be overlooked given that nematodes are the most abundant animals on the Earth [10] . Although Y . pseudotuberculosis does not readily colonise fleas , biofilm formation may alternatively be involved in the prevention of predatory feeding as has been noted for other soil bacteria [11] . Whether the bacteria-invertebrate biofilm relationship is bacterially driven or is a two way interactive process between the bacteria and nematode is not fully understood . It has however been postulated that nematodes accumulate the bacterially derived extracellular matrix ( ECM ) passively by virtue of their movement through a lawn of bacteria [12] and there is evidence to show that biofilms do not accumulate on the surface of non-motile C . elegans . This implies that a prerequisite for biofilm formation is nematode translocation which provides the necessary contact between bacteria and nematode [12] . However , Y . pseudotuberculosis is unable to form biofilms on a number of motile C . elegans mutants such as srf-2 , srf-3 and srf-5 [6] and bah-1 , bah-2 and bah-3 . Conversely many natural strains of Y . pseudotuberculosis fail to form biofilms on C . elegans as do a number of Y . pseudotuberculosis strains with mutations in lipopolysaccharide biosynthesis , signal transduction and hms genes [6] . Such findings imply the existence of an adaptive interaction between the nematode and the bacterium rather than simply the passive adherence of bacterially derived ECM [6] . Bacteria possess multiple integrated sensory systems that govern adaptation to environmental challenges including the local cell population density . Such population-dependent adaptive behaviour often takes the form of perception and processing of chemical information and is termed quorum sensing ( QS ) . For many Gram negative bacteria this involves the use of self-generated diffusible signal molecules such as the N-acyl homoserine lactones ( AHLs ) . These are usually synthesised and sensed via members of the LuxI AHL synthase and LuxR response regulator protein families respectively . QS enables bacteria to determine , by monitoring the concentration of a signal molecule , when the number of individuals in the population are sufficient ( a quorum ) to make a collective ‘decision’ to alter their behaviour in response to environmental challenges [13]–[16] . Such behavioural decisions impact on bacterial motility , secondary metabolism , virulence , and biofilm development [17] . Y . pseudotuberculosis produces four major AHLs via a QS system consisting of two genetic loci termed ypsR/ypsI and ytbR/ytbI which control cell aggregation/flocculation and swimming motility [18] , [19] . This system is organized hierarchically with YpsR and its cognate AHLs regulating ytbR and ytbI as well as ypsR and ypsI . The YpsR/YpsI and YtbR/YtbI QS system in turn fine tunes swimming motility by governing the expression of two key regulators of the motility cascade , namely flhDC and fliA [19] . AHL-dependent QS also controls motility in Y . enterocolitica [20] Y . pestis produces a similar range of AHLs to Y . pseudotuberculosis [21] and retains an analogous QS system [22] , [23] . However the relationship between QS and regulators of the motility cascade such as flhDC or fliA may be different in Y . pestis when compared with Y . pseudotuberculosis or Y . enterocolitica because Y . pestis is non-motile because of a frame-shift mutation in the motility master regulator flhD [24] . There is considerable evidence to show that AHL-dependent QS plays a significant role during the biofilm mode of growth on an abiotic surface since AHL production has been detected in glass and metal surface associated biofilms produced by bacteria such as Pseudomonas aeruginosa [25] and Aeromonas hydrophila [26] . Furthermore , in a variety of bacteria , QS controls the target genes required for different stages of biofilm development from adherence and aggregation to maturation and dispersal ( for review see [27] ) . In addition QS determines the physiological response of biofilm communities to antimicrobial agents and host defences [28] , [29] . In the present paper we sought to determine whether biofilm formation by Y . pseudotuberculosis on a living motile surface i . e . on C . elegans is an interactive , QS-dependent process . The results obtained revealed that QS in Y . pseudotuberculosis reciprocally regulates the C . elegans biofilm phenotype with type III secretion via the major motility regulators flhDC and fliA . Consequently the induction of type III secretion attenuates biofilm formation on C . elegans which can be restored in a QS mutant either by curing the pYV virulence plasmid from the ypsI/ytbI mutant , by growing YpIII under conditions permissive for type III needle formation but not Yop secretion or by mutating the type III secretion apparatus gene , yscJ , a key component of the type III injectisome .
When C . elegans is infected with Y . pseudotuberculosis YpIII harboring the gfp-plasmid pSB2020 and examined by confocal microscopy , the bacterial microcolonies fluoresce green and are embedded in an ECM which fluoresces red ( yellow when both bacteria and matrix are combined ) ( Figure 1A ) when labelled with WGA-R consistent with the presence of bacterially generated N-acetyl-D-glucosamine [12] . An orthogonal image of Figure 1A showing the depth of the biofilm in x and y planes can be seen in Figure S1A . After 48 h incubation the biofilms on C . elegans became highly resistant to WGA-R labelling and only stained red on the outer surface while the inner mass remained green ( compare Figure 1A with Figure 1B ) . In common with bacterial biofilms formed on abiotic surfaces [30] , the Yersinia biofilm on C . elegans also contains extracellular DNA as revealed by DAPI staining ( Figure 2A ) . To determine qualitatively whether AHLs are produced in the biofilms which accumulate on the surface of C . elegans , the biofilm matrix from heavily infected nematodes grown in the presence of Y . pseudotuberculosis for 24 h was extracted into dichloromethane and the extracts analysed using the AHL bioreporter C . violaceum CV026 in a well plate overlay assay [31] . As negative controls , AHL extractions were also carried out on nematodes which had been grown on E . coli OP50 and from the cell pellet of an overnight Y . pseudotuberculosis culture . Culture supernatant from the latter served as a positive control . Figure 3A ( i ) shows a purple halo of violacein around the agar well which contained the concentrated nematode extract taken from worms infected with parent Y . pseudotuberculosis . A similar result was obtained for the positive control ( Figure 3A iv ) while no violacein was observed around the negative control wells . Taken together these data indicate that AHLs are produced by Y . pseudotuberculosis growing as biofilms on the surface of C . elegans . To confirm that AHLs are synthesised in situ in the biofilms , Y . pseudotuberculosis was transformed with the gfp-biosensor , pJBA89 which fluoresces green in the presence of AHLs [32] . When infected with Y . pseudotuberculosis pJBA89 the characteristic biofilms which form on the surface of C . elegans after 24 h show green fluorescent Y . pseudotuberculosis pJAB89 embedded in the red WGA-R labelled biofilm matrix ( Figure 3B ) which were indistinguishable from those presented Figure 1A . Since AHLs were detected in the biofilms formed on C . elegans , we used two approaches to determine whether QS was required for biofilm development on the nematode surface . Firstly , we exploited the lactonase , AiiA which hydrolyses the ester bond within the AHL homoserine lactone moiety generating the corresponding , inactive , N-acylhomoserine compound [33] . When aiiA is introduced into Y . pseudotuberculosis on the pSU18 derivative pSA236 , the AHLs produced are hydrolysed , so generating an AHL-negative phenotype [19] . By comparing the parent YpIII strain with YpIII transformed with either the pSU18 control vector or pSA236 , we evaluated the contribution of AHL-dependent QS to biofilm development . For these experiments , a biofilm severity incidence was calculated for the infected C . elegans population after 24 h incubation . Each nematode was assigned a score between 0 and 3 related to the severity of biofilm accumulation ( examples of scores 0 and 3 can be taken from the biofilms shown in Figure 1A and C; and scored of 1 and 2 from Figure S1 B and C respectively ) . These assays revealed that Y . pseudotuberculosis and Y . pseudotuberculosis pSU18 had biofilm severity indices of 77 . 3% and 62 . 0% respectively . When C . elegans infected Y . pseudotuberculosis pSU18 were compared to nematodes infected with Y . pseudotuberculosis pSA236 the biofilm severity incidence was reduced to 38 . 7% ( p = <0 . 05 and n = 3 respectively ) ( Figure 4A ) . Secondly we carried out C . elegans infection assays using Y . pseudotuberculosis YpIII QS mutants transformed with the constitutive gfp-plasmid , pSB2020 . These included an AHL negative mutant in which both AHL synthase genes ( ypsI and ytbI ) have been disrupted and a second double mutant in which the two QS response regulators , ypsR and ytbR have been disrupted [18] , [19] . When compared with the parent Y . pseudotuberculosis YpIII strain ( Figure 1A ) , biofilm development was severely delayed in the ypsI/ytbI double mutant formed little or no biofilm ( compare Figure 1A and 1C ) . Similar results were obtained for the ypsR/ytbR double mutant ( data not shown ) . In addition , nematodes grown on YpIII , in contrast to those grown on E . coli OP50 , exhibit exaggerated body bends ( Figure 5 ) , are unable translocate within 1 . 5 h and by 5 h become moribund . In contrast , C . elegans infected with either the ypsI/ytbI mutant or the ypsR/ytbR mutant translocate normally and make tracks in the agar which are identical to those presented in Figure 5A and only began to show signs of aberrant movement 3–4 h post infection . After 96 h growth , both the ypsI/ytbI and ypsR/ytbR mutants formed severe biofilms on the nematodes . In addition , we calculated a biofilm severity incidence for each yersinia strain . Figure 4B shows that after 24 h there is an ∼3 fold reduction in the amount of biofilm on nematodes infected by the ypsI/ytbI double mutant compared with the parent ( 32% compared with 89%; p = <0 . 01 n = 4 ) . Similar results were obtained for the ypsR/ytbR double mutant ( data not shown ) . Genetic complementation of the ypsI/ytbI mutation with pSA291 ( Figure 4C ) partially restored the biofilm severity incidence to that of the parent strain ( Parent pHG327 ( 82% ) compared with the ypsI/ytbI mutant pHG327 ( 35% ) ( p = 0 . 001 n = 3 ) and ypsI/ytbI mutant pSA291 ( 60% ) compared with ypsI/ytbI mutant pHG327 ( 35% ) ( p = <0 . 05 n = 3 ) ) . These data demonstrate that the loss of AHL synthesis either via enzyme-mediated inactivation or by mutagenesis of the AHL synthases results in the attenuation of biofilm formation on C . elegans . Consequently QS is pivotal to the timing and severity of biofilm development on C . elegans . Since the ypsR/ypsI and ytbR/ytbI loci are both involved in the regulation of motility via flhDC and fliA which code for the motility master regulator and flagellar specific sigma factor respectively [19] , we sought to determine whether these downstream regulators contribute to the Yersinia/C . elegans biofilm phenotype . Figure 6A shows that the flhDC mutant was impaired in its ability to form biofilms on the surface of C . elegans ( biofilm severity incidence for the parent of 57 . 5% compared to 24 . 7% for the flhDC mutant ( p = <0 . 05 , n = 3 ) ) and genetic complementation of flhDC using pSA220 increased the biofilm severity to 61 . 6% when compared with the flhDC mutant ( p = <0 . 02 , n = 3 ) . Figure 6B shows that the biofilm severity incidence for the fliA mutant was also reduced when compared with the parent ( p = <0 . 05 , n = 3 ) . Since both regulators control swimming motility and as flhDC and fliA mutants are non-motile , these data suggested that biofilm formation may depend on flagellar-mediated motility . To explore this possibility , we first constructed a flagellin-negative strain by mutating the flagellin structural gene , fliC . This non-motile mutant formed biofilms on nematodes which were indistinguishable from the parent Y . pseudotuberculosis strain ( Figure 6C and data not shown ) . Consequently , flagellar-mediated motility is not a necessary pre-requisite for biofilm formation on C . elegans . However , in Y . enterocolitica , the flagellar type III secretion apparatus may also secrete non-flagellar proteins termed ‘Fops’ ( for Flagellar outer proteins ) such as the phospholipase , YplA [34] . Since flagellin structural mutants still secrete Fops , we constructed a flhA mutant since this gene codes for a structural component of the flagellar protein export apparatus [35] and flhA mutants have been reported not to secrete Fops [34] . In common with the flhDC and fliA mutants and when compared to the parent , the flhA mutant exhibited attenuated biofilm formation ( Figure 6B ) ( p = <0 . 05 , n = 3 ) , a finding which implies a possible role for a secreted protein ( s ) . To determine whether any secreted proteins could be involved in biofilm development on C . elegans , we first examined the extracellular protein profiles of the Y . pseudotuberculosis flhDC , fliA , flhA and fliC mutants grown overnight in LBmops at 30°C . Figure 7 shows that compared with the parent strain and fliC mutant , numerous proteins are up-regulated in each of the other motility mutants . MALDI-TOF MS analysis identified three of the major protein bands as YopM/H ( 41/51 kDa , these two proteins often co-migrate and could not be distinguished by MALDITOF sequencing ) , LcrV ( 37 kDa ) and YopN ( 32 kDa ) all of which are encoded on the pYV virulence plasmid and secreted by the Ysc-Yop type III secretion system . Two further up-regulated proteins were identified as KatY and GroEL which are not related to the Yop virulon ( Figure 7 ) . In contrast to the YpIII parent strain which only secretes Yops at 37°C in the absence of Ca2+ both flhDC and fliA mutants clearly secrete Yops at 30°C in the presence of Ca2+ . Since both of these motility regulators are controlled by QS in Y . pseudotuberculosis [19] , these data suggested that elements of the Yop virulon are also likely to be QS-controlled . Figure 8 shows that when grown in LBMops at 30°C overnight , at least 4 extracellular proteins are up-regulated in the ypsI/ytbI and ypsR/ytbR double mutants compared with the parent strain . The same proteins are also up-regulated in the ypsR , ytbR and ytbI single mutants whereas the ypsI mutant exhibits the same profile as the parent strain . MALDI-TOF MS analysis identified the proteins as YopM/YopH , FliC , LcrV and YopN . These proteins were also present in supernatants from the same mutants after growth at 37°C in LBMops but absent from the parent and ypsI mutant ( Figure S2 ) . In contrast , Yop proteins were absent from the supernatants of all of the strains grown at 22°C although two proteins , the flagellar capping protein ( FliD; 48 . 6 KDa ) and flagellin ( FliC; 45 KDa ) were up-regulated ( data not shown ) . The attenuation of biofilm formation on C . elegans observed for both the motility and QS mutants in conjunction with the elevated secretion of Yop virulon proteins at non-permissive temperatures raised the possibility that induction of type III secretion blocks biofilm development . Consequently , we predicted that biofilm formation would be restored in Y . pseudotuberculosis ypsI/ytbI and flhDC mutants cured of the pYV plasmid . To explore this hypothesis , we cured the pYV plasmid from the parent , ypsI/ytbI and flhDC mutants by repeated selection on CRMOX agar plates . The presence or absence of the pYV plasmid had no effect on the ability of the Y . pseudotuberculosis YpIII parent strain to form a biofilm on C . elegans ( Figure 9A ) . However when similar experiments were performed using the ypsI/ytbI double mutant ( Figure 9A and compare with Figure 4B ) or flhDC ( data not shown ) cured of pYV , biofilm formation on C . elegans was restored to parental strain levels when compared with the biofilm levels observed on the ypsI/ytbI pYV+ double mutant ( p = <0 . 01 , n = 3 ) . These data suggest that under these conditions , AHL-mediated QS represses the expression of a pYV gene ( s ) which would otherwise prevent biofilm formation . To gain further evidence in support of a biofilm inhibitory role for pYV Yop virulon component ( s ) , C . elegans was infected with the parent Y . pseudotuberculosis grown in LBmops MOX , conditions which promote Yop secretion ( i . e . 37°C in the absence of Ca2+ ) rather than in LBmops at 30°C in which Yops will not be secreted . These seed cultures were then transferred to NGM plates containing MgCl2 and sodium oxalate to chelate Ca2+ . Under such pre-conditions , the type III system is induced and no biofilms were formed on C . elegans ( data not shown ) providing additional support that induction of the Yop virulon prevents biofilm formation on C . elegans . To demonstrate unequivocally that the inhibition of biofilm formation on C . elegans observed for the Y . pseudotuberculosis ypsI/ytbI mutant depends on the induction of functional type III secretion system rather than other genes present on the pYV plasmid , we modified the ypsI/ytbI mutant by mutating yscJ . This gene codes for a key component of the Ysc injectisome required for the assembly of a functional type III secretion apparatus [36] . Cell free culture supernatants taken from the ypsI/ytbI/yscJ triple mutant grown in LBMops at 30°C were examined by SDS-PAGE . This confirmed that , in contrast to the ypsI/ytbI mutant , Yop proteins were no longer secreted ( data not shown ) . Yop secretion in the triple mutant grown under these conditions could however be restored by complementation with a plasmid-borne copy of yscJ ( pHG::yscJ; data not shown ) . In the C . elegans biofilm assays , the biofilm severity index of the ypsI/ytbI/yscJ triple mutant was ∼4-fold higher than that of the ypsI/ytb double mutant ( p = <0 . 05 , n = 3 ) and comparable with that of the parent strain ( Figure 9B ) . When the triple mutant was compared to its complemented counterpart containing a functional copy of yscJ ( on plasmid pHG::yscJ ) biofilm severity was reduced ∼two-fold ( p = <0 . 01 , n = 3 ) back to levels comparable with the ypsI/ytbI double mutant ( Figure 9B ) . These results are consistent with a role for the type III injectisome in preventing biofilm development on C . elegans and demonstrate that either the type III needle or the secreted Yop proteins or both prevent biofilm development on C . elegans . To attempt to differentiate between these three possibilities , we grew the Y . pseudotuberculosis YpIII parent strain at 37°C in the presence of calcium which results in type III needle assembly but not Yop secretion [37] . This is because Ca2+ prevents Yop effector secretion even in the presence of a fully formed injectisome . YpIII was then subcultured onto NGM medium supplemented with calcium . When pre-cultured under these conditions and used to infect C . elegans at 22°C , Y . pseudotuberculosis YpIII failed to form a biofilm on C . elegans . The infected worms were indistinguishable from that shown in Figure 1C suggesting that the type III needle rather than the Yop effectors was responsible for preventing biofilm development .
On abiotic surfaces , bacterial biofilm formation is generally considered as a step-wise process initiating from individual cells adhering to a substratum leading to microcolony formation , biofilm maturation and finally dispersal to new sites [38]–[42] . Although the nature and development of biofilms formed on biotic surfaces have not been as thoroughly investigated , biofilm development by Y . pseudotuberculosis on C . elegans involves attachment and maturation stages and the ECM contains both carbohydrate and extracellular DNA . Whether the DNA present in the biofilm is bacterial or nematode-derived has yet to be established . However , the WGA-stained carbohydrate present in the ECM appears to be bacterially-derived since it is present in the lawns of Y . pseudotuberculosis prior to the addition of nematodes which are not labelled by WGA [43] . The WGA-stained ECM carbohydrate could be either peptidoglycan which contains N-acetyl glucosamine in the sugar backbone [44] or polymeric N-acetyl-D-glucosamine or both . Y . pestis strains with mutations in the hmsHFRS locus , which is responsible for the biosynthesis of a poly β-1 , 6-N-acetyl-D-glucosamine-like polysaccharide [45] , are defective for biofilm accumulation on C . elegans implying that this exopolysaccharide plays an essential role . An intact hmsHFRS is also required for biofilm formation on C . elegans by both Y . pseudotuberculosis and Xenorhabdus nematophila [46] . Apart from the hmsHFRS genes , other yersinia genes currently known to be required for biofilm formation on C . elegans include two genes involved in LPS biosynthesis , two genes of unknown function and a potential hybrid two component regulatory protein [6] . Both RcsA ( a phosphorelay accessory protein which functions in concert with the response regulator , RcsB ) and PhoP negatively regulate the formation of Y . pseudotuberculosis biofilms on nematodes [47] while the action of PhoP appears to be mediated at least in part by the down-regulation of HmsT [48] . This is interesting since HmsT is a cyclic diguanylate ( c-di-GMP ) synthase and c-di-GMP metabolism plays an important role in biofilm formation in many different bacteria including Y . pestis [49] , [50] . Depending on the organism , QS may be involved in the early attachment or later maturation stages of biofilm development on abiotic surfaces [27] . In pathogens such as P . aeruginosa , QS is responsible for controlling the expression of key components of the biofilm extracellular matrix including exopolysaccharides and extracellular DNA release as well as the refractory nature of biofilms to host defences and antimicrobials [27] . The contribution of QS to yersinia biofilm development on C . elegans has not previously been investigated although for Y . pseudotuberculosis , QS controls cell aggregation ( a type of suspended biofilm ) in liquid culture [18] . A Y . pestis strain with combined mutations in ypsR/ypsI , ytbR/ytbI and luxS formed a similar biofilm on glass cover slips to the parental strain which could not be distinguished by crystal violet or Congo red staining although a very mild defect was observed using confocal microscopy [51] . Here , for Y . pseudotuberculosis YpIII , we have shown that AHL-dependent QS is functional in biofilms formed on C . elegans by demonstrating ( i ) the presence of AHL signal molecules within the nematode-associated biofilm matrix and ( ii ) that YpIII strains in which AHL biosynthesis is abrogated either by expressing an AHL-inactivating enzyme in situ or by mutating the AHL synthases ( YpsI and YtbI ) are attenuated for biofilm formation . Because Y . pseudotuberculosis YpIII does not form biofilms on polystyrene surfaces [6] , these data indicate that the QS-dependent pathway for biofilm formation on C . elegans is different from that on abiotic surfaces . While QS signals have previously been identified in pseudomonas and aeromonas biofilms on abiotic surfaces [25] , [26] to our knowledge they have not previously been detected directly in biofilms growing on a living , biotic surface . AHLs have however been shown to be produced in the tissues of mice infected with Y . enterocolitica [52] although no evidence was presented for biofilm formation in this acute experimental infection model . In Y . pseudotuberculosis , YpsRI and YtbRI form a QS hierarchy in which ypsR is auto-regulated and also controls the expression of ypsI , ytbI and ytbR; YtbR also regulates ytbI expression [19] . In common with the ypsI/ytbI double synthase mutant , the ypsR/ytbR double response regulator mutant was also attenuated for biofilm development on C . elegans . The ypsR/ytbR mutant however produces a similar AHL profile to that of the parent strain [19] and therefore AHL production per se is not required for biofilm formation . The intermediate level biofilms formed by the single ypsR , ytbR and ytbI mutants ( data not shown ) reflect the interdependent nature of the Y . pseudotuberculosis QS system while the lack of biofilm attenuation observed for the ypsI mutant suggested that the AHLs synthesized via YtbI are primarily responsible for the biofilm phenotype observed . A number of Gram-negative bacterial species rely on flagellar-mediated motility for specific stages of biofilm formation [38] . For example , in E . coli , mutations which lead to either the loss of flagella or flagella function ( which include fliC or flhD ) are unable to form mature biofilms indicating that the presence of functional flagella is a pre-requisite for biofilm development in a PVC attachment model [53] . Similarly , non-motile yet flagellate P . aeruginosa PA01 flgK mutants and Erwinia carotovora fliC or motA mutants cannot form biofilms on PVC surfaces [54] , [55] . Furthermore , in Y . enterocolitica , mutations that abolish the structure or rotation of the flagellar greatly reduced biofilm formation in PVC microplate assays [56] . Thus , given the links between biofilm formation , flagella-mediated motility and the regulation of the two key motility regulators , flhDC and fliA by QS in Y . pseudotuberculosis [19] , we investigated the contribution of motility to biofilm formation on C . elegans . Surprisingly , a Y . pseudotuberculosis fliC mutant formed similar biofilms to the parent strain indicating that on the nematode , the presence of flagellar is not a pre-requisite for biofilm formation . This provides further evidence to suggest that the biofilm developmental pathway on the living nematode surface is distinct from that occurring on an abiotic surface . Since flagellins are potent inducers of the innate immune response and are often considered as flags revealing the presence of bacteria [57] , it may therefore be advantageous for Yersinia to repress their expression during growth on living surfaces . Despite the lack of biofilm attenuation for the fliC mutant , non-motile strains with mutations in flhA , a structural component of the flagellar export apparatus as well as the motility cascade regulators , flhDC and fliA were significantly attenuated . Since QS governs the expression of key motility regulators [19] these data suggested that biofilm formation on C . elegans by Y . pseudotuberculosis was linked to QS via the motility cascade . As Y . pestis has a frameshift mutation in flhD , biofilm formation on C . elegans in Y . pestis may well be governed differently to Y . pseudotuberculosis . Y . enterocolitica secretes FOP proteins such as YplA via the flagellar type III secretion apparatus [34] . Consequently , we considered it possible that the loss of Y . pseudotuberculosis FOP proteins by mutation of the motility genes may have been responsible for biofilm attenuation . However , SDS-PAGE analysis of the extracellular protein profile of these strains did not reveal any novel FOP proteins but rather the presence of several proteins associated with the Yop virulon and type III secretion . In particular , LcrV which is associated with the tip of the injectisome and with pore formation across the host cell membrane , YopN , a plug considered to limit Yop effector translocation through the needle and YopH , a phosphotyrosine phosphatase effector protein which inhibits phagocytosis ( reviewed by [2] ) . Our findings are consistent with observations made by [58] that deletion of flhDC resulted in the up-regulation of the yop regulon in Y . enterocolitica as a consequence of FlhDC-mediated repression of the Yop virulon regulator gene , virF . Since QS in Y . pseudotuberculosis regulates flhDC and fliA [19] we also examined cell free supernatants of strains with mutations in the ypsRI and ytbRI loci for the up-regulation of Yop virulon proteins . Apart from the single ypsI mutant , which exhibited the parental phenotype , each of the QS mutants exhibited the same protein profile on SDS-PAGE as the flhDC and fliA mutants when grown at 30°C in the presence of Ca2+ . Since both injectisome and Yop effector proteins were up-regulated , these data suggest that QS represses the Yop virulon via the actions of FlhDC on virF . In addition , it is clear that mutation of QS results in the loss of both the temperature and Ca2+ dependence characteristic of type III secretion in Yersinia . Thus in Y . pseudotuberculosis , QS positively regulates motility but negatively controls type III secretion indicating that both phenotypes are population dependent . This would suggest that in the planktonic phase at high population densities in the presence of eukaryotic target cells , Yop secretion would be shut down in favour of bacterial migration to new sites where a fall in QS signal concentrations would stimulate the resumption of Yop secretion . With respect to the biofilm phenotype of the QS and motility mutants , the de-repression of type III secretion at temperatures below 37°C suggested that type III secretion blocked biofilm formation on C . elegans . Since the Yop virulon genes are located entirely on the pYV plasmid , we examined the biofilm phenotype of the plasmid-cured parent , ypsI/ytbI and flhDC mutants respectively . The loss of pYV from the parent Y . pseudotuberculosis strain had no impact on biofilm formation an observation which is fully in agreement with Joshua et al . , ( 2003 ) [6] who examined both YpIII and a range of Y . pseudotuberculosis strains with or without the virulence plasmid . However the attenuation of biofilm formation observed for both the ypsI/ytbI and flhDC mutants could be overcome by curing pYV , a finding which implied that QS represses the expression of pYV encoded gene ( s ) which block biofilm formation in the presence of Ca2+ and at 22°C , the temperature at which the C . elegans assays are carried out . Additional support for these observations was obtained when seed cultures of the Y . pseudotuberculosis parent strain were grown under conditions permissive for Yop release ( 37°C in the absence of Ca2+ ) and then transferred onto Ca2+-free modified NGM plates at 22°C whereupon biofilms did not form on C . elegans . To rule out the possibility that other genes located on the pYV plasmid were responsible for the biofilm phenotype rather than the presence of a functional type III secretion system , we introduced a yscJ mutation into the ypsI/ytbI double mutant . The newly generated triple mutant resulted in the loss of type III secretion at 30°C in the presence of Ca2+ and the restoration of biofilm formation on C . elegans . This strongly implies that the presence of an intact injectisome blocks biofilm formation on C . elegans . However , these data alone could not determine whether the reduction in biofilm was due to the presence of an intact injectisome , extracellular Yops or both . Evidence to suggest that the type III injectisome rather than the Yop effectors were responsible for attenuating biofilm formation on C . elegans was obtained by first conditioning seed cultures of Y . pseudotuberculosis at 37°C in Ca2+ containing media prior to carrying out biofilm assays . We reasoned that the conditioned Y . pseudotuberculosis cells would possess intact injectisomes but would not release Yops [59]–[63] . Furthermore , the presence of Ca2+ in the NGM agar would continue to suppress Yop secretion during the biofilm assays . When biofilm assays were performed using pre-conditioned Y . pseudotuberculosis cells biofilm formation was suppressed . These data appear to preclude a requirement for extracellular Yops in order for biofilm formation to take place . The simplest explanation is that the presence of the fully formed needle acts as a physical barrier which blocks the interaction between a key , chromosomally encoded bacterial surface component and the nematode surface . This would also be consistent with the loss of biofilm formation which results from the mutation of a number of C . elegans surface-determining genes [64] , [65] . However , at this stage we cannot rule out the possibility that contact between the injectisome and C . elegans results in the repression of as yet unidentified genes required for biofilm formation .
The Y . pseudotuberculosis , Escherichia coli and C . elegans strains and the plasmids used in this study are listed in Table S1 and Table S2 respectively . To aid visualisation of Y . pseudotuberculosis in biofilm assays , the bacterial cells were transformed with pSB2020 [66] which constitutively expresses gfp3 . To determine whether biofilm formation on C . elegans could be attenuated by AHL hydrolysis , Y . pseudotuberculosis YpIII was also transformed with the lactonase gene , aiiA on pSA236 as described before [19] . Except where stated , bacterial cultures were routinely grown with shaking at 200 rpm in L broth Lennox [67] or on agar plates containing the appropriate antibiotics buffered to pH 6 . 8 with Mops ( 3-N-morpholino ) propanesulphonic acid ( YLBmops ) to reduce alkaline hydrolysis of AHLs during bacterial growth [68] . To promote yop expression at 37°C some experiments were performed in YLBmops supplemented with MgCl2 ( 20 mM ) and sodium oxalate ( 20 mM ) as previously described [58] . Where required , pYV was cured from Y . pseudotuberculosis by the repeated sub-culture of white colonies onto Congo red-magnesium oxalate ( CRMOX ) plates [69] . The C . elegans wild-type ( N2 Bristol ) strain was obtained from the Caenorhabditis Genetics Centre ( University of Minnesota , St . Paul , MN ) and maintained on modified NGM plates [70] lacking MgCl2 , seeded with E . coli OP50 unless otherwise stated . For Yop induction assays NGM was supplemented with MgCl2 ( 20 mM ) and sodium oxalate ( 20 mM ) but CaCl2 was omitted . NGM plates were seeded with 1 ml of the appropriate Y . pseudotuberculosis strain grown overnight at 30°C unless otherwise stated . For some C . elegans biofilm experiments , NGM agar plates were modified by the addition of sodium oxalate ( 20 mM ) and MgCl2 ( 20 mM ) to promote Yop secretion . For the assays in which biofilm severity incidence was calculated , Y . pseudotuberculosis were spread evenly over the agar surface , dried to remove excess liquid and 20–30 young adult C . elegans were aseptically transferred to the seeded plates . After incubation for 22°C for 24 h ( unless otherwise stated ) , the worms were examined under low magnification using a Nikon SMZ1000 microscope and biofilm accumulation was classed as level 0 if no biofilm formed ( e . g . Figure 1C ) ; level 1 indicating a small accumulation of biofilm around the anterior end of the worm ( e . g . Figure S1B ) ; level 2 denoted larger accumulations of biofilm around the anterior end of the worm with some pockets of biofilm spreading back from the head ( e . g . Figure S1C ) ; level 3 by large accumulations of biofilm around the anterior end of the worm which extended to other parts of the nematode body surface ( e . g . Figures 1A and 2B ) . Confocal images of C . elegans were taken using a Zeiss LSM700 inverted microscope . Replicate Z-stacks were taken at 5 µm intervals . The Zeiss Zen software package was used for image analysis . The level of biofilm accumulation on C . elegans was denoted as the biofilm severity incidence and was calculated according to the method of Tarr [71]: Biofilm severity incidence = {[∑ ( level X number of samples in this level ) ]/ ( highest level X total sample numbers ) } X 100% . All assays in which the level of biofilm severity was assessed were carried out double blind , with at least three or four replicates and each experiment was performed more than once . The error bars shown on figures 4 , 6 and 9 represent the standard deviation from the mean and when necessary independent two-sample t-tests were performed with values for p and n given in the text and on histograms where appropriate . For some experiments the presence of the N-acetyl-D-glucosamine in the ECM of Y . pseudotuberculosis biofilms was demonstrated using a wheat germ agglutinin ( WGA ) -rhodamine ( WGA-R ) conjugate as described by [12] . Extracellular DNA present in the biofilms was stained with DAPI following the method of Vilain et al . , [72] in which low concentrations of DAPI are demonstrated to label the extracellular biofilm matrix without penetrating the bacterial cell and staining the intracellular DNA . To determine whether biofilm formation was attenuated when worms were infected with Y . pseudotuberculosis containing the AHL lactonase AiiA , aiiA was excised from pSA302 [19] as an EcoRI fragment and then sub-cloned into the chloramphenicol resistant vector pSU18 [73] to give pSA236 which was transformed into Y . pseudotuberculosis . pSU18 was transformed into Y . pseudotuberculosis to act as a vector control . Plasmids were isolated using the Promega Wizard system , agarose gel electrophoresis and standard methods for the preparation of competent cells , DNA ligation and electroporation were performed as previously described [20] . For the purification of DNA fragments from agarose gels , Qiaquick DNA purification columns were used ( Qiagen Ltd ) . Restriction endonucleases , DNA ligase and other DNA modification enzymes were used according to the manufacturers' instructions ( Promega ) . C . elegans infected with Y . pseudotuberculosis were removed from 40 NGM plates in M9 wash solution [74] . The worms were washed , the pellet extracted into dichloromethane , reconstituted into 20 µl of acetonitrile and analysed using a well plate overlay assay using the C . violaceum CV026 biosensor which reports the presence of AHLs by producing the purple pigment violacein [31] . To detect AHLs produced in situ in biofilms on the surface of C . elegans , Y . pseudotuberculosis and the isogenic ypsI/ytbI double mutant were each transformed with the AHL biosensor , pJBA89 [32] which expresses gfp in the presence of AHLs . Infected worms were examined using fluorescent microscopy for the presence of green fluorescent bacteria within the biofilm matrix . Y . pseudotuberculosis YpIII strains with deletions in fliA , flhA , fliC and yscJ were constructed as follows . The fliA mutant was constructed using a modified method of [75] . The primers used for mutant construction are listed in Table S3 . Briefly , primer pairs fliA1up-F/fliA1up-R and fliA1down-F/fliA1down-R were used to amplify 510 and 511bp fragments of the up- and downstream regions of fliA ( positions 2069236 to 2069746 and 2070363 to 2070874 in the published Y . pseudotuberculosis IP 32953 genome sequence [76] ) . Primer fliA1up-R and fliAdown-F also contained 25 and 22 bp respectively of sequence homologous to the first 25 bp and last 22 bp of kanamycin from pUC4K [77] . The kanamycin cassette was amplified from pUC4K ( Pharmacia ) using primer km-F and km-R under the following PCR conditions: 95°C for 5 min followed by 30 cycles of 95°C for 30 s , 56°C for 30 s and 74°C for 1 min and ending with 74°C for 5 min . The second and third step PCR conditions were as follows: 95°C for 5 min followed by 30 cycles of 95°C for 30 s , 60°C for 30 s and 74°C for 2 min and ending with 74°C for 5 min . The strategy for constructing the flhA and yscJ mutants was similar to that of fliA . For flhA , primer pairs flhA1up-F/flhA1up-R and flhA1down-F/flhA1down-R were used to amplify the up- and downstream regions of flhA ( positions 2017164 to 2017699 and 2019587 to 2020179 on the published IP 32953 Y . pseudotuberculosis genome sequence ) whereas for yscJ , primer pairs YscJaFor/YscJupR-Tet and YscJdownF-tet/YscJbRev were used to amplify the up- and downstream regions of yscJ ( positions 59172 to 59743 and 60344 to 61135 ) on the published Y . pseudotuberculosis IP 32953 pYV virulence plasmid sequence . For flhA , primer flhA1up-R and flhA1down-F each contained 19 bp of sequence homologous to the first 19 bp or last 19 bp of the kanamycin cassette from pUC4K whereas for yscJ , YscJupR-Tet and YscJdownF-tet contained 21 bp or 22 bp of sequence homologous to a tetracycline cassette which was amplified as a 1191 bp product from pBlue-tet ( a source of the tetracycline cassette initially amplified from pBR322 using primers Tet1 and Tet2 [19] and cloned into pBluescript as an xhoI fragment ) . All PCR conditions were the same as those for the construction of the fliA mutant . To complement yscJ , primers YscJF-XbaI and YscJR-SalI were used to amplify an 842 bp product from Y . pseudotuberculosis ( positions 59686 to 59703 on the IP32953 published sequence ) which , after cloning into pBluescript and sequencing was excised as a KpnI and PstI fragment and sub-cloned into the low copy number vector pHG327 [78] . The resulting plasmid , pHG::yscJ was transformed into the Y . pseudotuberculosis ypsI/ytbI double mutant . Colony PCR was used to amplify a fliC homologue from Y . pseudotuberculosis using the primers DC1 and DC2 and cloned into pGEMT/easy ( Promega ) to give pfliC . Sequencing revealed the 1 , 515 bp fragment to have an open reading frame of 1 , 110 bp and predicted protein product of 396 amino acids that shared significant amino acid similarity to several FliC homologues and was subsequently termed fliCYp ( Genbank accession number AY244555 ) . To construct a fliC mutant 616 bp was removed from pfliC using Csp45I and replaced with a kanamycin cassette from pUC4K ( Pharmacia ) as a blunt end fragment . The resulting construct was cloned into pDM4 as a SphI-SpeI fragment ( pDM fliC-Km ) and stably integrated into the chromosome of Y . pseudotuberculosis as previously described [18] , [19] . To complement the Y . pseudotuberculosis YpIII flhDC mutant [19] flhDC was amplified by PCR ( primers FlhDF and FlhCR ) , cloned into pGEMT/Easy ( Promega ) and the resulting pGEM::flhDC construct , pSA220 was transformed into the Y . pseudotuberculosis flhDC mutant . The flhDC , flhA , fliA and fliC mutants were examined for motility using swim agar plate assays and microscopy and the presence of flagella proteins was determined by SDS-PAGE once isolated from 24 h overnight liquid cultures grown at 22°C as previously described [20] , [79] . Proteins present in 10 ml of cell-free supernatant taken from Y . pseudotuberculosis QS and motility mutants grown to the same OD600 ( overnight in YLB at 22°C , 28°C and 37°C ) were concentrated by trichloroacetic acid precipitation , subjected to SDS-PAGE and the relevant bands excised . After in-gel tryptic digestion , the resulting peptides were identified by matrix-assisted laser desorption ionization-time of flight ( MALDI-TOF ) -MS sequencing as previously described [20] .
|
Many Gram-negative bacteria communicate by producing and sensing the presence of chemical signal molecules such as the N-acylhomoserine lactones ( AHLs ) . Bacterial cells use AHLs to convey information about their environment , metabolism and population size . This type of chemical signalling is called ‘quorum sensing’ ( QS ) and is often used by pathogenic bacteria to promote acute or chronic infections through the control of motility , toxins , tissue degrading enzymes and surface-associated biofilms . Yersinia pseudotuberculosis is a human pathogen which forms biofilms on the surface of the nematode worm , Caenorhabditis elegans . This offers a simple means for investigating biofilm development on living tissues and can be used to identify genetic features of both the pathogen and the host that contribute to biofilm-associated infections . We have discovered that quorum sensing is required for Y . pseudotuberculosis biofilm formation on C . elegans through a regulatory pathway which involves the master motility regulator protein ( FlhDC ) reciprocally controlling bacterial swimming and the construction of a specialized secretion needle that delivers proteins into mammalian cells to disrupt their normal activities .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"biology",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"microbiology"
] |
2011
|
Biofilm Development on Caenorhabditis elegans by Yersinia Is Facilitated by Quorum Sensing-Dependent Repression of Type III Secretion
|
The prediction of phenotypic traits using high-density genomic data has many applications such as the selection of plants and animals of commercial interest; and it is expected to play an increasing role in medical diagnostics . Statistical models used for this task are usually tested using cross-validation , which implicitly assumes that new individuals ( whose phenotypes we would like to predict ) originate from the same population the genomic prediction model is trained on . In this paper we propose an approach based on clustering and resampling to investigate the effect of increasing genetic distance between training and target populations when predicting quantitative traits . This is important for plant and animal genetics , where genomic selection programs rely on the precision of predictions in future rounds of breeding . Therefore , estimating how quickly predictive accuracy decays is important in deciding which training population to use and how often the model has to be recalibrated . We find that the correlation between true and predicted values decays approximately linearly with respect to either FST or mean kinship between the training and the target populations . We illustrate this relationship using simulations and a collection of data sets from mice , wheat and human genetics .
Predicting unobserved phenotypes using high-density SNP or sequence data is the foundation of many applications in medical diagnostics [1–3] , plant [4 , 5] and animal [6] breeding . The accuracy of genomic predictions will depend on a number of factors: relatedness among genotyped individuals [7 , 8]; the density of the markers [7 , 9 , 10]; and the genetic architecture of the trait , in particular the allele frequencies of causal variants [11 , 12] and the distribution of their effect sizes [7] . Most of these issues have been explored in the literature , and have been tackled in various ways either from a methodological perspective or by producing larger data sets and more accurate phenotyping . However , the extent to which predictive models generalise from the populations used to train them to distantly related target populations appears not to have been widely investigated ( two exceptions are [7 , 13] ) . The accuracy of prediction models is often evaluated in a general setting using cross-validation with random splits , which implicitly assumes that test individuals are drawn from the same population as the training sample; in that case accuracy to predict phenotypes is only bounded by heritability , although unaccounted “missing heritability” is common [14 , 15] . However , this assumption is violated in many practical applications , such as genomic selection , that require predictions of individuals that are genetically distinct from the training sample: for instance , causal variants may differ in both frequency and effect size between different ancestry groups ( in humans , e . g . [16] for lactose persistence ) , subspecies ( in plants and animals , e . g . [17] for rice ) or even families [18] . In such cases cross-validation with random splits may overestimate predictive accuracy due to the mismatch between model validation and the prediction problem of interest [19 , 20] even when population structure is taken into account [21] . The more distantly the target population is related to the training population , the lower the average predictive accuracy of a genomic model; this has been demonstrated on both simulated and real dairy cattle data [20 , 22 , 23] . In this paper we will investigate the relationship between genetic distance and predictive accuracy in the prediction of quantitative traits . We will simulate training and target samples with varying genetic distances by splitting the training population into a sequence of pairs of subsets with increasing genetic differentiation . We will measure predictive accuracy with Pearson’s correlation , which we will estimate by performing genomic prediction from one subset to the other in each pair . Among various measures of relatedness available in the literature , we will consider mean kinship and FST , although we will only focus on the latter . We will then study the mean Pearson’s correlation as a function of genetic distance , which we will refer to as the “decay curve” of the former over the latter . This approach is valuable in addressing several key questions in the implementation of genomic selection programs , such as: How often ( e . g . , in terms of future generations ) will the genomic prediction model have to be re-estimated to maintain a minimum required accuracy in the predictions of the phenotypes ? How should we structure our training population to maximise that accuracy ? Which new , distantly related individuals would be beneficial to introduce in a selection program for the purpose of maintaining a sufficient level of genetic variability ?
A baseline model for genomic prediction of quantitative traits is the genomic BLUP ( GBLUP; [24 , 25] ) , which is usually written as y = μ + Zg + ε with g ∼ N ( 0 , K σ g 2 ) and ε ∼ N ( 0 , σ ε 2 ) , ( 1 ) where g is a vector of genetic random effects , Z is a design matrix that can be used to indicate the same genotype exposed to different environments , K is a kinship matrix and ε is the error term . Many of its properties are available in closed form thanks to its simple definition and normality assumptions , including closed form expressions of and upper bounds on predictive accuracy that take into account possible model misspecification [15] . Other common choices are additive linear regression models of the form y = μ + X β + ε ( 2 ) where y is the trait of interest; X are the markers ( such as SNP allele counts coded as 0 , 1 and 2 with 1 the heterozygote ) ; β are the marker effects; and ε are independent , normally-distributed errors with variance σ ε 2 . Depending on the choice of the prior distribution for β , we can obtain different models from the literature such as BayesA and BayesB [25] , ridge regression [26] , the LASSO [27] or the elastic net [28] . The model in Eq ( 1 ) is equivalent to that in Eq ( 2 ) if the kinship matrix K is computed from the markers X and has the form X XT and β ∼ N ( 0 , VAR ( β ) ) [29 , 30] . In the remainder of the paper we will focus on the elastic net , which we have found to outperform other predictive models on real-world data [31] . This has been recently confirmed in [32] . Predictive accuracy is often measured by the Pearson correlation ( ρ ^ ) between the predicted and observed phenotypes . When we use the fitted values from the training population as the predicted phenotypes , and assuming that the model is correctly specified , ρ ^ 2 coincides with the proportion of genetic variance of the trait explained by the model and therefore ρ ^ 2 ⩽ h 2 , the heritability of the trait . ( An incorrect model may lead to overfitting , and in that case ρ ^ 2 ⩾ h 2 . ) When using cross-validation with random splits , ρ ^ CV ⩽ ρ ^ and typically the difference will be noticeable ( ρ ^ CV ≪ ρ ^ ) . However , ρ ^ C V may still overestimate the actual predictive accuracy ρ ^ D in practical applications where target individuals for prediction are more different from the training population than the test samples generated using cross-validation [14] . This problem may be addressed by the use of alternative model validation schemes that mirror more closely the prediction task of interest; for instance , by simulating progeny of the training population to assess predictive accuracy for a genomic selection program . This approach is known as forward prediction and is common in animal breeding [19 , 33] . Another possible choice is the prediction error variance ( PEV ) . It is commonly used in conjunction with GBLUP because , for that model , it can be estimated ( for small samples ) or approximated ( for large samples ) in closed form from Henderson’s mixed model equations [34] . In the general case no closed form estimate is available , but PEV can still be derived from Pearson’s correlation [35] for any kind of model as both carry the same information: PEV = ( 1 - ρ ^ 2 ) * VAR ( y ) . ( 3 ) For consistency with our previous work [31] and with [4] , whose results we partially replicate below , we will only consider predictive correlation in the following . A common measure of kinship from marker data is average allelic correlation [24 , 36] , which is defined as K = [kij] with k i j = 1 m ∑ k = 1 m X ˜ i k X ˜ j k ( 4 ) where X ˜ i k and X ˜ j k are the standardised allele counts for the ith and jth individuals and the kth marker . An important property of allelic correlation is that it is inversely proportional to the Euclidean distance between the marker profiles Xi , Xj of the corresponding individuals: if the markers are standardised 2 n - 2 k i j = 2 n - 2 ∑ k = 1 m X ˜ i k X ˜ j k = ∑ k = 1 m X ˜ i k 2 + X ˜ j k 2 - 2 X ˜ i k X ˜ j k = ∑ k = 1 m ( X ˜ i k - X ˜ j k ) 2 . ( 5 ) This result has been used in conjunction with clustering methods such as k-means or partitioning around medoids ( PAM; [37] ) to produce subsets of minimally related individuals from a given sample by maximising the Euclidean distance [14 , 19 , 38] . At the population level , the divergence between two populations due to drift , environmental adaptation , or artificial selection is commonly measured with FST . Several estimators are available in the literature , and reviewed in [39] . In this paper we will adopt the estimator from [40] , which is obtained by maximising the Beta-Binomial likelihood of the allele frequencies as a function of FST . F ^ ST then describes how far the target population has diverged from the training population , which translates to “how far” a genomic prediction model will be required to predict . In terms of kinship , we know from the literature that the mean kinship coefficient k ¯ between two individuals in different populations is inversely related to F ^ ST [41]: kinship can be interpreted as the probability that two alleles are identical by descent , which is inversely related to FST which is a mean inbreeding coefficient . Intuitively , the fact that individuals in the two populations are closely related implies that the latter have not diverged much from the former: if k ¯ is large , the marker profiles ( and therefore the corresponding allele frequencies ) will on average be similar . As a result , any clustering method that uses the Euclidean distance to partition a population into subsets will maximise their FST by minimising k ¯ . The simulations and data analyses below confirm experimentally that k ¯ and F ^ ST are highly correlated , which makes them equivalent in building the decay curves; thus we will report results only for F ^ ST ( see Section C in S1 Text ) . We evaluate our approach to construct decay curves for predictive accuracy using two publicly-available real-world data sets with continuous phenotypic traits , and a third , human , genotype data set . We estimate a decay curve of ρ ^ D as a function of FST as follows: The pair of subsets produced by k-means corresponds to m = 0 , hence the notation ( F ^ ST ( 0 ) , ρ ^ D ( 0 ) ) , and we increase m by steps of 2 to 20 until the F ^ ST between the subsamples is at most 0 . 005 . We choose the stepping for each data set to be sufficiently small to cover the interval [ 0 , F ^ ST ( 0 ) ] as uniformly as possible . The larger m is , the smaller we can expect F ^ ST ( m ) to be . We repeat step 3 ( a ) and 3 ( b ) 40 times for each m to achieve the precision needed for an acceptably smooth curve . As an alternative approach , we also consider estimating the decay rate of ρ ^ D by linear regression of the ρ ^ D ( m ) against the F ^ ST ( m ) ; we will denote the resulting predictive accuracy estimates with ρ ^ L . For any set value of F ^ ST , we compare the ρ ^ L at that F ^ ST with the corresponding value ρ ^ D from the decay curve estimated by averaging all the ρ ^ D ( m ) for which | F ^ ST ( m ) - F ^ ST | ⩽ 0 . 01 . Assuming that the decay curve is in fact a straight line reduces the number of subsamples that we need to generate , enforces smoothness and makes it possible to compute ρ ^ L for values of FST larger than F ^ ST ( 0 ) . On the other hand , the estimated ρ ^ L will be increasingly unreliable as ρ ^ L → 0 , because the regression line will provide negative ρ ^ L instead of converging asymptotically to zero . We also regress the ( ρ ^ D ( m ) ) 2 against the ( F ^ ST ( m ) ) 2 to investigate whether they have a stronger linear relationship than the ρ ^ D ( m ) with the F ^ ST ( m ) , as suggested in [22] using simulated genotypes and phenotypes mimicking a dairy cattle population . The size of the training ( nTR ) and target ( nTA ) subsamples is determined by k-means . For the data used in this paper , k-means splits the training populations in two subsamples of comparable size; but we may require a smaller nTA ≪ nTR to estimate ρ ^ D ( 0 ) and the ρ ^ D ( m ) while at the same time a larger nTR is needed to fit the genomic prediction model . In that case , we increase nTR by moving individuals from the target subsample while keeping the F ^ ST ( 0 ) between the two as large as possible . The impact on the estimated F ^ ST is likely to be small , because its precision depends more on the number of markers than on nTR and nTA [40] . The estimated ρ ^ D 0 and ρ ^ D ( m ) might be inflated because we are altering the subsets , even when F ^ ST does not change appreciably . Its variance , which can be approximated as in [49] , decreases linearly in nTA except that can be compensated by generating more pairs of subsamples for each value of m . We study the behaviour of the decay curves via two simulation studies . Finally , we estimate the decay curves for some of the phenotypes available in the WHEAT and MICE data . For both data sets we also produce and average 40 values of ρ ^ CV using hold-out cross-validation . In hold-out cross-validation we repeatedly split the data at random into training and target subsamples whose sizes are fixed to be the same as those arising from clustering in step 1 of the decay curve estimation . Then we fit an elastic net model on the training subsamples and predict the phenotypes in the target subsamples to estimates ρ ^ CV . Ideally , the decay curve should cross the area in which the ( F ^ ST , ρ ^ CV ) points cluster .
The decay curves from the simulations are shown in Figs 1 , 2 and 3 , and the corresponding predictive correlations are reported in Tables 1 and 2 and S1 Text . The predictive correlations for the WHEAT and MICE data sets are reported in Table 2 , and the decay curves are shown in Figs 1 , 2 and 3 and S1 Text . A summary of the different predictive correlations defined in the Methods and discussed here is provided in Table 1 . In all the simulations and the real-world data analyses the ρ ^ D from the decay curve is close to the linear interpolation ρ ^ L; considering all the reference populations in Table 2 and the generation means in Tables A . 1 and A . 2 in S1 Text , | ρ ^ D - ρ ^ L | ≪ 0 . 02 41 times out of 47 ( 87% ) . Both estimates of predictive correlation are close to the respective reference values ρ ¯ and ρ ^ P; the difference ( in absolute value ) is ≪ 0 . 05 39 times ( 41% ) and ≪ 0 . 10 69 times ( 73% ) out of 94 . The proportion of small differences increases when considering only target populations that fall within the span of the decay curve: 23 out of 44 ( 52% ) are ≪ 0 . 05 and 38 are ≪ 0 . 10 ( 84% ) . This is expected because the decay curve is already an extrapolation from the training population , so extending it further with the linear interpolation ρ ^ L reduces its precision . Regressing ( ρ ^ D ( m ) ) 2 against the ( F ^ ST ( m ) ) 2 does not produce a stronger linear relationship than that represented by ρ ^ L ( p = 0 . 784 , see Section D in S1 Text ) . The range of the predictive correlations ρ ^ D ( m ) around the decay curves varies between 0 . 05 and 0 . 10 , and it is constant over the range of observed F ^ ST for each curve . It does not appear to be related to either the size of the training subsample or the number of causal variants . This is apparent in particular from the genomic selection simulation , in which both are jointly set to different combinations of values . Similarly , there seems to be no relationship between the spread and the magnitude of the predictive correlations ( ρ ^ D ( m ) ∈ [ 0 , 0 . 75 ] ) . This amount of variability is comparable to that of other studies ( e . g . , the range of the ρ ^ D ( m ) is smaller than that in the cross-validated correlations in [32] ) once we take into account that the ( F ^ ST ( m ) , ρ ^ D ( m ) ) are individual predictions and are not averaged over multiple repetitions . Furthermore , subsampling further reduces the size of the training subpopulations; and fitting the elastic net requires a search over a grid of values for its two tuning parameters , which may get stuck in local optima . Several interesting points arise from the analysis of the real phenotypes in the WHEAT and MICE data , shown in Table 2 and in Figs B . 1 , B . 2 and B . 3 in S1 Text . Firstly , cross-validation always produces pairs of subsamples with F ^ ST ⩽ 0 . 01 and high ρ ^ CV that are located at the left end of the decay curve . The average F ^ ST is 0 . 006 for the WHEAT data and 0 . 001 for the MICE data , and the difference between the average ρ ^ CV and the corresponding ρ ^ D is ≪ 0 . 02 10 times out of 12 ( 83% , see Table B . 4 in S1 Text ) . The spread of the ρ ^ CV is also similar to that of the ρ ^ D ( m ) . Secondly , we note that in the WHEAT data all decay curves but that for flowering time cross the 95% confidence intervals for the cross-country predictive correlations ρ ^ P for Germany and UK reported in [4] . Even in the MICE data , in which all families are near the end or beyond the reach of the decay curves , the latter ( or their linear approximations ) cross the 95% confidence intervals for the ρ ^ P 18 times out of 24 ( 75% ) . However , we also note that those intervals are wide due to the limited sizes of those populations . Furthermore , the decay curves for the phenotypes in the WHEAT data confirm two additional considerations originally made in [4] . Firstly , [4] noted that the distribution of the Ppd-D1a gene , which is a major driver of this flowering time , varies substantially with the country of registration and thus cross-country predictions are not reliable . Fig B . 1 in S1 Text shows that the decay curve vastly overestimates the predictive correlation for both Germany and the UK . Splitting the WHEAT data in two halves that contain equal proportions of both alleles of Ppd-D1a and that are genetically closer overall ( F ^ ST = 0 . 04 ) , we obtain a decay curve that fits the predictive correlations reported in the original paper ( ρ ^ D = 0 . 77 , ρ ^ P = 0 . 79 ) . Secondly , we also split the data according to their year of registration and use the oldest varieties ( pre-1990 ) as a training sample for predicting yield . Again the decay curve crosses the 95% confidence intervals for the predictive correlations reported in [4] and the correlations themselves are within 0 . 05 of the average ρ ^ D from the decay curve both for 1990-1999 ( F ^ ST = 0 . 028 , ρ ^ D = 0 . 44 , ρ ^ P = 0 . 40 ) and post-2000 ( F ^ ST = 0 . 033 , ρ ^ D = 0 . 44 , ρ ^ P = 0 . 42 ) varieties . The decay curves from the genomic selection simulation on the original training population ( 200 varieties ) , shown in blue in Fig 1 , span two rounds of selection and three generations . When considering 200 or 1000 causal variants , the curve overlaps the mean behaviour of the simulated data points ( shown in green ) almost perfectly: the difference between the generation means ρ ¯ and the decay curve is ⩽ 0 . 06 for the first three generations , with the exception of the first generation in the simulation with 1000 variants ( | ρ ¯ - ρ ^ D | = 0 . 09 ) . As the number of causal variants decreases ( 50 , 10 ) , the decay curve increasingly overestimates ρ ¯ , although the difference remains ⩽0 . 10 for the first two generations; and both show a slower decay than the ρ ¯ . This appears to be due to a few alleles of large effect becoming fixed by the selection , leading to a rapid decrease of ρ ¯ without a corresponding rapid increase in F ^ ST . The decay curves fitted on the augmented training populations ( 800 varieties , now including those available at the end of the second round of selection , Fig 2 ) fit the first four generations well ( | ρ ¯ - ρ ^ D | ⩽ 0 . 04 for the first two , | ρ ¯ - ρ ^ D | ⩽ 0 . 06 for the third and the fourth ) . As before , the only exception is the first generation in the simulation with 1000 variants , with an absolute difference of 0 . 09 . However , the decay curves are also able to capture the long-range decay rates through their linear approximations . When considering 200 causal variants , | ρ ¯ - ρ ^ L | ≈ 0 . 08 for generations 5 to 7 and ≈0 . 10 for generations 8 and 9; and | ρ ¯ - ρ ^ L | ≪ 0 . 05 for generations 4 to 9 when considering 1000 causal variants . This can be attributed to the increased sample size of the training population , which both improves the goodness of fit of the estimated decay curve; and makes the decay rate of the ρ ¯ closer to linear , thus making it possible for the ρ ^ L to approximate it well over a large range of FST values . To investigate this phenomenon , we gradually increased the initial training population to 4000 varieties through random mating and we observed that for such a large sample size ρ ¯ indeed decreases linearly as a function of FST . We conjecture that this is due to a combination of the higher values observed for ρ ¯ and their slower rate of decay , which prevents the latter from gradually decreasing as ρ ¯ is still far from zero after 10 generations . In addition , we note that increasing the number of causal variants has a similar effect; with 200 and 1000 causal variants ρ ¯ indeed decreases with an approximately linear trend , which is not the case with 10 and 50 causal variants . The cross-population prediction simulation based on the HUMAN data ( Fig 3 ) generated results consistent with those above . As before , the number of causal variants appears to influence the behaviour of the decay curve: while the ρ ^ D ( m ) decrease linearly for 20 , 100 and 2000 casual variants , they converge to 0 . 65 for 5 causal variants . However , unlike in the genomic selection simulation , the quality of the estimated decay curve does not appear to degrade as the number of causal variants decreases . This difference may depend on the lack of a systematic selection pressure in the current simulation , which made the decay curve overestimate predictive correlation when considering 10 variants in the previous simulation . Finally , as in the analysis of the MICE data , the linear approximation ρ ^ L to the decay curve provides a way to extend the reach of the decay curve to estimate predictive correlations ρ ^ P for distantly related populations ( AMERICA , AFRICA , OCEANIA ) . Again we observe some loss in precision ( see Table 2 ) , but the extension still crosses the 95% confidence intervals of those ρ ^ P 14 times out of 18 ( 78% ) .
Being able to assess the predictive accuracy is important in many applications , and will assist in the development of new models and in the choice of training populations . A number of papers have discussed various aspects of the relationship between training and target populations in genomic prediction , and of characterising predictive accuracy given some combination of genotypes and pedigree information . For instance , [51] discusses how to choose which individuals to include in the training population to maximise prediction accuracy for a given target population using the coefficient of determination . [52] separates the contributions of linkage disequilibrium , co-segregation and additive genetic relationships to predictive accuracy , which can help in setting expectations about the possible performance of prediction . [53] and [22] link predictive accuracy to kinship in a simulation study of dairy cattle breeding; and [54] investigates the impact of population size , population structure and replication in a simulated biparental maize populations . The approach we take in this paper is different in a few , important ways . Firstly , we choose to avoid the parametric assumptions underlying GBLUP and the corresponding approximations based on Henderson’s equations that provide closed-form results on predictive accuracy in the literature . It has been noted in our previous work [31] and in the literature ( e . g . [32] ) that in some settings GBLUP may not be competitive for genomic prediction; hence we prefer to use models with better predictive accuracy such as the elastic net for which the parametric assumptions do not hold . Our model-agnostic approach is beneficial also because decay curves can then be constructed for current and future competitive models , since the only requirement of our approach is that they must be able to produce an estimate of predictive correlation . Secondly , we demonstrate that the decay curves estimated with the proposed approach are accurate in different settings and on human , plant and animal real-world data sets . This complements previous work that often used synthetic genotypes and analysed predictive accuracy in a single domain , such as forward simulation studies on dairy cattle data . Finally , we recognise that the target population whose phenotypes we would like to predict may not be available or even known when training the model . In plant and animal selection programs , one or more future rounds of crossings may not yet have been performed; in human genetics , prediction may be required into different demographic groups for which no training data are available . Therefore , we are often limited to extrapolating a ρ ^ D to estimate the ρ ^ P we would observe if the target population were available . Prior information on F ^ ST values is available for many species such as humans [39 , 43]; and can be used to extract the corresponding ρ ^ D from a decay curve . We observe that the decay rate of ρ ^ D is approximately linear in F ^ ST for most of the curves , suggesting that regressing the ρ ^ D ( m ) against the F ^ ST ( m ) is a viable estimation approach . This has the advantage of being computationally cheaper than producing a smooth curve with LOESS since it requires fewer ( F ^ ST ( m ) , ρ ^ D ( m ) ) points and thus fewer genomic prediction models to be fitted . In fact , if we assume that the decay rate is linear we could also estimate it as the slope of the line passing through ( F ^ ST ≈ 0 , ρ ^ CV ) and ( F ^ ST ( m ) , ρ ^ D ( m ) ) for a single , small value of m . It should be noted , however , that several factors can cause departures from linearity , including the number of causal variants underlying the trait , the use of small training populations and the confounding effect of exogenous factors . In the case of the MICE data , for instance , predictions may be influenced by cage effects; in the case of the WHEAT data , environmental and seasonal effects might not be perfectly captured and removed by the trials’ experimental design . We also note that the decay curves for traits with small heritabilities will almost never be linear , because ρ ^ D converges asymptotically to zero . Unlike the results reported in [22] , we do not find a statistically significant difference between the strength of the linear relationship between ρ ^ D and F ^ ST and that between the respective squares . There may be several reasons for this discrepancy; the simulation study in [22] was markedly different from the analyses presented in this paper , since it used simulated genotypes to generate the population structure typical of dairy cattle and since it used GBLUP as a genomic prediction model . We also observe that when F ^ ST ( m ) ≈ 0 , both ρ ^ D ( m ) and ρ ^ L are , as expected , similar to the ρ ^ CV obtained by applying cross-validation to the training populations selected from the WHEAT and MICE data . This suggests that indeed ρ ^ CV is an accurate measure of predictive accuracy only when the target individuals for prediction are drawn from the same population as the training sample , as previously argued by [14] and [19] , among others . Some limitations of the proposed approach are also apparent from the results presented in the previous section . The most important of these limitations appears to be that in the context of a breeding program the performance of the decay curve depends on the polygenic nature of the trait being predicted , as we can see by comparing the panels in Fig 1 . This can be explained by the fact that causal variants underlying less polygenic , highly and moderately heritable traits will necessarily have some individually large effects . As each of those variants approaches fixation due to selection pressure , allele frequencies in key areas of the genome will depart from those in the training population and the accuracy of any genomic prediction model will rapidly decrease [21] . However , these selection effects are genomically local and so have little impact on F ^ ST . A similar effect has been observed for flowering time in the WHEAT data . [4] notes that the Ppd-D1a gene is a major driver of early flowering , but it is nearly monomorphic in one allele in French wheat varieties and nearly monomorphic in the other allele in Germany and the UK . As a result , even though the F ^ ST for those countries are as small as 0 . 031 and 0 . 042 , ρ ^ D widely overestimates ρ ^ P in both cases . A possible solution would be to compute F ^ ST only on the relevant regions of the genome or , if their precise location is unknown , on the relevant chromosomes; or to weight F ^ ST to promote genomic regions of interest . On the other hand , in the case of more polygenic traits a larger portion of the genome will be in linkage disequilibrium with at least one causal variant , and their effects will be individually small . Therefore , F ^ ST will increase more quickly in response to selection pressure and changes in predictive accuracy will be smoother , thus allowing ρ ^ D to track them more easily . Indeed , in the WHEAT data the genomic prediction model for flowering time has a much smaller number of non-zero coefficients ( 28 ) compared to yield ( 91 ) , height ( 286 ) and grain protein content ( 121 ) . Similarly , in the MICE data the model fitted on F010 to predict weight has only 168 non-zero coefficients while others range from 212 to 1169 non-zero coefficients . By contrast , all models fitted for predicting weight , which correspond to curves that well approximate other families’ ρ ^ P , have between 1128 and 2288 non-zero coefficients . The simulation on the HUMAN data suggests different considerations apply to outbred species . Having some large-effect causal variants does not necessarily result in low quality decay curves; on the contrary , if we assume that the trait is controlled by the same causal variants in the training and target populations it is possible to have a good level of agreement between the ρ ^ D and the ρ ^ P . Intuitively , we expect strong effects to carry well across populations and thus ρ ^ D does not decrease beyond a certain FST . However , this will mean that the curves will not be linear and ρ ^ L will underestimate ρ ^ P ( see Fig 3 , top left panel ) . We also note that effect sizes are the same in all the populations , which may make our estimates of predictive accuracy optimistic . Another important consideration is that since the decay curve is extrapolated from the training population , its precision decreases as FST increases , as can be seen from both simulations and by comparing the WHEAT and MICE data . Predictions will be poor in practice if the target and the training populations are too genetically distinct; an example are rice subspecies [17] , which have been subject to intensive inbreeding . The trait to be predicted must have a common genetic basis across training and target populations . However , the availability of denser genomic data and of larger samples may improve both predictive accuracy and the precision of the decay curve for large FST . Furthermore , the range of the decay curve in terms of FST depends on the amount of genetic variability present in the training population; the more homogeneous it is , the more unlikely that k-means clustering will be able to split it in two subsets with high F ^ ST ( 0 ) . One solution is to assume the decay is linear and use ρ ^ L instead of ρ ^ D to estimate ρ ^ P; but as we noted above this is only possible if ρ ^ P ≫ 0 . If ρ ^ P ≈ 0 , the decay curve estimated with LOESS from ρ ^ D can converge asymptotically to zero as F ^ ST increases; but the linear regression used to estimate ρ ^ L will continue to decrease until ρ ^ L ≪ 0 . Another possible solution is to try to increase F ^ ST by moving observations between the two subsets , but improvements are marginal at best and there is a risk of inflating ρ ^ D . Even with such limitations , estimating a decay curve for predictive correlation has many possible uses . In the context of plant and animal breeding , it is a useful tool to answer many key questions in planning genomic selection programs . Firstly , different training populations ( in terms of allele frequencies , sample size , presence of different families , etc . ) can be compared to choose that which results in the slowest decay rate . Secondly , the decay curve can be used to decide when genomic prediction can no longer be assumed to be accurate enough for selection purposes , and thus how often the model should be re-trained on a new set of phenotypes . Unlike genotyping costs , phenotyping costs for productivity traits have not decreased over the years . Furthermore , the rate of phenotypic improvements ( i . e . selection cycle time ) can be severely reduced by the need of performing progeny tests . Therefore , limiting phenotyping to once every few generations can reduce the cost and effort of running a breeding program . The presence of close ancestors in the training population suggests that decay curves are most likely reliable for this purpose , as we have shown both in the simulations and in predicting newer wheat varieties from older ones in the WHEAT data . The other major application of decay curves is estimating the predictive accuracy of a model for target populations that , while not direct descendants of the training population , are assumed not to have strongly diverged and thus to have comparable genetic architectures . Some examples of such settings are the cross-country predictions for the WHEAT data , the cross-family predictions for the MICE data and across human populations . In human genetics , decay curves could be used to study the accuracy of predictions and help predict the success of interventions of poorly-studied populations . In plant and animal breeding , on the other hand , it is common to incorporate distantly related samples in selection programs to maintain a sufficient level of genetic variability . Decay curves can provide an indication of how accurately the phenotypes for such samples are estimated , since the model has not been trained to predict them well and they are not as closely related as the individuals in the program .
|
The availability of increasing amounts of genomic data is making the use of statistical models to predict traits of interest a mainstay of many applications in life sciences . Applications range from medical diagnostics for common and rare diseases to breeding characteristics such as disease resistance in plants and animals of commercial interest . We explored an implicit assumption of how such prediction models are often assessed: that the individuals whose traits we would like to predict originate from the same population as those that are used to train the models . This is commonly not the case , especially in the case of plants and animals that are parts of selection programs . To study this problem we proposed a model-agnostic approach to infer the accuracy of prediction models as a function of two common measures of genetic distance . Using data from plant , animal and human genetics , we find that accuracy decays approximately linearly in either of those measures . Quantifying this decay has fundamental applications in all branches of genetics , as it measures how studies generalise to different populations .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"biotechnology",
"population",
"genetics",
"cereal",
"crops",
"plant",
"science",
"mathematics",
"forecasting",
"statistics",
"(mathematics)",
"crops",
"plant",
"genomics",
"mammalian",
"genomics",
"population",
"biology",
"plants",
"research",
"and",
"analysis",
"methods",
"curve",
"fitting",
"grasses",
"crop",
"science",
"mathematical",
"functions",
"mathematical",
"and",
"statistical",
"techniques",
"plant",
"genetics",
"wheat",
"animal",
"genomics",
"agriculture",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"evolutionary",
"biology",
"plant",
"biotechnology",
"statistical",
"methods",
"genomic",
"medicine",
"organisms",
"human",
"genetics"
] |
2016
|
Using Genetic Distance to Infer the Accuracy of Genomic Prediction
|
Onchocerciasis or "river blindness" is a chronic parasitic neglected tropical disease which is endemic both in mainland and insular Equatorial Guinea . We aim to estimate the current epidemiological situation of onchocerciasis in Bioko Island after vector elimination in 2005 and more than sixteen years of Community Directed Treatment with Ivermectin ( CDTI ) by using molecular and serological approaches for onchocerciasis diagnosis . A community-based cross-sectional study was carried out in Bioko Island from mid-January to mid-February 2014 . A total of 544 study participants were recruited . A complete dermatological examination was performed and three skin snips were performed in every participant for parasitological and molecular assessments . Blood spots were also taken for determination of Ov16 IgG4 antibodies trough an “in-house” ELISA assay . Overall , we found 15 out of 522 individuals suffering any onchocerciasis specific cutaneous lesions and 16 out of 528 ( 3 . 0% ) with onchocercal nodules in the skin . Nodules were significantly associated with age , being more common in subjects older than 10 years than in younger people ( 3 . 9% vs . 0% , p = 0 . 029 ) . Regarding the onchocerciasis laboratory assessment , no positive parasitological test for microfilaria detection was found in the skin snips . The calculated seroprevalence through IgG4 serology was 7 . 9% . No children less than 10 years old were found to be positive for this test . Only one case was positive for Onchocerca volvulus ( O . volvulus ) after skin PCR . The present study points out that the on-going mass ivermectin treatment has been effective in reducing the prevalence of onchocerciasis and corroborates the interruption of transmission in Bioko Island . To our knowledge , this is the first time that accurate information through molecular and serological techniques is generated to estimate the onchocerciasis prevalence in this zone . Sustained support from the national program and appropriate communication and health education strategies to reinforce participation in CDTI activities are essential to ensure progress towards onchocerciasis elimination in the country .
Onchocerciasis or "river blindness" is a chronic parasitic neglected tropical disease caused by the filarial nematode Onchocerca volvulus ( O . volvulus ) . It is transmitted to humans through exposure to repeated bites of infected blackflies of the genus Similium [1] . Adult worms live in subcutaneous nodules and form deeper worm bundles , where fertilized females can produce , during an average of ten years , millions of embryonic larvae ( microfilariae ) responsible for the morbidity associated with this disease [2] . Several simulid species have been incriminated in the transmission of O . volvulus . The complex S . damnosum and , to a lesser extent , the S . naevei complex are the most frequently found in Africa and the Arabian Peninsula [1] . Onchocerciasis affects many systems and organs , but most important morbidity is due to cutaneous and ophthalmologic manifestations , with different clinical grades [3] . The presence of one or another clinical manifestation varies depending on the most prevalent parasite strain circulating in the area: blindness tends to occur more frequently in the West African areas of savannah while onchocerciasis cutaneous disease ( OCD ) prevails in African forest areas [1] . Itching is usually the first clinical manifestation of onchocerciasis in the skin , and may occur alone or associated with OCD . Murdoch et al . ( 1993 ) described a grading system for OCD and defined five main categories , which can coexist together: acute papular onchodermatitis , chronic papular onchodermatitis , lichenified onchodermatitis , atrophic onchodermatitis and depigmented onchodermatitis [3] . This classification and grading system allowed to highlight the linkages between the different OCD forms with onchocerciasis epidemiology in different endemic areas [4] . Traditionally , the most common method for diagnosis of onchocerciasis was the detection of microfilariae ( MF ) in small , superficial skin biopsies ( skin snips ) . However , skin snip examination is not sufficiently sensitive for detection of early infections or for diagnosis in persons with low MF densities . The low sensitivity of this parasitological method makes it unreliable in hypo endemic areas . So its value is insufficient to support an epidemiological decision . More recent and sensitive approaches include antibody-based diagnostic tests and PCR . Improved methods are needed for field diagnosis of onchocerciasis , to support efforts aimed at elimination of the disease [5 , 6] . Actually , the updated WHO guidelines ( 2016 ) for onchocerciasis control suggest the use of molecular tools in situations where interruption of transmission is suspected [7] . Onchocerciasis is endemic to tropical regions both in Africa and Latin America and in the Yemen . In Latin America , it is found in 13 foci located in 6 different countries . In 8 of the 13 foci in the region onchocerciasis elimination and transmission interruption has been achieved thank to Onchocerciasis Elimination Program for the Americas ( OEPA ) efforts [8] . More than 99% of onchocerciasis infected people live in Africa . It has been estimated that 21 , 115 , 000 members of the 118 , 285 , 000 African population at risk for infection are infected . Among these , 690 , 000 experienced visual impairment and 220 , 000 were totally blind [9 , 10] . Between 1974 and 2002 , onchocerciasis was brought under control in West Africa through the work of the Onchocerciasis Control Program ( OCP ) , mainly focused on by vector control . OCP was supplemented by large-scale distribution of ivermectin from 1989 . The African Program for the Control of Onchocerciasis ( APOC ) was launched in 1995 , with the aim to establish sustainable community-based systems for distribution of ivermectin ( CDTI ) in those countries where onchocerciasis was still a public health problem [11] . As a result of the success due to sustained onchocerciasis control activities , APOC paradigm has recently changed from control to a strategy of onchocerciasis elimination ‘where feasible’ [12] . In this new context , monitoring and evaluation activities become especially necessary in order to document if the transmission has been interrupted . The new elimination goal requires new approaches for assessment of CDTI needs in areas with lower infection prevalence , that may not be efficiently identified by methods such us Rapid Epidemiological Mapping for Onchocerciasis ( REMO ) [13] . Onchocerciasis is an endemic disease both in mainland and insular Equatorial Guinea [14] . Onchocerciasis control activities started in 1989 , mainly based on massive distribution of ivermectin , but it was only in 1998 when CDTI activities were started as part of the proposed APOC mandate [15] . Since then , ivermectin coverages on the eligible population have ranged yearly between 50% and 75% [16] . Despite the fact that ivermectin coverage was below 85% , which is the coverage needed in a sustained fashion to interrupt transmission , progress towards this goal in Bioko Island has been made . This is partially due to the success in aerial larviciding campaign in 2005 , which achieved elimination of S . yahense from Bioko Island , with no evidence of vector reappearance in the following three years . No further entomological assessments have been performed after 2008 [17] . Data from 1989 showed that overall onchocerciasis prevalence ( measured through MF skin snips assessment ) in Bioko Island was 75 . 2% ( range 51 . 9% to 87 . 1% ) . This prevalence dropped to 38 . 4% after eight years of treatment with ivermectin ( from 1989 to 1998 ) , according to Mas et al [15] . The most recent available data from the Equatoguinean Ministry of Health ( MoH ) , based on MF skin snip assessment , estimated a prevalence of 0–3% in Bioko Island ( 2013 data ) . Despite these achievements , the follow-up of CDTI activities has been irregular and there is no accurate information on the current prevalence of onchocerciasis in the whole country . Thus , we aim to estimate the current epidemiological situation of onchocerciasis in Bioko Island after more than sixteen years of CDTI activities , by using molecular and serological approach for onchocerciasis diagnosis . To our knowledge , this is the first time that the current epidemiological situation of onchocerciasis in Bioko Island is estimated by using non-traditional methods for onchocerciasis diagnosis .
The Island of Bioko is a part of the Republic of Equatorial Guinea , which also includes Rio Muni on the mainland and the island of Annobon . It is located in the Bay of Guinea in Central Africa , about 40 km southwest of the Cameroon coast . The surface area of Bioko Island is of approximately 2 , 017 km2 , and it is about 72 km in length . Most of its 260 , 000 inhabitants live in the northern part of the island . The interior of the island is covered with dense forests on the steep slopes of volcanoes and calderas . The highest peak on the island reaches 3 , 011 m above sea level . The island has a humid tropical environment . Mean daily maximum and minimum temperatures range between 29–32°C and 19–22°C , respectively . A cross-sectional study was conducted from mid-January to mid-February 2014 . Sampling was carried out by multistage cluster survey . The sample size was computed using Epi-Info version 3 . 4 . 1 free software considering the following parameters: 10% hypothesized prevalence and 2% standard error . We assumed a design effect of 2 corresponding to the complex design . The initial sample size was 450 . It was increased ( +20% ) in prevision of missing data . n = DEFF p ( 1-p ) /e^2 , where DEFF is the design effect , e is the desired standard error and p is the prevalence Firstly , twenty communities were randomly selected with probability proportional to size ( Fig 1 ) . Second sampling units were randomly selected households from an updated census from each community , provided by the head of the village ( in rural areas ) or neighbourhood ( in urban zones ) . In every selected household , all individuals aged 5 years or above who had permanently lived in Bioko Island during the last five years were recruited . Children from 5 to 9 years old were included in the study in order to detect exposure to O . volvulus through Ov16 , IgG4 serology test . A closed ended structured questionnaire was administered to every study participant by trained medical personnel . The questionnaire was pre-tested on close areas not included in this study for clarity and cultural acceptability . It comprised the following parts: socio-demographic characteristics , risk factors for onchocerciasis and clinical data . Each interview was made by house-to-house visit . If the study participant was aged less than 15 years old , the questionnaire was answered by a parent or guardian of the teenager . A complete dermatological examination was performed in each participant in a well-lit private room . Palpable nodules and signs of onchocercal skin disease ( onchodermatitis ) were assessed by trained health staff . Results were recorded on the form of “positive” or “negative” . Three skin snips specimens were collected from every participant ( two from right iliac crest , one from left iliac crest ) . No special collection time was considered as the microfilariae of O . volvulus are non-periodic . Two samples were immersed in normal saline solution to prevent the preparation from drying out . Then , they were sent to the local hospital laboratory to be read under a 10X microscope after 24 hours . Results were expressed for each individual as ‘positive’ or ‘negative’ . Laboratory results were recorded on the original ( field ) registration form . The third skin snip was stored at 4°C before shipping the samples to the National Centre of Microbiology , Health Institute Carlos III ( Spain ) , where further PCR analysis were performed . The collected data from the questionnaires and lab assessments were merged with a unique individual id and double entered into a data entry file using EpiData software , V . 3 . 1 . The data were then transferred to SPSS version 18 . 0 ( SPSS Inc . , Chicago , Illinois , USA ) . Frequencies , means and standard deviations ( SD ) were computed to summarize the data . Prevalence results with 95% confidence intervals ( CI ) were also calculated . Bivariate analyses by age group were performed with χ2 test for categorical data . Where a cell value was below 5 , Fisher’s exact test for two–way tables was applied . The criterion for significance was set at p<0 . 05 based on a two-sided test . The study was approved by the ethical advisory boards of the Health Institute Carlos III in Spain and the Ministry of Health ( MoH ) in Equatorial Guinea ( CEI PI 21_2014 ) . The study complied with current national and international regulations and standards for biomedical research in human subjects . The village and neighborhood representatives were informed of the day of the visit and the scope of the study by an official letter from the Equatoguinean MoH . Written informed consent was obtained from all patients prior to study inclusion . Anonymity was assured . A written statement was also included on the introductory part of the questionnaires in which further information concerning the purpose of the study and the confidentiality of the research information was given . The written consent was obtained from parents or guardians in those individuals younger than 18 years old . Data were analysed in anonymous form .
The 16 . 4% of the interviewees referred to have suffered onchocerciasis in the past . Overall , 15 out of 522 individuals pointed out suffering any onchocerciasis specific cutaneous lesions . During the clinical examination , we found that 78 out of 523 individuals ( 14 . 9% ) presented itching at the time of answering the questionnaire , and 16 out of 528 ( 3 . 0% ) had onchocercal nodules in the skin . The 1 . 3% of the study sample ( 7/544 ) showed clinical signs of onchodermatitis while 10/544 ( 1 . 8% ) interviewees had any degree of leopard skin . Only 2 out of 544 presented both cutaneous manifestations ( onchodermatitis and leopard skin ) . After stratification by age group , we found that nodules were significantly associated with age , being more common in subjects older than 10 years than in younger people ( 3 . 9% vs . 0% , respectively , p = 0 . 029 ) . Also pruritus was more frequently found in adults ( 17 . 6% ) than children ( 5 . 9% , p = 0 . 002 ) ( Table 2 ) . Regarding laboratory results , none of the 544 skin snips assessments for MF detection was found positive . In parallel , 541 skin snips were submitted to specific skin PCR for Loa loa , Mansonella spp and O . volvulus ( 3 skin samples were missing ) . Skin PCR test was positive in 11/541 cases ( 2 . 0% ) , all of which were adults . After DNA sequencing only one case was positive for O . volvulus . The remaining ones included: seven positive cases for Mansonella perstans , two positive cases for Mansonella streptocerca and one case with Loa Loa , which was considered to come from blood contamination ( Table 3 ) . Blood samples were obtained from 531 out of 544 individuals and analyzed for identification of Ov16 IgG4 antibodies by ELISA . Globally , the seroprevalence calculated was 7 . 9% , with and CI95%: ( 5 . 9%-10 . 6% ) . From the samples studied , 43 samples were positive and 15 undefined by OV16 -ELISA ( Table 3 ) . No children less than 10 years old were found to be positive for this test . In relation to preventive practices among our study population , we found that 274/544 individuals ( 50 . 4% ) referred that they had never taken the drug . Overall , 28% had taken ivermectin more than twice in the last five years , while 15 . 6% had taken it less than twice . Among those who referred to have taken ivermectin , 65 . 6% pointed out that they received the drug from a community distributor . When asking about the reasons for not taking ivermectin , we observed that the lack of information was the most common reason ( 29 . 9% ) , followed by not having suffered the disease ( 20 . 5% ) , age ( 16 . 5% ) , not being at home at the time of the drug campaign ( 14 . 3% ) and lack of access to the drug ( 11 . 6% ) ( Table 4 ) . Children under 10 years old referred having taken ivermectin significantly less frequently than older participants ( p<0 . 005 ) , while they were more likely to have taken the drug last year than those with older age ( p = 0 . 002 ) . Moreover , being younger was significantly associated with getting ivermectin from a community distributor ( p = 0 . 023 ) . The most frequent reason for not taking ivermectin in children younger than 10 years old was age ( 41 . 0% ) while for adults it was mainly due to the lack of information ( 37 . 7% ) ( p = 0 . 000 ) .
Overall , we found a low number of individuals presenting clinical cutaneous manifestations of onchocerciasis in our study , and most of them were adults , as expected . Pruritus was the most common symptom , and only a few individuals pointed out suffering onchocerciasis specific cutaneous lesions ( mainly leopard skin and onchodermatitis ) . Moreover , we found a very low number of onchocercal nodules carriers , which is coherent with the hypoendemic situation in Bioko Island . O . volvulus transmission in Bioko Island focus has been extensively documented since 1990 to date [14] . In a survey carried out in the mid 80’ in the Island of Bioko the 28 . 8% of the study population presented with dermatitis , pigmentation changes and cutaneous atrophy [23] . The global prevalence of nodules was 27 . 2% . This research showed that onchocerciasis was widespread over the Island , with a high rate of population at risk of infection . More recent data ( 1998 ) showed that prevalence of skin depigmentation , the proxy of longstanding infection of onchocerciasis in the community , was 9 . 4% , with a reduction of 7 . 7% after eight years of vertical ivermectin distribution [15] . In this study , it was pointed out that carriers of nodules in children aged 0–4 years decreased 2 . 6 times , while no significant changes were observed in other age groups [10] . This result reflected a reduction in onchocerciasis general transmission after several cycles of ivermectin distribution and vector elimination in Bioko Island . The impact of ivermectin on OCD in onchocerciasis endemic countries has been described in the literature [24] . A multi-country study by Ozoh et al ( 2011 ) in meso and hyperendemic communities of seven study sites in Cameroon , Sudan , Nigeria and Uganda showed a substantial reduction in itching and all forms of OCD after five or six years of CDTI , including a reduction in nodules [24] . Low rates of nodule prevalence can also found in countries with long history of onchocerciasis control activities , such as most areas in Malawi , Kenya and Rwanda [25] . In our study , we observed most cases presenting itching and OCD manifestations ( including nodules ) were older than 10 years old , and depigmentation was exclusive of adults . Linkages between OCD and age have been well documented [24] . Age is considered to be a risk factor for depigmentation while reactive skin lesions overall have been described to be linked with younger ages [26] . No positive MF skin snip assessments were found in our study population . Only one individual was found to be positive for O . volvulus in skin by PCR . Moreover , almost 8% of individuals were positive for IgG4 antibodies for Ov16 by using ELISA test . None of them were children . Since the introduction of ivermectin distribution in 1987 in Bioko Island , an increasing number of communities have been enrolled into the treatment programme , resulting in a substantial reduction in disease prevalence [14] . Data from 1989 showed that overall prevalence ( measured through MF skin snips assessment ) and mean microfilarial density in Bioko Island were 75 . 2% ( range 51 . 9% to 87 . 1% ) and 32 . 2 mf/snip respectively [15] . Later on , Mas et al described a reduction in prevalence to 38 . 4% after eight years of treatment with ivermectin ( from 1989 to 1998 ) [15] . In this study , both prevalence and intensity of infection dropped among the children under 5 years who had never been treated with an anti-filarial drug , suggesting an indirect effect of ivermectin treatment towards younger groups . Similarly , studies in Burundi [27] and Cameroon [28] have also shown a high reduction in prevalence and the intensity of microfilaridermia in young children who have never received ivermectin , but lived in four annual rounds of mass ivermectin treated communities . The most recent available data from the Equatoguinean MoH ( 2013 ) showed a prevalence of 0–3% in Bioko Island by MF skin snip assessment . This is coherent with the parasitological results in our study . Both microscopic detection of MF in skin snips and nodular palpation are useful epidemiological tools in the field for diagnosis and monitoring of onchocerciasis prevalence in many endemic countries [21 , 29–31] . Nevertheless , improved methods are needed for those areas with lower infection prevalences , which might be close to elimination [13] . Determining IgG4 antibodies for parasite-specific 16 kDa antigen ( Ov16 ) is a recognized epidemiological tool to certificate the interruption of onchocerciasis transmission in endemic countries [32–35] . According to recent recommendations from WHO ( 2016 ) , the critical threshold for interruption or elimination of transmission is the upper bound of the 95% confidence interval of less than 0 . 1% confirmed seropositivity to Ov-16 in children under 10 years of age [7] . In our study , the absence of positive cases for IgG4 antibodies for Ov16 by using ELISA in children younger than 10 years old suggest that the interruption of transmission might have been achieved in Bioko Island . However , the sample size should be increased to 2 , 000 children , in order to meet the WHO requirement needed to verify the interruption of onchocerciasis transmission [7] . Only one case was positive for O . volvulus by PCR , while the rate of onchocerciasis exposure by IgG4 antibodies was remarkably higher . The unique positive case by PCR came from a rural community in Riaba district , usually not targeted by CDTI . Surprisingly , this case´ serological test was negative , thus it might be a new infection . In a study carried out by Evans et al in Nigeria , the prevalence of onchocerciasis in children was also higher when measured by IgG4 antibodies than with skin snip . Probably , these infections were below the sensitivity of a skin snip . Other possible explanation is that their antibody response fitted a recent exposure rather than a patent one [35] . Molecular technics based on PCR have demonstrated to be more sensitive than skin snip microscopy or nodule palpation for detecting onchocerciasis [13 , 20] . In future research , the use of PCR in skin snips from children who are Ov16 positive may be considered as a confirmatory test . The mentioned remarkable decrease on onchocerciasis prevalence in Bioko Island happened in spite of a moderate current involvement in preventive practices in the community . We found that more than half the population had never taken ivermectin . Among those who had taken ivermectin , most of them received the drug from a community distributor . The most important identified reason for not taking ivermectin included lack of information , followed by not suffering the disease . Previous research from APOC ( 2008 ) highlighted a progressive increase in ivermectin therapeutic coverages in Bioko in the last years [16] . According to the most recent data from the Equatoguinean MoH , 80% of therapeutic coverage was reached during 2013 ivermectin campaign . It should be taken into account that this assessment was only performed in sentinel sites while our analysis but the whole geographical territory of Bioko Island . This might explain differences with our results . Weaknesses in targeting younger ages ( < 10 years old ) were also identified as less than one fourth of the children had ever taken the drug . Within this group , age was identified as the main reason only in less than half . Bearing in mind that ivermectin distribution campaigns target children after 5 years , our findings suggest a reduction in community participation in ivermectin campaigns as success in control program is achieved . This could be explained by a decrease on risk perception [36] or a relaxation of the control program activities due to financial constraints . Evidence that long-term ivermectin treatment alone might interrupt and eliminate onchocerciasis in African countries has been well documented [22 , 34 , 35] . Overall , our findings support that interruption of transmission might have been achieved in Bioko Island , and interruption of ivermectin distribution could be considered in the near future . Nevertheless , decisions about ceasing CDTI activities in Bioko Island should be cautious in order to avoid recrudescence . This situation has been described in some neighbouring countries such as Cameroon after more than a decade of onchoderciasis control activities , which showed a reduction of meso- and hyperendemic onchocerciasis areas to hypoendemic areas . Nevertheless , transmission appears to continue in many areas as it was found that children under 10 years of age in the follow-up surveys had positive skin snips for MF [28] . Considerations about when to interrupt ivermectine distribution and initiate Post Treatment Survaillance ( PTS ) phase in Bioko Island are still unclear . Decisions on that respect will require additional entomologic studies ( not performed since 2008 ) . Moreover , extended serology studies with children of 10 years of age and under are needed in order to reach at least a 2000 children population , according to WHO guidelines ( 2016 ) [7] . Pre-intervention data on the prevalence and intensity of onchocerciasis infection were not available in some of the assessed communities as background information in order to assess the long-term impact of ongoing CDTI program . Moreover , the sample size was not large enough to draw definite conclusions about interruption of transmission in Bioko . Some limitations related to the comprehension level of the questionnaire among the study population should also be noted . Regarding the Ov16 ELISA test , it is currently not known how long the IgG4 antibody response to the Ov16 antigen persists in exposed individuals and specificities vary according to authors . Cross-reactions with Mansonella species have been described [37] . Finally , although PCR is considered a highly sensitive test , it should be noted that cost would limit its use as regular epidemiological tool to apply on entire populations . Our findings support the idea that Equatorial Guinea is moving fast towards elimination after long term ivermectin distribution and the elimination of the vector Simulium yahense Bioko form from Bioko Island in 2005 , among other factors . Our results will contribute to strengthen and optimize the current onchocerciasis control activities supported by the National Control Program . Appropriate communication and health education strategies to reinforce participation in CDTI activities are essential to ensure progress towards onchocerciasis elimination in the country .
|
Onchocerciasis or “river blindness” is a chronic parasitic disease which is mainly found in Sub-Saharan Africa . Onchocerciasis is endemic in both mainland and insular Equatorial Guinea . Huge achievements have been made on onchocerciasis control in Bioko Island in the last years , and the country is moving fast towards elimination . In the new elimination context , monitoring and evaluation activities with more sensitive diagnostic tools become especially necessary in order to confirm that transmission has been interrupted . Previous data on the epidemiological situation of onchocerciasis in Bioko Island are mainly based on microfilaria ( MF ) skin snip assessments . We aim to create evidence towards the fact that onchocerciasis transmission might have been achieved in Bioko Island after more than sixteen years of onchocerciasis control activities by using molecular and serological technics for onchocerciasis diagnosis .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"onchocerca",
"volvulus",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"animals",
"onchocerca",
"parasitology",
"neglected",
"tropical",
"diseases",
"onchocerciasis",
"infectious",
"disease",
"control",
"immunologic",
"techniques",
"africa",
"research",
"and",
"analysis",
"methods",
"equatorial",
"guinea",
"infectious",
"diseases",
"serology",
"immunoassays",
"people",
"and",
"places",
"helminth",
"infections",
"nematoda",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2016
|
Evidence for Suppression of Onchocerciasis Transmission in Bioko Island, Equatorial Guinea
|
Unlike mammals and birds , teleost fish undergo external embryogenesis , and therefore their embryos are constantly challenged by stresses from their living environment . These stresses , when becoming too harsh , will cause arrest of cell proliferation , abnormal cell death or senescence . Such organisms have to evolve a sophisticated anti-stress mechanism to protect the process of embryogenesis/organogenesis . However , very few signaling molecule ( s ) mediating such activity have been identified . liver-enriched gene 1 ( leg1 ) is an uncharacterized gene that encodes a novel secretory protein containing a single domain DUF781 ( domain of unknown function 781 ) that is well conserved in vertebrates . In the zebrafish genome , there are two copies of leg1 , namely leg1a and leg1b . leg1a and leg1b are closely linked on chromosome 20 and share high homology , but are differentially expressed . In this report , we generated two leg1a mutant alleles using the TALEN technique , then characterized liver development in the mutants . We show that a leg1a mutant exhibits a stress-dependent small liver phenotype that can be prevented by chemicals blocking the production of reactive oxygen species . Further studies reveal that Leg1a binds to FGFR3 and mediates a novel anti-stress pathway to protect liver development through enhancing Erk activity . More importantly , we show that the binding of Leg1a to FGFR relies on the glycosylation at the 70th asparagine ( Asn70 or N70 ) , and mutating the Asn70 to Ala70 compromised Leg1’s function in liver development . Therefore , Leg1 plays a unique role in protecting liver development under different stress conditions by serving as a secreted signaling molecule/modulator .
The process of liver development includes 1 ) the specification of hepatoblasts from the endoderm , 2 ) proliferation of hepatoblasts to form the liver primordium ( liver bud ) , and 3 ) differentiation and proliferation of hepatocytes to form the embryonic liver [1–5] . Liver organogenesis is not only controlled by intrinsic transcription factors such as FoxA factors [6] , GATA factors [7] , Hhex[8] and Prox1 [9] but also by secreted signaling molecules [10] including FGF [11] , BMP [12] , Wnt[13] and RA [14] produced by neighboring mesodermal cells/tissues . Strikingly , studies of mouse , chick/quail , Xenopus and zebrafish have shown that the molecular events controlling liver development are robustly conserved across these different species although evolutionally , these species are distantly related , especially when considering the obvious anatomic differences in organ initiation and patterning [1–4] and the differences in the circumstances of their embryogenesis . Unlike mammals and birds , teleost fish complete the process of embryogenesis externally . To cope with the stresses brought about by environmental changes , teleost fish have to evolve anti-stress mechanism ( s ) to protect the process of embryogenesis/organogenesis . However , the molecule ( s ) mediating such activity is ( are ) currently unknown . Therefore , it is of great interest to determine whether other such factors , in addition to the aforementioned common factors , protect liver development during external embryogenesis in teleost fish . leg1 ( liver enriched gene 1 ) is an evolutionally conserved gene in vertebrates that encodes a novel secreted protein Leg1 , which contains only a domain of unknown function 781 ( DUF781 ) [15–17] . In zebrafish , there are two copies of the leg1 gene , namely leg1a and leg1b , which are closely linked on chromosome 20 [15] . Previous reports showed that knockdown of total Leg1 results in defective liver development , and the expression of leg1 is modulated by hypoxia conditions [18] . On the basis of the detailed analysis of leg1a mutants generated by the TALEN method , we provide strong evidence to demonstrate that Leg1 functions probably as a novel signaling molecule/modulator to protect liver development through Erk phosphorylation under stress conditions , and glycosylation at N70 in Leg1a is essential for this function .
We reported previously that leg1a but not leg1b was the predominant form expressed during the embryonic stage in zebrafish and that knockdown of leg1a resulted in a small liver phenotype [15] . To unequivocally prove the role of leg1a in liver development , we generated two leg1a mutant alleles , one with a 13-bp insertion ( leg1azju1 ) and another with a 12-bp deletion ( leg1azju2 ) ( leg1b is intact in these two leg1a alleles ) , via the TALEN technique[19] by targeting exon 1 of leg1a ( Fig 1A ) . To our surprise , unlike the leg1a morphants[15] , the leg1a homozygous mutant obtained from a cross between either leg1azju1/+ or leg1azju2/+ heterozygous male and female did not show an obvious small liver phenotype at 3 . 5 days post fertilization ( dpf ) when examined with a liver-specific molecular marker fatty acid binding protein 10a ( fabp10a ) ( Fig 1B ) , and both leg1azju1 and leg1azju2 homozygous mutants could grow to adulthood and were fertile . We examined the total Leg1 levels by western blot analysis and found no drastic difference between unfertilized eggs from wild-type ( WT ) and leg1azju1/+ heterozygous females ( Fig 1C ) . In fact , the leg1azju1 homozygous mutant obtained from leg1azju1/+ crosses retained , though showing variations , a considerable level of Leg1 at 4 dpf ( Fig 1D ) , suggesting that maternal Leg1 compensated for the need for Leg1a during early hepatogenesis[20] . The leg1azju1 homozygous mutant was propagated and allowed to produce progenies . We determined that such leg1azju1 homozygous progenies ( maternal-zygotic mutants ) lacked Leg1 ( Fig 1E ) at 1 dpf but started to express the Leg1b homolog at 3 . 5 and 7 dpf . Surprisingly , whole-mount in situ hybridization ( WISH ) using the fabp10a probe revealed that the maternal-zygotic mutant exhibited a small liver phenotype in a season-dependent manner ( Fig 1F , S1A Fig ) . For example , majority of the maternal-zygotic mutants exhibited a small liver phenotype in 14 cases recorded during the cold season whereas the mutant liver showed a great variation in sizes ranging from normal to small in 18 cases recorded during the warm/hot seasons ( Fig 1G , S1B and S1C Fig ) . These results suggest that liver development in the maternal-zygotic leg1azju1 mutant is amenable to its living environment . We tested this hypothesis by growing fish in different mild stress conditions . Growing maternal-zygotic leg1azju1 mutants in relative high temperature ( 32°C ) and high density ( 200 embryos per 10-cm diameter Petri dish ) ( Fig 2A ) or briefly treating the maternal-zygotic leg1azju1 mutants at 24 hpf with 2 . 5 mJ/cm2 ultraviolet ( UV ) irradiation ( UV25 ) ( Fig 2B , S2A and S2B Fig ) sharply increased the proportion of the mutant embryos displaying the small liver phenotype at 3 . 5 dpf . High density alone also caused a small liver phenotype to the maternal-zygotic leg1azju1 mutants ( S2C Fig ) . In addition , we found that incubating the zygotic leg1azju1 mutant in the egg water containing a mild but not lethal dose of H2O2 ( 0 . 5mM ) also led to the small liver phenotype ( Fig 2C ) . Interestingly , the maternal-zygotic leg1azju1 mutant embryos did not exhibit a small liver phenotype at 3 . 5 dpf when they were grown in the egg water containing 0 . 5% or 1% ethanol starting at 24 dpf ( S2D Fig ) , a concentration range not causing overall abnormality [21] . UV25 treatment also enhanced the small liver phenotype in leg1azju2 , another mutant allele of the leg1a gene ( S2E Fig ) . UV treatment , high temperature , andH2O2 treatment all would lead to oxidative stress [22] . To find out whether the maternal-zygotic leg1azju1 mutant is compromised in scavenging ROS caused the oxidative stress , we compared the ROS level at different time points between the UV25 treated WT and maternal-zygotic leg1azju1 embryos by DCFH-DA [23] . The result showed that the maternal-zygotic leg1azju1 embryos accumulated a higher ROS level at all time points examined within the first hour after UV treatment ( Fig 2D ) . We wondered whether the development of the small liver phenotype in leg1a mutants could be prevented by blocking the production of reactive oxygen species ( ROS ) . Diphenyleneiodonium ( DPI ) and apocynin ( APO ) are two specific inhibitors of the Duox/Nox enzyme often used to block the production of ROS[24 , 25] . We treated the maternal-zygotic leg1azju1 mutants with DPI or APO one hour prior to the UV25 treatment and found that both DPI and APO prevented the mutants from developing the small liver phenotype ( Fig 2E and 2F ) . Liver , exocrine pancreas and intestine are all derived from the endoderm [26] . Previous genetic screening found that mutants with defects in liver development often showed defective development of the exocrine pancreas and/or intestine [27 , 28] , likely because liver and exocrine pancreas share common progenitors [29 , 30] . Leg1a expression is enriched in the embryonic liver , but meanwhile , Leg1a is also a secretory protein [15] . Considering this fact , we wanted to determine whether leg1azju1 also affects development of the pancreas and other digestive organs . We used fabp10a , trypsin , insulin and fabp2a probes in WISH to mark the liver , exocrine pancreas , endocrine pancreas and intestine , respectively . Interestingly , it appeared that UV25 treatment drastically reduced the liver size , but only subtly affected the exocrine pancreas development and did not show observable effect on the intestinal tube development in the maternal-zygotic leg1azju1 mutant ( Fig 3A ) . We then used prox1 and hhex , two earlier hepatic markers , in WISH to examine the effect of UV25 treatment on liver bud formation at 30 hpf[13] . The result showed that UV25 treatment halted liver bud formation in most of the mutant embryos but not the WT ( Fig 3B ) . A TUNEL assay did not reveal any obvious differences in the apoptotic activity between the UV25-treated WT and mutant liver cells ( S3 Fig , total 12cryosections from 6 embryos examined ) . Immunostaining of phosphorylated histone 3 ( PH3 , a molecular marker for cell proliferation ) showed that leg1azju1 liver cells ( defined by immunostaining of the hepatic marker Betaine homocysteine S-methyltransferase ( BHMT ) , in red ) contained significantly fewer ( p<0 . 05 ) PH3 positive cells ( 12 of 604 total cells counted or 2 . 4% , data obtained from 6 embryos ) when compared with those in the WT ( 30 of 684 total cells counted or 4 . 43% , data obtained from 6 embryos ) at 54 hpf after UV25 treatment ( Fig 3C and 3D ) . Therefore , the maternal-zygotic leg1azju1 mutant develops a small liver phenotype under UV stress due to cell cycle arrest . One possible explanation for the inhibitory effect of UV25 treatment on liver development in the maternal-zygotic leg1azju1 mutant is that UV25 treatment induces the expression of Leg1 . However , we observed that the UV- or H2O2-treatment of the WT embryos at 24 hpf did not cause significant changes to the levels of total leg1 transcripts at 3 and 6 hours after treatment ( S4 Fig ) . UV25 treatment of the WT embryos at 24 hpf neither affected the level of total Leg1 protein at 3 , 6 , 9 and 12 hours after treatment ( Fig 4A ) . In zebrafish , 24–34 hpf is a crucial stage for hepatogenesis when signaling molecules including FGF , BMP , Wnt2bb and RA orchestrate the initiation of the liver bud [13 , 31–33] . Based on all of the above , we speculated that the signaling pathway promoting cell proliferation is probably impaired due to loss of the maternal-zygotic Leg1a . This prompted us to investigate whether Leg1 , being a secretory protein , is involved in known signaling pathways . We treated 24hpf WT and maternal-zygotic leg1azju1 mutant embryos with UV25 and found that UV25 treatment up-regulated the level of p-Erk in WT but showed an inhibitory effect on the mutant at 6 hours post treatment ( i . e . at 30 hpf ) ( Fig 4B ) . Notably , UV25 treatment did not affect the Bmp signaling as indicated by the level of pSmad1/5/8 ( Fig 4B ) . Importantly , upon UV25 treatment , Leg1a over-expression by leg1a mRNA injection at one-cell stage increased the level of p-Erk but did not alter the level of pSmad1/5/8 ( Fig 4C ) . Considering that the activation of the expression of Bmp2 by heatshock of Tg ( hsp70l:bmp2b ) [34] embryos at 18 or 24 hpf increased only the level of pSmad1/5/8 and not that of p-Erk ( Fig 4D and 4E ) , we speculated that Leg1 does not signal through the Bmp pathway but probably through the Erk signaling pathway to protect liver development under stress conditions . To test whether Leg1 acts through the Erk-signaling pathway to protect liver development , we generated a constitutively active form of Erk mutant ( caErk ) by substituting L84 to P84 ( L84P ) , S162 to D162 ( S162D ) , D330 to N330 ( D330N ) simultaneously[35] . It has been shown that over-activating Erk signaling at the early stage ( up to 80% epiboly ) negatively regulates the endoderm formation[36] . Indeed , we found that injection of caErk mRNA into one-cell stage embryos caused small liver both in WT and mutant embryos ( S5A Fig ) . To overcome the effect of Erk-signaling on early embryogenesis we injected caErk mRNA or fgf8 mRNA into the yolk at 22hpf and treated the embryos with UV25 at 24hpf . The effectiveness of this way of injection is demonstrated by the fact that the injected Cy3-labled oligo-dT can successfully reach to the prospective liver bud region ( S5B Fig ) . We found that such injection rescued the mutant liver development to a great extent ( S5C Fig ) . Next , we cloned the caErk gene downstream of the doxycycline ( Dox ) inducible promoter tetracycline response element ( TRE ) promoter[37 , 38] ( S6A Fig ) . The expression of caErk is effectively induced by Dox after a low dosage ( 10 pg ) of plasmid injection although a weak leakage of the TRE promoter was observed ( S6B Fig ) . The TRE-caErk plasmid ( 10 pg ) was injected into one-cell stage maternal-zygotic leg1zju1mutant embryos and the injected embryos were treated with UV25 at 24 hpf followed immediately by addition of the drug Dox ( final concentration 30 μg/mL ) . The liver development in these embryos was examined with the fabp10a probe at 3 . 5 dpf . The result showed that induction of the caErk expression between 24hpf and 33hpf achieved a significant rate of rescue of the liver growth in maternal-zygotic leg1zju1mutant ( Fig 4F ) while overall features of the injected embryos appeared relatively normal ( S6C Fig ) . We showed previously that Leg1 is a classical secretory protein [15] . Because glycosylation is a common modification for a secretory protein [39] , we checked whether Leg1 is also modified by glycosylation . There are two types of glycosylation , N-glycosylation and O-glycosylation [40] . N-glycosylation can be cleaved by PNGase F [41]whereas O-glycosylation can be cleaved by the combination of endo-α-N-acetylgalactosaminidase plus neuraminidase [42] . We previously reported that adult fish serum contains a high level of Leg1 [15] . We used these enzymes to treat the serum protein and also total protein extracted from embryos at 3dpf , respectively , and found that only PNGase F treatment caused a band shift ( Fig 5A and 5B ) . Because both leg1a and leg1b are expressed in the adult liver to produce the serum Leg1 [15] , the fact that PNGase F treatment caused a clear band shift to total Leg1 protein from the serum suggests that both Leg1a and Leg1b are modified by N-glycosylation . To confirm this hypothesis , leg1a and leg1b were cloned into the expression vector PCS2+ , and the obtained plasmids were used to transfect the human liver cancer cell line HepG2 . PNGase F treatment caused a band shift to both the expressed Leg1a and Leg1b in HepG2 ( Fig 5C ) . To determine which amino acid residue is glycosylated in Leg1a and Leg1b , we used a web-based platform , NetNGly ( http://www . cbs . dtu . dk/services/NetNGlyc/ ) , to predict the site of modification ( s ) based on the Asn-Xaa-Ser/Thr motif [43] . The prediction showed that the 70th asparagine ( N70 ) was a putative glycosylation site for both Leg1a and Leg1b ( Fig 5D and 5E ) . For Leg1a , N298 was also predicted to be a candidate site for glycosylation ( Fig 5D ) . We then mutated N70 and N298 in Leg1a to alanine ( A ) to obtain the leg1aN70A and leg1aN298A plasmids . We transfected the leg1a WT plasmid and the two leg1aN70A and leg1aN298A mutant plasmids into the HepG2 cell line , respectively , and performed western blot analysis of Leg1 in the extracted total protein at 24 hours post transfection . The result showed that leg1aN70A produced a product with a mobility like that of Leg1a treated with PNGse F , whereas leg1aN298A produced a product with a mobility like that by the leg1a WT plasmid ( Fig 5F ) . We also mutated the N70 to A70 in Leg1b and found that Leg1bN70A was no longer sensitive to PNGase F treatment and exhibited a mobility like that of Leg1b treated with PNGase F ( Fig 5G ) . In addition , we injected leg1aN70A and leg1bN70A mRNA and their respective WT control mRNA into zebrafish embryos at the one-cell stage and extracted total protein at 9 hours post injection . Western blot analysis of the protein samples showed that both Leg1aN70A and Leg1bN70A exhibited a mobility like that of Leg1a or Leg1b treated with PNGase F ( Fig 5H ) . All of these results demonstrated that N70 was the only N-linked glycosylation site for both Leg1a and Leg1b . Glycosylation in secretory proteins often facilitates the proper folding of the protein so that the protein can be licensed to be transported to the Golgi apparatus [44] . To assess the secretory ability of Leg1aN70A and Leg1bN70A , WT leg1a , WT leg1b , leg1aN70A or leg1bN70A plasmid was each co-transfected with HA-tagged rnasel1 plasmid into HepG2 cells . Rnasel1 ( NCBI accession no . AI476973 ) encodes a known secretory protein Rnasel1 [45] and was used as a control here . Total proteins were extracted from the culture medium and the cell pellet , respectively . Western blot analysis of the protein samples showed that Leg1a , Leg1b , Leg1aN70A and Leg1bN70A were all detected in the cell pellet fraction ( Fig 6A ) . Leg1a , Leg1b and Leg1aN70A were also detected in the culture medium fraction ( Fig 6A ) , whereas no Leg1bN70A was detected ( Fig 6A ) . Meanwhile , we noticed that the secretion of HA-Rnase1l in the cells expressing Leg1bN70A was also greatly reduced ( Fig 6A ) . To determine where the un-secreted Leg1bN70A was located in the protein trafficking route , we co-immunostained Leg1 with ER and Golgi markers , respectively . Consistent with western blot analysis , Leg1a and Leg1aN70A were secreted normally ( Fig 6B ) , as was the WT Leg1b , which was hardly detectable in the leg1b plasmid-transfected cells ( Fig 6C and 6D , upper panels ) . In contrast , Leg1bN70A nicely co-localized with the cis-Golgi indicated by cis- and medial-Golgi marker Giantin ( Fig 6C , lower panels ) but not with the ER marker PDI [46 , 47] ( Fig 6D , lower panels ) . Strikingly , cells transfected with the leg1bN70A plasmid appeared to harbor more cis-Golgi components ( revealed by Giantin staining ) than those in the WT leg1b-transfected cells ( Fig 6C ) , indicating that the Leg1bN70A mutant protein is retained in the cis-Golgi apparatus , which caused a traffic jam in the cells such that the secretion of Rnase1l was also severely blocked in these cells ( Fig 6A ) . The fact that accumulation of Leg1bN70Amutant protein in the cis-Golgi but not in the ER might explain why we did not observe an activation of the markers ( including Bip , Chop , and p-eIF2a ) for the ER-stress response either in the cultured cells ( S7A Fig ) or in the maternal-zygotic leg1azju1 mutants ( S7B and S7C Fig ) . Next , we tested whether N70 glycosylation in Leg1a is required to promote liver development by injecting leg1a andleg1aN70A mRNA , respectively , into the maternal-zygotic leg1azju1 mutant embryos at the one-cell stage ( S8 Fig ) . These injected embryos were briefly treated with UV25 . WISH analysis using the fabp10a probe showed that leg1aN70A mRNA injection failed to rescue the mutant liver development ( Fig 7A ) . In fact , Leg1aN70A was greatly compromised in promoting Erk phosphorylation under UV25 ( Fig 7B ) . FGFis a key effector of the Erk signaling pathway . We wanted to determine whether Leg1 interacts with the FGF receptor ( FGFR ) to activate the phosphorylation of Erk . Extracted serum protein containing Leg1a and Leg1b ( total Leg1 ) was incubated with human 293T cells transfected with a plasmid expressing FLAG-tagged FGFR3 . Co-immunoprecipitation ( Co-IP ) analysis showed that Leg1 interacted with FGFR3 ( Fig 7C ) . To determine whether N70-glycosylation is necessary for Leg1 to bind to FGFR3 , we treated total serum proteins ( containing both Leg1a and Leg1b ) with PNGase F to get a mixture of Leg1 and de-glycosylated Leg1 under the undenaturized condition ( Fig 7C , left panels ) . The mixture of Leg1 plus de-glycosylated Leg1 was incubated with 293T cells expressing FGFR3 . Co-IP analysis showed that only Leg1 but not the de-glycosylated Leg1 interacted with FGFR3 ( Fig 7C , right panels ) . We also over-expressed Leg1a and Leg1aN70A in 293T cells ( Fig 7D , left panels ) and harvested the culture medium containing Leg1a or Leg1aN70A to incubate with 293T cells overexpressing FGFR3 , respectively . Co-IP showed that Leg1a but not Leg1aN70A interacted with FGFR3 ( Fig 7D , right panels ) . In zebrafish , FGFR3 was co-immunoprecipitated by the Leg1 antibody when Leg1a and FGFR3 were co-expressed by their mRNA co-injection ( Fig 7E ) . The zebrafish transgenic line Tg ( hsp70:dnfgfr1-gfp ) expresses the dominant-negative Fgfr1 ( dn-Fgfr1 ) by the hsp70heakshock promoter[48] . The expressed dn-Fgfr1 , whose tyrosine kinase domain is replaced by GFP ( as a reporter of the transgenic embryos ) , can form heterodimer with all FGFR subtypes so that to block the FGF signaling . When this line was treated with UV25 only we found that the level of p-Erk was increased ( S9 Fig ) . However , when the embryos were heat-shocked at 22 hpf to express dn-Fgfr1 followed by treatment with UV25 at 24 hpf , the effect of UV25 on activation of p-Erkin the GFP+ embryos was down-regulated to a similar level to that observed in the UV25 untreated GFP+ embryos . This result further suggests that the activation of Erk by UV25-treatment is through the FGF pathway , Considering the importance of the liver for a living organism and the viable and fertile nature of theleg1a mutant , we wondered whether the small liver in the leg1a mutant would be recovered to normal during later growing stages . We treated WT and maternal-zygotic leg1azju1 mutant with UV25 at 24hpf , and check the liver size at 3 . 5dpf and 10dpf . While , as expected , almost all the maternal-zygotic leg1azju1 mutant displayed a small liver compared to the WT at 3 . 5dpf , the liver sizes in the mutants were recovered to normal at 10dpf ( Fig 8A ) . However , the maternal-zygotic leg1azju1 mutant exhibited a lower survival rate ~32% ( 28/87 ) when compared to 66% ( 50/76 ) for the WT counted at 10 dpf . Examining the expression of leg1b , the homolog of leg1a , in the leg1azju1 mutant revealed that the levels of leg1b transcripts were up-regulated both in zygotic homozygous mutant at 4 dpf ( Fig 8B ) and maternal-zygotic mutant at 3- , 5- , and 7-dpf ( Fig 8C ) . WISH using the leg1 probe also showed that the total leg1 transcripts in the maternal-zygotic leg1azju1 mutant was enriched in the liver ( S10B Fig ) . These data suggest that the compensatory mechanism[49] is activated in the leg1azju1 mutant to support the liver development at the later stages . At the adult stage , although the liver to body ratio of the leg1azju1 mutant fish did not show significant difference to that of the WT fish ( Fig 8D ) the leg1azju1 mutant fish exhibited a shorter stature and higher mortality compared to the WT fish ( Fig 8E and 8F ) , suggesting that the Leg1a anti-stress pathway also functions in the adult fish and that theLeg1bonly partially compensates for the function of Leg1a .
In addition to the precise spatial and temporal control of genetic programs instructing oganogenesis , successful completion of organogenesis also relies on the maintenance of an optimal environment through the elimination or neutralization of the stress-induced harmful reagents , and how this is achieved is of tremendous interest in the field of developmental biology [50] . Although undergoing external embryogenesis , teleost fish harbor a robust genetic program dictating liver development as long as any environmental change , including temperature or natural UV irradiation , is not detrimental . It is therefore of interest to explore the mechanism ( s ) behind this phenomenon . We showed that Leg1 plays a unique role in protecting liver development under different stress conditions by serving as a secretory signaling molecule/modulator to activate the Erk pathway . This finding may explain the adaption of teleost fish in coping with environmental changes . The process of liver organogenesis is governed by key transcription factors ( e . g . , HNF , GATA , Prox and Hhex ) and signaling molecules ( e . g . FGF , Bmp and Wnt ) [1–3 , 10] . Meanwhile , each of these stages has to deal with the oxidative stress constantly imposed intrinsically or externally . In zebrafish , hepatoblasts are specified at around 24 hpf and start to form the liver primordium at around 30 hpf . Both FGF and Bmp play crucial roles during this period [10 , 31 , 32] . In general , FGF acts through the FGFR-RAS-ERK signaling pathway [51] , Bmp through activation of Smad1/5/8 phosphorylation [52] and Wnt2bb through the β-catenin-TCF pathway to control organ/tissue development , respectively [53] . It is envisaged that molecules mediating anti-oxidative stress during liver organogenesis might act as a tuner of the pathways controlling cell proliferation or elimination . Based on the facts that 1 ) Leg1a expression is enriched in the yolk syncytial layer between 24–48 hpf ( S10A Fig ) and this layer is directly exposed to external stress such as UV irradiation or low level of oxygen , 2 ) Leg1a expression is obviously enriched in the embryonic liver at 48 hpf , 3 ) Leg1a is a secretory protein , 4 ) the leg1azju1 maternal-zygotic mutant exhibited a small liver only under the cold season , UV irradiation , high temperature , or H2O2 treatment , and 5 ) the small liver phenotype was rescued by the antioxidant chemicals DPI and APO , we conclude that Leg1a defines a novel anti-stress pathway to protect the liver development . Besides , we noticed that the leg1azju1 maternal-zygotic adults displayed a shortened body length and reduced survival rate , suggesting that the Leg1-meidated anti-stress pathway is also necessary for wellbeing of an adult fish . Then , the question is how there is a season in a fish facility which is maintained at relatively constant temperature throughout the year ? Since the small liver exhibited by the maternal-zygotic leg1azju1 mutant is ROS-dependent we speculate that the difference in the oxygen content in the fish water in different seasons might be the cause of the stress-related phenotype although the temperature is maintained in the facility . We know that the oxygen content in the water is related to atmospheric pressure and that the atmospheric pressure is higher in the cold seasons and lower in the warm seasons . We checked the weather record between Dec 30 , 2013 to Jan 30 , 2015 in Hangzhou and plotted the liver size against the record of atmospheric pressure . The small liver phenotype in leg1a mutant nicely correlates with high atmospheric pressure in the cold seasons ( S1D Fig ) . However , we cannot exclude other possibilities at this moment . Utilizing specific morpholinos ( MOs ) , we previously showed that leg1a is required for liver bud outgrowth[15] . However , zygotic leg1azju1 mutants do not show this phenotype , and maternal-zygotic leg1azju1 mutant phenotype is milder than the phenotype generated by the leg1a-MO . The discrepancy between the phenotype caused by MO-injection and the phenotype exhibited by a loss-of-function mutant is indeed a concern in the zebrafish community . The explanations for the discrepancy observed could be: 1 ) previous used leg1-MO might have yielded an off-target effect on the liver development in the morphants , this fits with the observation that leg1a or leg1b mRNA or even combination of leg1a and leg1b mRNA only partially rescued the morphant small liver phenotype[15]; 2 ) injecting morpholino itself works as a stress cue to induce the small liver phenotype when Leg1 is knocked down; and 3 ) since the leg1b gene is still intact in the leg1azju1 mutant , the mild phenotype exhibited by the maternal-zygotic leg1azju1 mutant might be due to the functional compensation by Leg1b[49] . To narrow down the possibilities , we tried to get the leg1a and leg1b double knockout mutant , however , failed in obtaining such double mutant . We also injected standard control morpholino ( ST-MO , derived from human β-globin antisense morpholino ) into the leg1azju1 mutant embryos and found that ST-MO did not enhance the leg1azju1 mutant phenotype ( S11 Fig ) . We then compared the expression of leg1b in the WT and leg1azju1 and found that the expression of leg1b is elevated in the leg1azju1 mutant embryos . These data suggest that the expression of leg1b is mobilized to compensate , at least partially , for the loss of function of Leg1a in the leg1azju1 mutant . Mechanistically , it appears that Leg1a does not signal through the Bmp pathway because Leg1a over-expression does not promote the phosphorylation of Smad1/5/8 as done by the over-expression of Bmp at 18 or 24 hpf . However , Leg1a over-expression does promote the phosphorylation of Erk upon UV25 treatment . Furthermore , UV-treatment caused up-regulation of the phosphorylation of Erk in WT but not in the maternal-zygotic leg1azju1 mutant . In addition , as revealed by immunostaining , it appeared that more p-Erk cells were in the WT endoderm than that in the maternal-zygotic leg1azju1 mutant after UV25 treatment ( S12 Fig ) . The intriguing question is why Leg1a promotes Erk phosphorylation after UV exposure . We speculate that it maybe because UV causes certain modification or conformation change to Leg1a that facilitates the interaction between Leg1a and Fgfr3 to promote Erk phosphorylation . Nevertheless , these data suggest that Leg1a signals through the Erk pathway . Since FGF is a key effector of the Erk pathway and is essential for liver development , our data suggest that there might be a crosstalk between the FGF and Leg1-meidated anti-stress signaling pathways . Therefore , it is of great interest to determine how Leg1 promotes Erk phosphorylation in the future . For example , being a secretory protein , does Leg1a have its own receptor or shares the FGF receptor to mediate its activity ? If Leg1a does share the FGF receptor with FGF , then which type of receptor do they share ? Or does Leg1a simply serve as an agonist to facilitate the binding of FGF to its receptor ? Leg1 is an evolutionally conserved protein across the vertebrates[15] . A recent report showed that Leg1 homologs in monotreme is highly expressed in monotreme milk and appears to be modified by N-glycosylation[17] . This implies that the tissue expression specificity and function of Leg1 might vary among different animal species . Here we showed that zebrafish Leg1a is glycosylated at N70 . Although this glycosylation modification is not essential for the secretion of Leg1a , it is important for Leg1a in the promotion of liver development , for the phosphorylation of Erk and interaction with FGFR3 . All available data have suggest that Leg1a is a novel signaling molecule/modulator , which has urged us to identify more downstream signaling molecules involved in this pathway , which may ultimately reveal the importance of this pathway in the evolution of vertebrates .
All animal procedures were performed in full accordance to the requirement by ‘Regulation for the Use of Experimental Animals in Zhejiang Province’ . This work is specifically approved by the Animal Ethics Committee in the School of Medicine , Zhejiang University ( ETHICS CODE Permit NO . ZJU2011-1-11-009Y , issued by the Animal Ethics Committee in the School of Medicine , Zhejiang University ) . The zebrafish ( Danio rerio ) AB strain was used as WT in this study . To generate the leg1a mutant , we constructed a TALEN vector against the first exon of the leg1a gene ( Fig 1A ) according to the “Unit Assembly” protocol[19] . The TALEN mRNA was synthesized using the SP6 mMESSAGEmMACHINE Kit ( Ambion ) and was injected into the WT embryos at one-cell stage . These embryos were bred to the adulthood as founders to mate with a WT fish . Eight embryos from each cross were genotyped using the primer pair leg1a 4244 Fw ( CTTACAAGTTACAGCAGCTCC ) and legg1a 7748 Rv ( CACAACGGACCAGTACATCG ) followed by the second primer pair TALEN ID fw ( CTCCCAGAGGATGACCATGT ) and TALEN ID Rv ( ACTCCAGAGCGGATTCTCCT ) to identify leg1a mutants , and the rest embryos were bred to adulthood for identification of individuals carrying the mutation . The Tg ( hsp70:dnfgfr1-gfp ) and Tg ( hsp70l:bmp2b ) fish lines were obtained from Dr Feng Liu . Fish was raised and maintained in the fish facility ( Ai-Sheng Zebrafish Facility Manufacturer Company , Beijing , China ) in Zhejiang University according to the standard procedure . HepG2 cells were grown in the DMEM medium ( high glucose , GIBCO ) , supplemented with 10% newborn calf serum ( NBCS , GIBCO ) . Plasmids were transfected into cells mediated with lipofectamine 2000 ( InVitrogen ) according to the manufacturer’s instruction . Total protein was extracted 24 hours after transfection and was subjected to western blotting analysis . The ORF region of leg1and erk was cloned into PCS2+ vector . All leg1and erk point mutations were generated by site-directed mutagenesis . The primers for leg1 mutant used in the PCR reaction were designed by the webtoolPrimerX ( http://bioinformatics . org/primerx/index . htm ) . The sequences of primers are listed in S1 Table . All primers used for erk point mutation was designed as previously described[35] . mRNAs were obtained via in vitro transcription using the mMessagemMachine ( Ambion ) according to the manufacturer’s instruction . Cells were seeded in glass slide when the optimal cell density was achieved and fixed by 3% PFA for half an hour at 4°C . The cells were washed twice with 50mM NH4Cl and three times with PBS and were then penetrated with PBST ( PBS+0 . 1% Triton X 100 ) for 15min and blocked with blocking buffer ( 5% goat serum , 5% fetal bovine serum and 2% bovine serum albumin ) for 30min , sequentially . Cells were finally incubated with corresponding primary antibody and then Alexa Fluor conjugated second antibody . The samples were visualized under a confocal microscope . Leg1 antibody [54] and BHMT antibody [55] was generated as described . PDI antibody ( Sigma , P7496 ) , Giantin antibody ( Abcam , ab24586 ) , PH3 antibody ( Santa Cruz , SC-8656-R ) , Actin antibody ( Huabio , R1207-1 ) , GAPDH antibody ( Huabio , M1211-1 ) , p-Erk antibody ( Cell Signalling Technology , 9101 ) , total Erk antibody ( Cell Signalling Technology , 4695 ) , pSmad1/5/8 ( Cell Signalling Technology , 9511 ) antibody were purchased from the companies as indicated . WISH was performed as previously described [27] . prox1 , hhex , fabp10a ( liver fatty acid binding protein 10 ) , insulin and trypsin , fabp2 ( intestinal fatty acid binding protein 2 ) were cloned into expression vectors , respectively [27 , 31] . Corresponding probes were synthesized via in vitro transcription and were labeled with digoxigenin ( DIG , Roche , Diagnostics ) . Liver size was measured as previously described [54] . Briefly , liver was marked out after WISH suing the fabp10a probe , and imaged by Nikon AZ100 from left lateral view after aligning two eyes of the embryo vertically . The fabp10a signal area in each image was calculated by Nikon image system ( NIS-elements D v3 . 0 ) and used as the index of the liver size . PNGase F ( NEB , P0704 ) was used to cleave N-linked glycosylation , and a combination of Endo-alpha-α-Acetylgalactosaminidase ( NEB , P0733 ) and neuraminidase ( NEB , P0720 ) was used to cleave O-linked glycosylation . All the enzyme treatment was performed according to the manufacturer’s instruction . For either fish embryo or cultured cells , total protein was extracted using an extraction buffer ( 63mM Tris-HCl , PH6 . 8 , 10% glycerol , 5% β-Mercaptoethanol , 3 . 5% SDS ) containing 1X Complete ( Roche , 11873580001 ) . Western blotting was performed as described previously [27] using corresponding antibodies as indicated in the figures . Actin antibody ( Huabio , R1207-1 ) , GAPDH antibody ( Huabio , M1211-1 ) , p-Erk antibody ( Cell Signalling Technology , 9101 ) , total Erk antibody ( Cell Signalling Technology , 4695 ) , pSmad1/5/8 antibody ( Cell Signalling Technology , 9511 ) , Bip antibody ( Sigma , G9043 ) , Chop antibody ( Sigma , G6916 ) , phosphorylated eIF2a ( p-eIF2a ) antibody ( Cell Signaling Technology , 9721S ) , and total eIF2a antibody ( Cell Signaling Technology , 9720S ) , and Flag antibody ( Sigma , F1804 ) were purchased from the companies as indicated . Signal intensity of a desired band was calculated by ImageJ software ( v . 1 . 48 ) . Embryos were treated with different dosage of UV energy supplied by Ultraviolet Crosslinker ( UVP , CL-1000 ) at 24 hpf and then allowed to grow in the egg water . For H2O2 treatment , 24 hpf old embryos were treated with different concentration of H2O2 for half an hour . For APO and DPI treatment , embryos were incubated with 0 . 5 μM APO ( Sigma , W508454 ) or 10μM DPI ( Sigma , D2926 ) for 1 hour at 23 hpf , followed by UV25 treatment , and then allowed to grow in normal egg water . Embryos injected with TRE-caErk plasmid were treated with Dox at 24hpf , and replaced with fresh egg water at 30hpf ( 6hpt ) and 33hpf ( 9hpf ) , respectively . ROS content measurement was performed as described previously [56] with some modifications . Briefly , embryos were sunk in 100μl of 10μM DCFH-DA ( Beyotime , S0033 ) solution for one hour prior to UV25 treatment . For each sample , embryos were divided into four groups ( containing three embryos in each group ) and placed in a 96-well plate . After UV25 treatment the fluorescence signal was measured at a 10 min interval for one hour on a Synergy H1 Reader ( Biotek ) ( excitation 485 nm , emission 560 nm ) . For studying the interaction between Leg1 and FGFR3 in the cell culture system , Leg1 sample was prepared either by diluting 30μl of the fish serum with 1000μl of serum-free DMEM media ( Gbico ) or by transfecting 293T cells with the leg1 plasmid ( cloned into the PCS2+ expression vector ) and collecting the culture media 30 hrs after transfection . The Leg1 samples were incubated with the 293T cells expressing FGFR3 ( cells transfected with the FGFR3 plasmid in the PLX304 expression vector , the vector is provided by Dr Bing Zhao ) at 4°C for one hour . After incubation , the cells were washed with PBS for three times , and lysed with NP40 lysis buffer ( 50mM Tris-Hcl , PH 8 . 0 , 150mM NaCl , 1% NP40 , 2mM EDTA ) . For studying the interaction between Leg1 and FGFR3 in the embryos , embryos injected with leg1a and fgfr3 mRNA at one cell stage were harvested at 7hpf and lysed with NP40 lysis buffer . All lysates were incubated with Leg1 antibody or Flag antibody at 4°C overnight , followed by incubation with Protein A/G argrose beads ( beyotime , Cat . No . P2012 ) for further 2 hrs . The beads were washed with cold PBS for three times and eluted by 100mM PH2 . 2 glycine . The elution was subjected to western blot analysis . More than 50 embryos were pooled for total RNA extraction . Reverse transcription was performed by SuperScript II Reverse Transcriptase ( Invitrogen , 18064–014 ) according to the manufacturer’s protocol . The transcribed cDNA was used as the template in qPCR with SYBR Green Master Mix ( Vazyme ) . The CFX96 real time system ( Bio-Rad ) was used to obtain the threshold cycle ( Ct ) value , and the relative expression of each gene was determined after being normalized to the actin gene . Primer pairs used are listed in S2 Table . In considering of relative small sizes of samples with skewed phenotype distribution among individuals in this study , the conventional statistical analysis by showing mean and standard error/derivation is apparently not suitable for presenting the liver size measurement data . Instead , quartiles are more intuitive in presenting data with relative small sample size with skewed distributions [57] . Therefore , we used the quartile boxplot to present our data [58] . The box plot was drawn by ggplot2 [59] . Survival ratio statistical analyses were carried by Chi-squared test . Other statistical analyses were performed with the Student’s T-test . * , p<0 . 05 , ** , p<0 . 01 , *** , p<0 . 001 , N . S , no significant difference .
|
Although being challenged by stresses from their living environment during embryogenesis , teleost fish harbor a robust genetic program dictating liver development as long as any environmental change , including temperature or natural UV irradiation , is not detrimental . It is therefore of interest to explore the mechanism ( s ) behind this phenomenon . We showed that Liver-enriched gene 1 ( Leg1 ) plays a unique role in protecting liver development under different stress conditions by serving as a secretory signaling molecule/modulator that binds to FGF receptor and activates the Erk signaling pathway . This finding may explain the adaption of teleost fish in coping with environmental changes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"protein",
"transport",
"cell",
"processes",
"vertebrates",
"animals",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"model",
"organisms",
"organism",
"development",
"glycosylation",
"embryos",
"extraction",
"techniques",
"research",
"and",
"analysis",
"methods",
"erk",
"signaling",
"cascade",
"embryology",
"protein",
"extraction",
"fishes",
"proteins",
"liver",
"development",
"organogenesis",
"biochemistry",
"signal",
"transduction",
"zebrafish",
"protein",
"secretion",
"cell",
"biology",
"post-translational",
"modification",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"cell",
"signaling",
"glycobiology",
"organisms",
"signaling",
"cascades"
] |
2016
|
Liver-Enriched Gene 1, a Glycosylated Secretory Protein, Binds to FGFR and Mediates an Anti-stress Pathway to Protect Liver Development in Zebrafish
|
Analyses investigating low frequency variants have the potential for explaining additional genetic heritability of many complex human traits . However , the natural frequencies of rare variation between human populations strongly confound genetic analyses . We have applied a novel collapsing method to identify biological features with low frequency variant burden differences in thirteen populations sequenced by the 1000 Genomes Project . Our flexible collapsing tool utilizes expert biological knowledge from multiple publicly available database sources to direct feature selection . Variants were collapsed according to genetically driven features , such as evolutionary conserved regions , regulatory regions genes , and pathways . We have conducted an extensive comparison of low frequency variant burden differences ( MAF<0 . 03 ) between populations from 1000 Genomes Project Phase I data . We found that on average 26 . 87% of gene bins , 35 . 47% of intergenic bins , 42 . 85% of pathway bins , 14 . 86% of ORegAnno regulatory bins , and 5 . 97% of evolutionary conserved regions show statistically significant differences in low frequency variant burden across populations from the 1000 Genomes Project . The proportion of bins with significant differences in low frequency burden depends on the ancestral similarity of the two populations compared and types of features tested . Even closely related populations had notable differences in low frequency burden , but fewer differences than populations from different continents . Furthermore , conserved or functionally relevant regions had fewer significant differences in low frequency burden than regions under less evolutionary constraint . This degree of low frequency variant differentiation across diverse populations and feature elements highlights the critical importance of considering population stratification in the new era of DNA sequencing and low frequency variant genomic analyses .
In the field of human genetics research , there has been increasing interest in the role of low frequency variation in complex human disease ( defined in this text as variants with a minor allele frequency between 0 . 5%–3% ) . This is in many ways a response to changing technology , but more importantly a response to the inability to completely explain heritability in common complex diseases and recognition of the true multifactorial mechanisms of genetic inheritance [1] . Since low frequency variants are likely essential in understanding the etiology of common , complex traits , it is critical to elucidate the genetic architecture and population substructure of low frequency variants for future work in this field . Factors such as rapid population growth and weak purifying selection have allowed ancestral populations to accumulate an excess of low frequency variants across the genome . This affects genomic analyses in two ways: proportion of deleterious versus neutral variation expected in low frequency variants and population stratification . It has been suggested that slightly deleterious single nucleotide variants ( SNVs ) subjected to weak purifying selection are major players in common disease susceptibility [2] , [3] . For example , Nelson et al . found that in 202 drug target genes , 2/3 of the low frequency variants were nonsynonymous mutations . This is a much higher ratio than found for common variants , and reflect the expected proportion given random mutation and degenerate coding . This ratio also suggests low frequency variants are only weakly filtered by selection [2] , [4] . In addition , low frequency variants represent a considerable proportion of the genome due to recent explosive population growth [3] . Gorlov estimates up to 60% of SNVs in the genome are variants with an allele frequency <5% [5] . Since the allele frequency distribution is skewed towards more low frequency variants , a higher number of low frequency deleterious variants are expected . Subsequently , low frequency variants appear to be enriched for functional variation , including protein coding changes and altered function [6] . Further , low frequency variants exhibit extreme population stratification . Demonstrating the magnitude of low frequency population stratification between two populations , Tennessen et al . identified more than 500 , 000 SNVs using 15 , 585 protein-coding genes from 2 , 440 individuals . Of these SNVs , 86% had a MAF<0 . 5% and 82% were population specific between European Americans and African Americans [3] . Low frequency allele sharing between populations on the same continent can be between 70% and 80% . In contrast , low frequency allele sharing between populations on different continents can be lower than 30% and variants are often unique to a single population . This extreme geographic stratification can lead to higher false positives and difficulty in replicating associations across genetic studies when not considered as part of the experimental design for low frequency SNV analyses [6] . To study the “landscape” of low frequency variant stratification across populations , we grouped low frequency variants across pertinent genome-wide biological features in a series of pairwise population comparisons across multiple ancestries . We define the boundaries of grouping by features , which consist of genomic regions ( one or many ) that belong to a genomic category , for example , a gene or a set of genes in a pathway . Methods that aggregate variants have been shown to be much more powerful than single-variant association testing for low frequency variants [7]–[10] , and thus are reliable to detect population stratification . Our collapsing method , BioBin , provides the opportunity to cast a broader net and uncover stratification across meaningful elements such as genes , pathways , and evolutionary conserved regions by aggregating low frequency variants based on expert biological knowledge . Herein we have applied BioBin to individuals from 1000 Genomes Project Phase I data; we defined “cases” and “controls” randomly between exhaustive pairwise population comparisons . Our goal was to identify features across the genome with differences in low frequency burden between populations; specifically , to look for aggregate differences in low frequency variation between populations , not to detect individual population-specific variants . We show that BioBin is effective in identifying differences in low frequency variant burden centered on biological criteria and highlights the considerable differences in low frequency variants across ancestry groups . These results further emphasize the critical importance of considering low frequency population substructure in future rare and low frequency variant analyses .
We applied BioBin to whole-genome population data using the 1000 Genomes Project Phase I data . The populations , sample sizes , and total number of loci , variants , low frequency variants , and private variants are listed in Table 1 . Although the Iberian population ( IBS ) is listed in Table 1 , this population was not used in the analyses presented in this paper . There was not a sufficient sample size to meet our low frequency criteria ( N = 14 , MAF cutoff = 0 . 03 ) . In addition to the differences in overall magnitude of variation seen in Table 1 between these population groups , there were also differences in the distribution of this variation . In Figure 1 , we present an allele frequency density distribution plot of all autosomal chromosomes for all 13 populations . African descent populations have the highest density of low frequency variation . Others have found a similar trend genome-wide [11] . In general , the African ancestral populations not only have more variants overall than other ancestral groups ( see Table 1 ) , these populations also have a higher distribution of low frequency variants than other ancestral groups ( see Figure 1 ) . Although low coverage next generation sequence data is prone to errors , we found no evidence that sequence technology led to differential bias in a way that could explain the trends found in this paper ( Text S1 , Table S1 , Figures S1 , S2 ) . We investigated sample-relatedness with respect to common and low frequency variants using both identity-by-descent ( IBD ) and identity-by-state ( IBS ) estimations , and in each analysis , we found evidence of increased relatedness in ASW ( African ancestry , USA ) , CHB ( Han Chinese Beijing , China ) , CHS ( Han Chinese Shanghai , China ) , CLM ( Medellin , Columbia ) , GBR ( England and Scotland ) , JPT ( Japan ) , LWK ( Luhya , Kenya ) , and MXL ( Mexican Ancestry , California ) . We performed iterative IBD calculations in plink to eliminate related individuals from continental groups . Seventy-five individuals of 1080 total individuals were parsimoniously removed to achieve a pi_hat< = 0 . 3 in each continental population . The remaining 1 , 005 individuals were used for the binning analyses presented in this paper . An alternate allele sharing method described by Abecasis et al . uses IBS rather than IBD to review allele sharing [12] , [13] . In the case of low frequency or rare variants , IBS approximates IBD . Figure 2 shows within population IBS for all 13 populations for variants with a MAF<3% , where each point represents a pairwise IBS calculation within the same population ( i . e . YRI-YRI but not YRI-CEU ) . In Figure 2A , the pairs with average IBS calculations that fall outside of the cluster are cryptically related individuals with increased allele sharing . Figure 2B shows the IBS calculations after removing 75 individuals with cryptic relatedness . Complete details of these and additional sample-relatedness analyses are available in Text S2 , Figure S3 , Figure S4 , and Figure S5 . Knowledge of population substructure in low frequency variants is critical for genomic studies . We applied BioBin to test for low frequency ( MAF≤0 . 03 ) variant burden differences between 13 populations from the 1000 Genomes Project across different genomic features: genes ( intronic and exonic variants , filtered nonsynonymous and predicted damaging variants ) , intergenic regions , ORegAnno annotated regulatory regions , pathways , pathway-exons , evolutionary conserved regions , and regions considered to be under natural selection . Results are shown in Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure S6 , and Figure S7 . In each matrix plot , we have indicated the proportion of significant bins ( after Bonferroni correction ) out of the total number of bins generated between two populations . The color intensity represents the proportion of total bins that were significant [0 , 1] . Overall , there are large differences across populations with regard to low frequency variant burden and the distribution of low frequency variants is not random across the genome . The magnitude of stratification corresponds to the mutational landscape of the region . We chose NCBI Entrez to provide the boundaries for gene regions and created a custom role file of intron and exon boundaries using data provided from UCSC Genome Browser [14] . In Figure 3 , the top matrix corresponds to bins created using gene-exon boundaries , the middle matrix corresponds to bins created using gene-intron boundaries , and the bottom matrix corresponds to bins created using regions between genes ( intergenic ) . The values and color intensity within each block represent the proportion of significant bins after Bonferroni correction out of the total number of bins generated between two populations . The coding regions show a trend of a lower proportion of significant bins with low frequency variant burden differences than either the intron or intergenic bins . For example , in the CEU ( Northern/Western European Ancestry , USA ) −YRI ( Yoruba African ) comparison , approximately 44% of the gene-exon bins had significant differences in low frequency variant burden . In contrast , the noncoding region bins , gene-introns and intergenic bins had 66% and 70% of bins with significant differences in low frequency variant burden . The coding regions appear to be under more constraint across populations than noncoding regions . Comparing only the noncoding regions , introns tend to have slightly fewer variation differences than intergenic bins , most likely because introns are by default nearest neighbors to the selective pressures on coding regions . We filtered the gene-exon bins using annotations from the Variant Effect Predictor Software ( VEP ) [15] . We created gene bins with only nonsynonymous variants and a second analysis using only predicted damaging variants annotated by SIFT or PolyPhen2 [15]–[17] . The results in Figure 4 indicate that these potentially functional and significant changes are even more conserved between populations than coding regions ( Figure 3A ) . We used ORegAnno ( Open Regulatory Annotation database ) to define regulatory region boundaries for the bin analysis . The top matrix of Figure 5 shows the 78 population comparisons for the ORegAnno regulatory feature analysis . In comparison to Figure 3 , the annotated regulatory regions have fewer significant bins . For example , in gene-exon analysis shown in Figure 3 , approximately 44% of the ASW-CHB gene-exon bins contained significant differences in low frequency burden . However , in Figure 5 , only 28% of the ASW-CHB annotated regulatory bins contained significant differences in low frequency burden . This trend is consistent across the matrix of population comparisons; regulatory regions have fewer significant bins than the coding or noncoding features of the same population comparison . Several biological pathway and group sources from LOKI ( the Library of Knowledge Integration , which is described in detail in the methods ) were used to generate low frequency variant bins; including , PFAM , KEGG , NetPath , PharmGKB , MINT , GO , dbSNP , Entrez , and Reactome . The Figure 5B shows the 78 population comparisons for the pathway group feature analysis . Of all of the feature analyses , pathway bins consistently show the highest proportion of significant differences in low frequency variant burden between populations . There are several potential explanations . First , since pathway bins are generally much larger than the other feature types , it is possible that large bins increase the false positive rate . Second , the same genes and regions can recur in multiple pathways . If the region has significant differences in low frequency variant burden , then each pathway or group containing that region will have a higher chance of having significant differences in low frequency variant burden . Following this logic , a pathway containing many genes has a higher chance of having at least one gene with extreme low frequency variant stratification . To compare only coding regions within a pathway , we filtered the pathway analysis to include only variants within exons . The proportions are reduced ( shown in Figure S6 ) but still higher than the gene-exon proportions shown in Figure 3A . PhastCons output downloaded from UCSC Genome Browser was used to derive evolutionary conserved feature boundaries for primates , mammals , and more than 40 species of vertebrates . Figure S7 shows the 78 population comparisons for the ECR feature analysis . Of all of the feature analyses , ECR bins had the smallest proportion of significant bins . More ancestrally similar populations tended to have negligible low frequency burden differences in these conserved segments . For example , approximately 7% of the ECR region bins ( vertebrate alignment ) were significantly different between FIN ( Finnish ) and JPT ( Japanese ) individuals . However , the significant number of bins between the two ancestrally similar GBR ( British ) and CEU individuals was less than 1% . To investigate regions of natural selection , we created a feature list using regions recently identified/confirmed by Grossman et al . with the Composite Multiple Signals algorithm on the 1000 Genomes Project data [18] . In addition , a publication by Barreiro et al . provided a list of specific genes with the strongest signatures of positive selection; i . e . genes that contained at least one nonsynonymous or 5′ UTR mutation with an FST value greater than 0 . 65 [19] . After lifting positions to build 37 , there were only 368 regions from the regions identified by Grossman et al . The results are shown in Figure 6 . The top plot corresponds to regions identified in African ancestry , the middle plot corresponds to regions identified in populations of Asian ancestry , and the bottom plot corresponds to regions identified specifically in populations of European ancestry . The trends in these three matrix plots are distinctly different from the trends shown in Figures 3–5 . The blocks of comparisons within a continental group ( shown in black boxes on each matrix plot ) still have very little color , which means that the low frequency variant burden between populations within a continental group is very similar . The main difference is the gain of intensity outside of the continental groups . For example , in Figure 6B ( regions identified in Asian populations ) , the European continental group and Spanish continental group mostly have proportions over 60% when compared to populations of Asian descent . In the same plot , the populations in the African group have proportions over 85% when compared to populations in the Asian group . In general , we found regions considered to be under natural selection unlikely to have significant differences in low frequency burden between ancestrally similar populations , and very likely to have significant differences in regions considered to be under natural selection between ancestrally distant populations ( see Figure 6 ) . Additional analyses were performed using regions identified in other publications and can be found in Text S3 , Table S2 , and Table S3 . Although low frequency variants are commonly assumed as independent ( in low linkage disequilibrium ( LD ) with other variants ) , there are rare haplotypes within related individuals and populations [20] . In Figure 7 , three pairwise population comparisons are shown . We investigated the top 10 ranked bins from the CEU-CHB ( A ) , CHB-YRI ( B ) , and CEU-YRI ( C ) coding and noncoding analyses for presence of LD ( r2>0 . 3 ) between two variants in the same bin . Figure 7 shows bins predominately filled with white-space indicating low to no pairwise LD between variants in those bins . In the top ten bins from these three analyses , rare haplotypes do not appear to be driving the significant differences seen in low frequency variant burden . Since the proportion of significant bins in the feature analyses is considerably higher for pathway bins than any other feature , we wanted to investigate the correlation between pathway p-value and bin size . We chose to assess the correlation between significance and several characteristics of the pathways using the pathway feature CEU-YRI population comparison . Figure S8 and Figure S9 show the correlations between six untransformed and transformed variables ( with outliers removed ) , where each pairwise correlation is significant ( p-value<0 . 05 ) . A bin was considered an outlier if the number of loci in the bin was more than 2 . 5 standard deviations from the mean transformed loci value . The most interesting correlations were the nonlinear correlations between the loci/variants/genomic coverage and p-values . Figure S9B is a higher magnification of the highlighted correlation in Figure S9A , specifically; we plotted the correlation between −log10 p-values and log10 variants . The lowess smoothing function is shown in red , and the function appears to change slope twice . From x = 1 to x = 3 , the slope is increasing with increasing number of variants . From x = 3 to x = 4 , the slope is near 0 . From x = 4 to x = 5 , the slope is increasing with increasing numbers of variants . When the log10-transformed value of the number of variants is less than 3 or greater than 4 , there appears to be a positive correlation between the number of variants in a bin and increasing significance of that bin . However , the data is not uniformly distributed and is sparse in those same areas . Therefore , the trends in the tails are most likely unreliable . We created boxplots describing certain characteristics from each data source . Figure S10 shows that specific sources ( i . e . KEGG ) consistently have larger bin characteristics ( number of loci , number of genes , coverage ( kb ) , etc . ) and also have much more significant bin p-values ( Figure S10B ) . It appears that certain sources might inherently have more significant groups by nature of the information that source provides , or because of the size of groups found in the source . From the matrix plots shown above , there is undoubtedly a functional component that influences the evolution of low frequency variants . However , from the correlations in the pathway analyses , it is also clear that larger bins can contain more stratification and thus more likely to have significance differences in low frequency burden . In more traditional case/control analyses , large bins are less likely to be significant because increasing binsize generally means more noise to mitigate the signal . However , in this study , when diverse populations are compared , larger bin sizes have more opportunity to capture population stratification . Figure S11 shows the relative number of loci across tested features and varying interregion parameters . Boxplots in Figure S11A represent each feature tested in the population comparison . The small inset figure shows the magnitude of difference between the numbers of loci in pathways ( peak ) versus other feature types . The main plot in Figure S11A shows the same information , but is limited to 2000 loci . In general , ECRs/exonic regions/nonsynonymous gene variants/ORegAnno annotated regions/predicted deleterious gene variants/UTRs are very small bins . Pathway bins have a broader distribution , but in general are much larger . For comparison , we varied the size of intergenic regions ( only noncoding regions ) between 10 kb and 200 kb , those results are shown in Figure S11B . We also split the entire genome ( including coding and noncoding regions ) by various windows between 10 kb–200 kb . Figure S11C represents a genome “average , ” and both Figure S11B and S11C can be used as comparison for feature tests . Figure S11B and S11C show increasing bin size as windows increase , the proportion of significant bins increase as window size increases as well ( see Figure S12 and S13 ) . For example , Figure S13 shows matrix plots from whole-genome “average” analyses ( A–G correspond to 10 kb , 25 kb , 50 kb , 75 kb , 100 kb , 150 kb , and 200 kb respectively ) . According to Figure S11A and S11C , exon bins from the original feature analysis are roughly comparable in size to 10 kb bins from the whole-genome “average” analysis . In the gene analysis results , approximately 43 . 9% of bins are significant after Bonferroni correction between CEU-YRI . Comparatively , the genome average between CEU-YRI for 10 kb bins is 57 . 64% . This supports the idea that coding regions are presumably more functional and perhaps more conserved than other regions in the genome of comparative size . According to Figure S11A and S11C , pathway bins from the original feature analysis are roughly comparable in size to 150 kb bins from the whole-genome “average” analysis . In the pathway analysis results , approximately 81 . 28% of bins are significant after Bonferroni correction between CEU-YRI . Comparatively , the genome average between CEU-YRI for 150 kb bins is 86 . 09% . The gap between pathway bins and “average” genome stratification given similar size is much smaller for pathways than it is for exons . This particular pathway analysis includes introns ( which typically have more variation than coding regions and larger bins are expected to collect more stratification . However , there are still fewer significant bins than expected on average .
Since the reference genome is predominantly of European ancestry [21]–[23] , populations with non-European ancestry generally have more variation with respect to the reference genome than those of European ancestry ( see Table 1 ) . Therefore , to interpret the results of this study , one might conclude that non-European populations have higher rates of sequencing error than European descent populations . However , in the most recent 1000 Genomes Project publication , the authors report an accuracy of individual genotype calls at heterozygous sites more than 99% for common SNVs and 95% for SNVs at a frequency of 0 . 5% . Furthermore , the authors found that variation in genotype accuracy was most related to sequencing depth and technical issues than population-level characteristics [11] . Therefore , neither the sequencing error nor the predominantly European reference genome adequately explains the trends seen in the genomic feature exploration ( see Text S1 , Table S1 , Figures S1 , S2 ) . Both sequence generation ( technology and/or site ) and population identity strongly contributes to underlying stratification in next-generation sequence data . After removing individuals with cryptic relatedness , 4 out of 13 Phase I populations were sequenced entirely using a single sequence technology ( CHB , CHS , JPT , and TSI ) . The other 9 populations had between 3–18 individuals or ∼5%–57% of the population sequenced on technologies other than Illumina ( ABI SOLID or LS454 ) . Note: all three of the Asian populations ( after removing individuals with cryptic relatedness ) were sequenced only with Illumina technologies . In our IBD analysis using variants with a minor allele frequency of 5% or greater and linkage disequilibrium r2< = 0 . 2 , we identified and eliminated 75 individuals of various population backgrounds . In addition to the previously documented relatedness in 1000 Genomes Project [http://www . 1000genomes . org/phase1-analysis-results-directory] , we also found additional cryptic relatedness seen in other work [24] , [25] . The differences are likely because we used continental groups ( not a single population or the entire 1080 individuals ) to identify cryptically related individuals and in our analysis that could include variants with fixed opposite frequencies and are overall common . This is infrequent in populations of the same continental group , but could be stratification introduced by different sequencing technologies . The major goal of this study was to investigate population stratification across multiple biological features . We created matrix plots to illustrate the proportion of significant bins in each comparison ( shown in Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure S6 , and Figure S7 ) . Our results show an interesting trend between functional regions of the genome and variant tolerance . Mutations appear to be less tolerated in functional regions . Similarly , ECRs , which are known to be conserved among species , are also the features least likely to have variation burden differences between two populations . There is some debate about selection and functional significance in these conserved regions , it is unknown what factors have the largest effect on mutation rates [26] , but it is possible that consistently low mutation rates in these features have generated conserved regions throughout evolution [27] . There are two potential explanations: 1 ) additional level of repair of DNA damage in transcriptional active regions by transcription coupled repair ( TCR ) , 2 ) approximately 3% of the genome is subject to negative selection , however it is estimated that functionally dense regions contain up to 20% of the sites under selection [26] , [27] . A number of the top results in each comparison have an interesting context , particularly in light of natural selection . Perhaps one of the most notable is SLC24A5 ( Ensembl ID:ENSG00000188467 ) , which is one of the top ten results in 19 out of 78 populations comparisons in the gene feature analysis . European specific selective sweeps estimated in the last 20 , 000 years suggest that SLC24A5 is key in skin pigmentation and Zebrafish with “golden” mutations exhibit melanosomal changes [28]–[30] . The presence of selection in particular populations due to environmental factors such as distance to the equator has led to the evolution and expansion of low frequency variants in some populations but not others . A second notable top result is DARC ( Ensembl ID:ENSG00000213088 ) , which encodes the Duffy antigen . The DARC gene bin was in the top ten results in 14 out of 78 population comparisons in the gene feature analysis . It has long been known that populations of African descent have increased diversity due to natural selection at this location , which prevents Plasmodium vivax infection . The top result from the regulatory region analysis was a region on chromosome 20 ( chr20:45395536–45396346 ) which was in the top ten bins in 24 out of 78 populations comparisons in the ORegAnno feature analysis . This region also overlaps ENCODE transcription factor binding sites in multiple cell lines , including: CTCF , POLR2A , NFYA , E2F1 , FOS , and more . It was also annotated as an insulator in multiple cell lines in ENCODE Chromatin State Segmentation analyses using Hidden Markov Models [14] , [31] . One last example , chr15 . 968 , contains variants in the genome location chr15:48400199–48412256 . This bin is one of the top ten bins in 17 out of 78 population comparisons in the intergenic analysis . The region covered by the chr15 . 968 bin is less than 1 kb upstream of SCL24A5 on chromosome 15 and overlaps with several transcription factor-binding sites ( including CTCF ) , regions thought to be weak enhancers , and regions thought to be insulators . According to Grossman et al . , there are defined regions under natural selection before and after this region ( chr15:45145764–45258860 and chr15:48539026–48633153 ) , and all are very likely to participate in the transcriptional regulation of SLC24A5 [18] . The natural selection features require knowledge of three things for interpretation: 1 ) population A , 2 ) population B , and 3 ) the population where this signature was identified . When all three of these are within the same ancestral or continental group , we expect very few differences in low frequency burden . However , if population A is the same or similar to the population possessing the selection signature and population B is different , we expect significant differences in low frequency burden between population A and population B . In our results , we found that the vast majority of regions considered to be under natural selection had significant differences in low frequency burden between disparate ancestral populations , which support the theory of selection in these regions . In general , size of bins can influence the number of stratified variants contained and thus the significance of that bin . It is important to prove that this is because larger bins have a greater opportunity to “collect” variants that are stratified and not because of inflated type I error . We have tested type I error rates in bins between approximately 40 variants to over 100 , 000 variants , which covers all analyses presented in this paper , and found no correlation between bin size and Type I error rate ( unpublished data ) . However , it should also be noted that while larger bins have more chances to collect stratified variants , there is also a larger capacity to collect neutral variants that contribute noise and decrease the signal . Using CEU-YRI pathway burden analysis , we reviewed the correlation between pathway size and significance . The number of genes in pathways ranged from 1 to over 700 genes , with the average around 5 genes per group . Correlations for this data are shown in Figure S9 . Not surprisingly , there were very linear and positive correlations between number of loci , number of variants , and genomic coverage . However , each of these had a nonlinear and somewhat complex relationship with the log-transformed p-value . This is highlighted in Figure S9B , which shows the relationship between the −log10 transformed p-value and the log10 transformed number of variants in the bin . The trend indicates that p-values are positively correlated ( become more significant ) with numbers of variants in a bin when the numbers of variants are relatively small or very large . Two reasons could explain this correlation: 1 ) the false-positive rate is influenced by bin size ( number of variants per bin ) , and 2 ) true signals from gene bins with burden differences perpetuate higher numbers of significant pathway bins . After extensive simulation testing ( unpublished data ) and recent publications in the literature , we believe the later is true [32] , [33] . A single or small number of child bins ( gene bins in this example ) , can drive parent bins ( pathways in this example ) to be significant even if no other child bin contains stratification . The comparison in Figure S10 between group sources available in LOKI suggests KEGG , NetPATH , PharmGKB , and Reactome have consistently larger bins ( higher number of loci , variants , and coverage ) . On average , these same four sources also tend to have bins with smaller p-values . Therefore , larger pathways are more likely to contain a gene with extreme low frequency variant stratification . Population stratification is incredibly important in genomic analyses , particularly when low frequency variants are being studied . Expected stratification and potential bias is related to bin size and functional significance of region studied . Regions with more selective pressure often have fewer differences between populations than one would expect by chance . However , it is also important to consider the size of the region since population stratification tends to become more of a problem in large bins . The x-axis of each matrix plot ( i . e . Figure 3 ) are oriented with African continental populations on the far right and the continental group with the highest proportion of significantly different low frequency variant bins on the far left . According to these matrix plots , Asian populations have more bins that are significantly different when compared to African populations than European/African population comparisons . Popular evolutionary theories suggest that the population that left Africa split before travelling East and West . One would expect low frequency burden differences ( compared to African populations ) to be very similar . However , populations from the Asian continental group tend to have more low frequency burden differences with African groups than European descent populations differences with African groups . There are at least three possible explanations; first , the Asian populations were the only continental group to be sequenced on the same technology , which could introduce a different bias when testing any of these populations with populations outside of Asian ancestry . While this is true of the 1 , 005 unrelated individuals , there were cryptically related individuals sequenced using SOLID technologies in all three of the Asian populations . The only population ( including cryptically related individuals ) to be sequenced exclusively on Illumina was TSI . When we examined the Asian populations and included the cryptically related individuals ( and thus individuals sequenced with different technologies ) , the trend was the same . Asian populations are the most different from African populations with regard to low frequency variant burden . The second potential explanation is that Asian populations had considerable proportions of cryptic relatedness that had to be removed for our analysis , 49 of the 75 individuals removed were from Asian populations . Perhaps there was something unique about how those samples were collected . The third and most interesting explanation is a speculation that involves the journey for early Asian populations after leaving Africa . Travelling east was much different geographically than travelling west . For example , early Asian migrants would have traversed the Himalayan Mountains . The harsh travel could have induced bottlenecks and other evolutionary mechanisms that would uniquely change the genetic architecture , specifically the architecture of low frequency variation . The course of travel for European descent populations was very different; they were exposed to unique challenges and climates . As each continental group diverged from Africa , their separate paths could explain why the difference in burden exists ( EAS/AFR and EUR/AFR ) . As we continue in pursuit of genetic etiology explaining heritability in common , complex disease , it is important to consider multiple types of genomic data , specifically variation beyond common variants . Low frequency variants are more frequent in the genome than common variants and are likely to have significant functional impact on human health . However , as we look forward to many successes in next-generation data analysis , it has become increasingly clear that we can't apply the same methods and corrections to low-frequency variants as we did in GWAS . Since low frequency variants are often recent mutations , they are specific to continental ancestry groups . This provides two important conclusions . First , potentially functional low-frequency variants are likely not the same across distantly related individuals . Second , low frequency population substructure leads to substantial differentiation and cannot be ignored [11] . Until relatively recently , we have not focused on the challenges presented by low frequency population stratification . Current methods used for GWAS to correct for ancestry are not likely adequate for low frequency stratification [34] , [35] . Therefore , it is imperative that researchers are aware of potential pitfalls stratification can introduce to low frequency genomic analyses . In summary , we were able to expose the magnitude of low frequency population stratification between all populations available in 1000 Genomes Project Phase I release across multiple interesting biological features . The magnitude of low frequency stratification appeared to be dependent on the functional location of the variation and the genomic size of the pertinent features . For example , there were fewer differences in low frequency burden in coding regions than intergenic regions . We found features with less variant tolerance and possibly more evolutionary constraint to have fewer differences in low frequency variant burden between different populations , i . e . significant low frequency bins seemed to be consistent with mutation theory . In addition , larger features were more likely to contain stratified variants and thus be significantly different between two populations . Low levels of stratification existed even between populations of the same continental group . The results of this study serve as a warning to researchers whom wish to use population control groups such as 1000 Genomes Project or shared control sets , unmatched case and control groups can contribute significantly to type I error rates . Future studies should focus on methods to accurately control for low frequency population stratification .
BioBin is a standalone command line application written in C++ that uses a prebuilt LOKI database described below ( software paper in preparation ) . Source distributions are available for Mac and Linux operating systems and require minimal prerequisites to compile . Included in the distribution are tools that allow the user to create and update the LOKI database by downloading information directly from source websites . The computational requirements for BioBin are quite modest; for example , during testing , a whole-genome analysis including 185 individuals took just over two hours using a single core on a cluster ( Intel Xeon X5675 3 . 06 GHz processor ) . However , because the vast amount of data included in the analysis must be stored in memory , the requirements for memory usage can be high; the aforementioned whole-genome analysis required approximately 13 GB of memory to complete . Even with large datasets , BioBin can be run quickly without access to specialized computer hardware or a computing cluster . The number of low frequency variants is the primary driver of memory usage [36] . BioBin is open-source and publicly available on the Ritchie lab website ( http://ritchielab . psu . edu/ritchielab/software/ ) . Harnessing prior biological knowledge is a powerful way to inform collapsing feature boundaries . BioBin relies on the Library of Knowledge Integration ( LOKI ) for database integration and boundary definitions . LOKI contains resources such as: the National Center for Biotechnology ( NCBI ) dbSNP and gene Entrez database information ( downloaded dbSNP b137: Dec 21 2012 , Entrez: Feb 1 2013 ) [37] , Kyoto Encyclopedia of Genes and Genomes ( KEGG , downloaded Dec 21 2012 , Release 64 ) [38] , Reactome ( downloaded Dec 12 2012 ) [39] , Gene Ontology ( GO , downloaded Feb 1 2013 ) [40] , Protein families database ( Pfam , downloaded Dec 1 2011 ) [41] , NetPath - signal transduction pathways ( downloaded Sept 3 2011 ) [42] , Molecular INTeraction database ( MINT , downloaded Oct 29 2012 ) [43] , Biological General Repository for Interaction Datasets ( BioGrid , downloaded Feb 1 2013 , version 3 . 2 . 97 ) [44] , Pharmacogenomics Knowledge Base ( PharmGKB , downloaded Jan 6 2013 ) [45] , Open Regulatory Annotation Database ( ORegAnno , downloaded Jan 10 2011 ) [46] , and evolutionary conserved regions from UCSC Genome Browser ( downloaded Nov 10 2009 ) [14] . LOKI provides a standardized interface and terminology to disparate sources each containing individual means of representing data . The three main concepts used in LOKI are positions , regions and groups . The term position refers to single nucleotide polymorphisms ( SNPs ) , single nucleotide variants ( SNVs ) or low frequency variants . The definition of region has a broader scope . Any genomic segment with a start and stop position can be defined as a region , including genes , copy number variants ( CNVs ) , insertions and deletions , and evolutionary conserved regions ( ECRs ) . Sources are databases ( such as those listed above ) that contain groups of interconnected information , thus organizing the data in a standardized manner . LOKI is implemented in SQLite , a relational database management system , which does not require a dedicated database server . The user must download and run installer scripts ( python ) and allow for 10–12 GB of data to be downloaded directly from the various sources . The updater script will automatically process and combine this information into a single database file ( ∼6 . 7 GB range ) . A system running LOKI should have at least 50 GB of disk storage available . The script to build LOKI is open source and publicly available on the Ritchie lab website ( http://ritchielab . psu . edu/ritchielab/software/ ) . Users can customize their LOKI database by including or excluding sources , including additional sources , and updating source information as frequently as they like [36] . We chose NCBI dbSNP and NCBI Entrez Gene as our primary sources of position and regional information due the quality and reliability of the data , clearly defined database schema , and because they contain gene IDs that map to the majority of group sources in LOKI . Gene boundary definitions were derived from NCBI Entrez . Pathway/group bins , regulatory regions , and evolutionary conserved regions were created using sources available in LOKI ( sources detailed in Software section ) . Some sources explicitly provide lists of genes in pathways; others provide groups of genes , which share a biological connection ( i . e . protein-protein interactions ) . For the purposes of this study , any bin created by multiple regions/genes will be analyzed in the “Pathway-Groups” feature analysis . External custom input files were generated using boundaries of annotated exon regions from UCSC to bin exon and intron specific variants . For example , if Gene A has three exons and two introns , only two bins would be created: GeneA-exons and GeneA-introns . GeneA-exons would contain all variants that fell within any of the three Gene A exon boundaries . External custom feature files were also generated for regions under natural selection by combining regions provided by previously published work [18] , [19] . Example binning strategies can be seen in Figure 8 . Using hierarchical biological relationships and optional functional or role information , BioBin can create many combinations of variants to bin . Custom feature files and additional binning details are explained in Text S4 , Table S4 , and Table S5 . BioBin is a bioinformatics tool used to create new feature sets that can then be analyzed in subsequent statistical analyses . Statistical tests used with BioBin can be chosen according to the hypothesis being tested , the question of interest , or the type of data being tested [36] . Unless otherwise noted , the results presented herein were calculated using a Wilcoxon 2-sample rank sum test implemented and graphed in the R statistical package [47] , [48] . P-values presented have been corrected using a standard Bonferroni correction , adjusting for the number of bins created and tested in a given analysis . Simulations confirming the power and validity of using the Wilcoxon 2-sample rank sum test are described in Text S5 and Table S6 . To investigate population stratification using BioBin , we analyzed the 1000 Genomes Project Phase I data . The 1000 Genomes Project was started in 2008 with the mission to provide deep characterization of variation in the human genome . As of October 2011 , the sequencing project included whole-genome sequence data for 1080 individuals , and aimed to sequence 2 , 500 individuals by its completion [49] . We removed 75 cryptically related individuals and conducted a pairwise comparison of low frequency variant burden differences between all 13 ancestry groups included in the phase I release of the 1000 Genomes Project ( October 2011 release ) . Table 1 provides the total number of variants ( common and low frequency ) and individuals included in Phase I VCF files of 1000 Genomes Project data for 1080 individuals in all 13 populations . In any genetic study , and especially in consideration of low frequency variants , it is important to evaluate sample relatedness . We combined populations by continental ancestry ( i . e . AFR continental group includes ASW , LWK , YRI ) and evaluated sample relatedness between and within the general ancestry groups using identity-by-state ( IBS ) and identity-by-descent ( IBD ) . Pairwise IBS represents the number of shared alleles at a specific locus between two individuals . IBS can be observed as 0 , 1 , or 2 depending on how many alleles are in common between the pair . If the shared alleles are inherited from a recent common ancestor , they are also considered IBD . Pairwise IBS calculations for low-frequency variants approximate IBD since the variants are likely to be recent and the chance of being identical because of recurrence is rare [50] . We used plink and plink-seq to estimate pairwise IBS and IBD for individuals of the same general ancestry group ( http://atgu . mgh . harvard . edu/plinkseq/ , http://pngu . mgh . harvard . edu/~purcell/plink/ ) [51] . For common variants , we created an independent subset of SNVs with a minor allele frequency greater than 5% and r2 values less than 0 . 2 to calculate pairwise IBD between individuals . For example , for the populations of African descent ( LWK , ASW , and YRI ) we grouped all of the individuals from these three populations and calculated the IBD . We removed maximally connected or related individuals in a parsimonious and iterative manner and repeated the IBD analysis until the maximum pairwise pi_hat score was less than or equal to 0 . 3 . After repeating this analysis in each continental group , 75 individuals were dropped from BioBin analyses based on our threshold for cryptic relatedness . We also evaluated allele sharing within and between major ancestral groups using plink-seq to calculate IBS for low frequency variants and common variants ( threshold 0 . 03 MAF and 0 . 25 MAF , respectively ) . Even though we estimated IBD in common variants ( described above ) , we calculated the IBS in low frequency and common variants separately to ensure the results were consistent . Using the ratio of shared alleles divided by the total number of genotyped alleles between two individuals , we evaluated excess sharing of low frequency variants ( MAF<0 . 03 ) compared to excess sharing of common variants ( MAF>0 . 25 ) . Feature selection in BioBin is a clear innovation over other available collapsing methods . Knowledge of biological features , such as genes and pathways , are available through LOKI for binning . In this analysis , we used the feature options of BioBin to investigate a variety of biologically relevant bins for differences in low frequency variant burden across 13 populations . We implemented a minimum bin size of two variants , inter-region bin size of 50 kb , and set the MAF binning threshold to 0 . 03 . We chose a 3% MAF binning threshold to focus our analysis on rare and near rare variation that differs between population groups . Additional details concerning binning parameters can be found in the Text S4 . We binned genes ( introns , exons , nonsynonymous variants , and predicted deleterious variants ) , intergenic regions , pathways , pathway-exons , regulatory regions , evolutionary conserved regions , and regions thought to be under natural selection . Natural selection can alter genomic variation in features , particularly in regions with some impact on protein function ( regulatory regions , coding regions ) . Positive selection on a specific variant allows the advantageous variant to sweep through a population , which can lead to an excess of common variants . Alternatively , weak negative selection or purifying selection can result in selective removal of deleterious alleles . This can decrease variation around the locus under selection and lead to an excess of rare or low frequency variation [52] . Commonly , evidence of natural selection is found only in one ancestral group , which is consistent with the idea that these selection events postdate population separation [53] . Because of this differentiation among populations , we were interested in using regions identified as being under selective pressures as features in a BioBin analysis . Table 2 shows the analysis plan , features tested , sources used , and the mean number of bins generated across all pairwise comparisons . After evaluating the population comparisons for the features described in Table 2 , we investigated the linkage disequilibrium ( LD ) in 10 top-ranked bins for three population comparisons , CEU-CHB , CHB-YRI , CEU-YRI . We calculated the LD between binned variants and determined the number of variants inside of a bin in LD with an r2> = 0 . 3 . We also evaluated the correlation between pathway significance and bin size . We took all of the pathways in the YRI/CEU analysis and compiled the following information for each pathway bin; total genomic coverage , number of genes , number of independent genes , number of loci , number of variants , and BioBin p-value . Because the majority of pathways or groups are not very large , the data was heavily skewed ( see Figure S8 ) . We performed a log10 transformation on all six variables: number of genes in the pathway or group , number of unique genes ( not present in any other pathway or group ) , number of loci in the pathway bin , number of variants in the pathway bin , genomic coverage of the pathway bin , and the BioBin reported Bonferroni adjusted p-value . Because of the skewness , we removed any pathway bins that had transformed loci values outside of 2 . 5 standard deviations of the log-transformed loci mean .
|
Low frequency variants are likely to play an important role in uncovering complex trait heritability; however , they are often continent or population specific . This specificity complicates genetic analyses investigating low frequency variants for two reasons: low frequency variant signals in an association test are often difficult to generalize beyond a single population or continental group , and there is an increase in false positive results in association analyses due to underlying population stratification . In order to reveal the magnitude of low frequency population stratification , we performed pairwise population comparisons using the 1000 Genomes Project Phase I data to investigate differences in low frequency variant burden across multiple biological features . We found that low frequency variant confounding is much more prevalent than one might expect , even within continental groups . The proportion of significant differences in low frequency variant burden was also dependent on the region of interest; for example , annotated regulatory regions showed fewer low frequency burden differences between populations than intergenic regions . Knowledge of population structure and the genomic landscape in a region of interest are important factors in determining the extent of confounding due to population stratification in a low frequency genomic analysis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"genome",
"sequencing",
"genome",
"analysis",
"tools",
"biology",
"genomics",
"computational",
"biology"
] |
2013
|
Low Frequency Variants, Collapsed Based on Biological Knowledge, Uncover Complexity of Population Stratification in 1000 Genomes Project Data
|
In the glucose-free environment that is the midgut of the tsetse fly vector , the procyclic form of Trypanosoma brucei primarily uses proline to feed its central carbon and energy metabolism . In these conditions , the parasite needs to produce glucose 6-phosphate ( G6P ) through gluconeogenesis from metabolism of non-glycolytic carbon source ( s ) . We showed here that two phosphoenolpyruvate-producing enzymes , PEP carboxykinase ( PEPCK ) and pyruvate phosphate dikinase ( PPDK ) have a redundant function for the essential gluconeogenesis from proline . Indeed , incorporation of 13C-enriched proline into G6P was abolished in the PEPCK/PPDK null double mutant ( Δppdk/Δpepck ) , but not in the single Δppdk and Δpepck mutant cell lines . The procyclic trypanosome also uses the glycerol conversion pathway to feed gluconeogenesis , since the death of the Δppdk/Δpepck double null mutant in glucose-free conditions is only observed after RNAi-mediated down-regulation of the expression of the glycerol kinase , the first enzyme of the glycerol conversion pathways . Deletion of the gene encoding fructose-1 , 6-bisphosphatase ( Δfbpase ) , a key gluconeogenic enzyme irreversibly producing fructose 6-phosphate from fructose 1 , 6-bisphosphate , considerably reduced , but not abolished , incorporation of 13C-enriched proline into G6P . In addition , the Δfbpase cell line is viable in glucose-free conditions , suggesting that an alternative pathway can be used for G6P production in vitro . However , FBPase is essential in vivo , as shown by the incapacity of the Δfbpase null mutant to colonise the fly vector salivary glands , while the parental phenotype is restored in the Δfbpase rescued cell line re-expressing FBPase . The essential role of FBPase for the development of T . brucei in the tsetse was confirmed by taking advantage of an in vitro differentiation assay based on the RNA-binding protein 6 over-expression , in which the procyclic forms differentiate into epimastigote forms but not into mammalian-infective metacyclic parasites . In total , morphology , immunofluorescence and cytometry analyses showed that the differentiation of the epimastigote stages into the metacyclic forms is abolished in the Δfbpase mutant .
Trypanosomes of the Trypanosoma brucei species complex are the etiological agents of Human African Trypanosomiasis , a parasitic disease that affects about 20 countries in sub-Saharan Africa [1] . T . brucei adapts to the different environments encountered in its insect ( tsetse fly ) and mammalian hosts by remodeling its metabolism . In the glucose-rich environment of mammalian blood , the bloodstream forms of T . brucei rely solely on glucose to produce energy . However , in the glucose-free midgut environment of its insect vector , the procyclic form ( PCF ) of the parasite develops an elaborated energy metabolism based on amino acids such as proline , which has recently been proved to be essential for the parasite development , at least in the tsetse midgut [2 , 3] . Although glucose is absent from the tsetse midgut lumen between blood meals , the PCF of T . brucei prefers glucose to proline when both carbon sources are available [4] . In these conditions , glucose is converted by aerobic fermentation to the partially oxidised and excreted end products , succinate and acetate [5 , 6] . The first seven steps of glycolysis are sequestered within peroxisome-like organelles , called glycosomes [7 , 8] . Phosphoenolpyruvate ( PEP ) is produced in the cytosol , where it is located at a branching point to feed the glycosomal ‘succinate branch’ and the mitochondrial ‘acetate and succinate branches’ ( Fig 1B ) . Both succinate branches are initiated by the glycosomal PEP carboxykinase ( PEPCK , EC 4 . 1 . 1 . 49 , step 16 in Fig 1 ) by conversion of PEP into oxaloacetate , which is further converted into malate in the glycosomes ( step 15 ) , before being metabolised into succinate in both the glycosomes ( steps 13 and 14 ) and the mitochondrion ( steps 6B and 7 in Fig 1B ) [9 , 10] . PEPCK , together with the glycosomal pyruvate phosphate dikinase ( PPDK , EC 2 . 7 . 9 . 1 , step 17 ) [11] , are directly involved in the maintenance of the glycosomal ADP/ATP balance , by regenerating the ATP consumed by the first and third glycolytic steps ( hexokinase , EC 2 . 7 . 1 . 1 , step 31 and phosphofructokinase , PFK , EC 2 . 7 . 1 . 11 , step 26 , ) [12] . Part of PEP is also converted in the cytosol to pyruvate , which enters the mitochondrion to feed the pyruvate dehydrogenase complex ( PDH , EC 1 . 2 . 4 . 1 , step 9 ) for acetyl-CoA production , which is further converted into excreted acetate by mitochondrial acetate:succinate CoA-transferase ( ASCT , EC 2 . 8 . 3 . 8 , step 10 ) and acetyl-CoA thioesterase ( ACH , EC 3 . 1 . 2 . 1 , step 11 ) [13–15] . It is noteworthy that a canonical tricarboxylic acid cycle , with acetyl-CoA being converted into CO2 , is not operative in trypanosomes [16] . In glucose-rich condition , proline contributes moderately to central carbon metabolism and is primarily converted into the excreted end product succinate ( Fig 1B ) [4] . In contrast , when the PCF are incubated in the absence of glucose , a situation likely encountered by trypanosomes in their different in vivo niches , proline becomes the main carbon source used by the parasite , being consumed up to 6-fold more [4] . In these conditions , proline feeds the whole central carbon and energy metabolism ( Fig 1A ) . For instance , proline-derived succinate produced in the mitochondrion , even in the presence of glucose ( steps 1–5 in Fig 1B ) , is further metabolised and converted into excreted alanine [4] . The absence of glucose also implies that glucose 6-phosphate ( G6P ) , a precursor for essential pathways , such as the pentose phosphate pathway ( PPP ) and nucleotide sugar biosynthesis [17 , 18] , needs to be produced by reversal of glycolysis through the so-called gluconeogenesis . Under physiological conditions , all glycolytic enzymes involved in G6P conversion into PEP catalyse a reversible reaction used for glyconeogenesis , except PFK that is replaced by fructose-1 , 6-bisphosphatase ( FBPase , EC 2 . 7 . 1 . 40 , step 25 ) , as observed in many organisms including Leishmania major [19] . In addition , the production of the main gluconeogenic precursor , PEP , can theoretically be performed in trypanosomatids by PPDK from pyruvate and/or by PEPCK from oxaloacetate . In the promastigote form of L . major , PPDK and PEPCK both contribute to mannogen biosynthesis through gluconeogenesis from alanine and aspartate , respectively [20] . In contrast , proline has been reported to be incorporated into hexose phosphates through gluconeogenesis in the PCF of T . brucei grown in the absence of glucose , however the contributions of PPDK and PEPCK in these conditions have not been addressed so far [21] . Here , we used a combination of multiple mutations ( up to three genes targeted at the same time by knock-out and/or RNAi-mediated knock-down ) and metabolomic analyses to study the role of the key gluconeogenic enzymes in the PCF trypanosomes . We showed that PPDK and PEPCK have a redundant essential function for the incorporation of [13C]-enriched proline into gluconeogenic intermediates . In addition , gluconeogenesis from glycerol is abolished in the glycerol kinase ( GK , EC 2 . 7 . 1 . 30 , step 32 ) null background . FBPase gene knock-out abolished colonisation of the tsetse fly vector salivary glands by the parasite , as well as in vitro differentiation of the epimastigote forms into the metacyclic form , demonstrating the crucial role of gluconeogenesis , at least during the last part of the trypanosome cyclical development in the tsetse fly .
PCF trypanosomes incubated in glucose-free and proline-rich conditions are able to maintain significant levels of ( glycolytic ) sugar-phosphates , thought at lower extent than cells incubated with glucose and proline ( S1 Table ) . This means that proline is able to support the synthesis of the sugar-phosphates by a gluconeogenesis process . According to the current view of the central carbon metabolism of the PCF trypanosomes , PPDK and PEPCK are the only enzymatic steps that can produce PEP , the precursor of gluconeogenesis , in proline catabolism ( steps 16 and 17 in Fig 1A ) . These two enzymes have been shown to contribute to glucose catabolism , however , their roles in glucose-depleted conditions have not been investigated so far [12 , 22] . To investigate this poorly explored gluconeogenic metabolic pathway from proline , we have determined by mass spectrometry the incorporation levels of 13C atoms from uniformly [13C]-enriched proline ( [U-13C]-proline ) into key gluconeogenic intermediates of parental and knock-out ( Δppdk , Δpepck and Δppdk/Δpepck ) PCF cell lines . Cells were incubated in PBS containing 2 mM [U-13C]-proline in the presence or absence of the same amounts of non-labelled glucose . Incorporation of 13C into glycolytic intermediates was quantified by IC-MS/MS and the values for selected glycolytic metabolites are shown in Fig 2 . Incorporation of 13C atoms into hexose phosphate glycolytic intermediates from [U-13C]-proline is very low in the presence of glucose ( 3 . 1% on average ) , compared to cells incubated in the absence of glucose ( 94 . 1% on average ) . The same labelling pattern was also observed for the intermediates of the pentose phosphate pathways , i . e . 6-phosphogluconolactone and sedoheptulose 7-phosphate . This confirms that proline feeds gluconeogenesis and the pentose phosphate pathway in the absence of glucose , while the presence of glucose down-regulates conversion of proline into metabolite intermediates of these pathways , as we previously observed [21] . Incorporation of at least two 13C atoms into the same hexose phosphates is more reduced in the Δppdk mutant than in the Δpepck mutant ( 13 . 8% versus 49 . 3% on average , respectively ) , suggesting that PPDK might contribute more than PEPCK to gluconeogenesis from proline ( Fig 2 ) . More importantly , this 13C-incorporation is abolished in the Δppdk/Δpepck double mutant ( not detectable in Fig 2 ) , although proline consumption is increased by 2-fold [12 , 21] . This demonstrates that PPDK and PEPCK are the only enzymes significantly contributing to PEP production from proline catabolism and have a redundant function for the essential production of glycolytic intermediates from proline . These data also confirm that , under the conditions pertaining in the cell , pyruvate kinase ( EC 2 . 7 . 1 . 40 , step 18 in Fig 1 ) irreversibly converts PEP into pyruvate , even in the absence of PEPCK and PPDK . To explore whether blocking gluconeogenesis from proline affects parasite growth , we developed an SDM79-derived glucose-free medium ( SDM79-GlcFree ) containing less than 1 μM glucose , as determined by NMR spectrometry and supplemented with 50 mM N-acetyl D-glucosamine , a competitive inhibitor of glucose transport [21] . The growth of the parental EATRO1125 . T7T PCF cell line is moderately affected by the absence of glucose with a doubling time increased by ~20% in the SDM79-GlcFree compared to the same medium supplemented with 10 mM glucose ( Fig 3A ) . Similarly , growth of the Δppdk/Δpepck cell line is not compromised in the absence of glucose , suggesting that another gluconeogenic carbon source can be used by the PCF in addition to proline ( Fig 3A ) . The abolition of PEP production from gluconeogenic amino acids ( including proline ) , as demonstrated above , strongly supports the view that glycerol might be a possible alternative to proline that would feed gluconeogenesis in the Δppdk/Δpepck cell line . To address this hypothesis , we have measured by IC-MS/MS the [13C]-incorporation into glycolytic intermediates of the parental PCF incubated with [U-13C]-glycerol . In this experiment , most hexose phosphate glycolytic intermediates ( 83 . 2% on average ) are fully [13C]-enriched after 2 h of incubation with [U-13C]-glycerol as the only carbon source ( S1 Fig ) . In order to abolish all possible gluconeogenic pathways , the expression of the key enzymes of glycerol conversion pathway was successively down-regulated in the Δppdk/Δpepck null background , i . e . glycerol kinase ( GK , EC 2 . 7 . 1 . 30 , Tb927 . 9 . 12550-Tb927 . 9 . 12630 , step 32 ) , glycerol-3-phosphate dehydrogenase ( GPDH , EC 1 . 1 . 1 . 8 , Tb927 . 8 . 3530 , step 33 ) and triose phosphate isomerase ( TIM , EC 5 . 3 . 1 . 1 , Tb927 . 11 . 5520 , step 23 ) . We hypothesised that the abolition of gluconeogenesis from both proline and glycerol , in the Δppdk/Δpepck/RNAiGK , Δppdk/Δpepck/RNAiGPDH and Δppdk/Δpepck/RNAiTIM triple mutants , should lead to the death of parasites grown in glucose-free conditions . Indeed , the growth of the tetracycline-induced Δppdk/Δpepck/RNAiGK cell line ( Δppdk/Δpepck/RNAiGK . i ) in SDM79-GlcFree medium stopped after 10 days of tetracycline induction and cells ultimately died two weeks later , while the addition of glucose restored their growth ( Fig 3B ) . Similarly , the growth of the Δppdk/Δpepck/RNAiTIM . i mutant was strongly affected by the absence of glucose , but this was not lethal to the cells ( Fig 3D ) . However , the growth of the Δppdk/Δpepck/RNAiGPDH . i cells appears to be only moderately affected by the absence of glucose , although GPDH was not detected by western blotting anymore ( Fig 3C ) , probably because the glycerol 3-phosphate/dihydroxyacetone phosphate shuttle could bypass the GPDH step ( see Fig 2 in [5] for this pathway ) . To determine whether the glycerol metabolism is abolished in the Δppdk/Δpepck/RNAiGK . i cell line that shows the strongest growth alteration in SDM79-GlcFree medium , quantitative analyses of 13C-enriched end products excreted from [U-13C]-glycerol metabolism were performed by proton NMR spectrometry [6] . It is noteworthy that ( i ) the use of 13C-enriched glycerol as carbon source allows us to distinguish between the products excreted from metabolism of [U-13C]-glycerol and an uncharacterised internal carbon source [6 , 23] , and ( ii ) the succinate excretion is abolished in the Δppdk/Δpepck/RNAiGK . i mutant ( Fig 4 ) , as previously observed for the Δppdk/Δpepck and Δpepck cell lines [12 , 22] , since the succinate branch is interrupted in the absence of PEPCK ( see Fig 1B ) . The Δppdk/Δpepck/RNAiGK . i mutant shows a 15-fold decrease of the total end products excreted from glycerol breakdown ( acetate , alanine , pyruvate and lactate ) compared to the parental Δppdk/Δpepck mutant , while glucose metabolism is not affected ( Fig 4 ) . This confirms that the growth arrest observed for the Δppdk/Δpepck/RNAiGK . i is due to the abolition of glycerol metabolism that is required to feed gluconeogenesis in the PPDK/PEPCK null background in glucose-free conditions . Gluconeogenesis from glycerol and proline leads to the production of fructose 1 , 6-bisphosphate ( F1 , 6BP ) , that is converted into fructose 6-phosphate ( F6P ) by the fructose-1 , 6-bisphosphatase ( FBPase , EC 3 . 1 . 3 . 11 , step 25 in Fig 1A ) , a well-known gluconeogenic enzyme that is , however , not involved in glycolysis . To confirm the gluconeogenic role of FBPase in PCF , both alleles of the single copy FBPase gene ( Tb927 . 9 . 8720 ) were replaced by the blasticidin and puromycin markers ( Δfbpase ) and FBPase expression was conditionally down-regulated by RNAi ( RNAiFBPase ) . Both the Δfbpase and RNAiFBPase . i cell lines are viable in the absence of glucose , although with a similarly increased doubling time compared to the parental cells ( 18 . 4 h and 18 . 5 h , respectively , compared to 13 . 7 h ) ( Fig 5B and 5C ) . These data suggest that , although the absence of FBPase affects parasite growth in glucose-free conditions , an alternative enzyme or pathway could substitute for the FBPase reaction , in such a way that proline and/or glycerol would contribute to G6P production even in the absence of FBPase . Re-expression of a rescue ectopic copy of the FBPase gene in the Δfbpase background using the tetracycline-inducible pLew100 vector ( Δfbpase/FBPase . i cell line ) restored the WT growth ( Fig 5D ) . To confirm the role of FBPase in gluconeogenesis , we determined by IC-MS/MS the incorporation levels of [U-13C]-proline into glycolytic intermediates of the WT , Δfbpase and Δfbpase/FBPase . i rescue cell lines ( Fig 6 ) . The incorporations of [13C]-enriched atoms from proline into triose phosphates ( PEP , 2/3PG , 1 , 3BPG and Gly3P , see Fig 1 for abbreviations ) and F1 , 6BP are equivalent in the parental , Δfbpase and Δfbpase/FBPase . i cell lines . However , the incorporations of [13C]-enriched atoms in F6P ( the product of the reaction catalysed by FBPase ) , as well as in metabolites produced from F6P , i . e . G6P , mannose 6-phosphate ( M6P ) and 6-phosphogluconolactone ( 6PG ) and sedoheptulose 7-phosphate ( S7P ) , are strongly reduced in the Δfbpase mutant compared to the parental cell line . Indeed , the relative amounts of non-enriched F6P , G6P , M6P and 6PG and S7P in the Δfbpase cells are increased by 6 . 5- , 4 . 6- , 6 . 9 , 7 . 2- and 5 . 8-fold , respectively . As expected , incorporation of 13C-carbons into hexose phosphates is restored in the Δfbpase/FBPase . i rescue cell line ( Fig 6 ) . These data clearly demonstrate that gluconeogenesis from proline , although strongly reduced , still occurs in the absence of FBPase , which implies the existence of an alternative step to FBPase producing G6P from triose phosphates . To order to highlight the alternative to FBPase , the FBPase activity was determined on the parental , Δfbpase and Δfbpase/FBPase . i ( rescue ) cell lines . This enzymatic activity was not detectable on total cellular extracts , but was detectable in the glycosomal fractions of the parental cell line at a low level compared to the glycerol kinase activity ( 14 . 8 ±4 . 5 versus 11 , 100 ±2 , 300 nmol min-1 mg-1 of proteins ) ( Fig 7A ) . Interestingly , this activity is not abolished in the Δfbpase mutant , although decreased by 3 . 8-fold . Reintroduction of the FBPase gene in the Δfbpase cell line only partially restored the FBPase activity ( 7 . 1 ±2 . 3 versus 14 . 8 ±4 . 5 nmol min-1 mg-1 of proteins ) , which is consistent with the lower level of FBPase expression in the rescue cell line compared to that of the parental cells ( Fig 7A ) . Altogether , these data strongly support the view that an unknown glycosomal enzyme is responsible for approximately one fourth of the glycosomal FBPase activity . Incidentally , the T . brucei genome contains a single gene potentially coding for a sedoheptulose-1 , 7-bisphosphatase ( SBPase , Tb927 . 2 . 5800 ) , which belongs to the same superfamily as the FBPases . Western blot analysis of the glycocomal and cytosolic fractions of the parental trypanosomes with the anti-FBPase and anti-SBPase immune sera demonstrated the glycosomal localisation of these two proteins ( Fig 7B ) , which is in agreement with previously published proteomic analyses of glycosomal fractions [24 , 25] . We then attempted to determine whether the absence of FBPase would affect parasite virulence in vivo in their tsetse fly vectors . To perform such experimental infections , a cell line bearing a single FBPase allele ( Δfbpase-/+ ) , a FBPase null mutant ( Δfbpase ) , and two distinct constitutive add-back rescue cell lines ( Δfbpase/FBPase-1 and 2 ) were generated in the AnTat1 . 1E pleomorphic strain [26] . Indeed , this strain has maintained the ability to efficiently colonise salivary glands and to produce infectious metacyclic forms , as opposed to parasites with the EATRO1125 genetic background [27] . Two strategies were developed to produce the Δfbpase/FBPase rescue cell lines , i . e . the addition of a FBPase ectopic copy under the control of the PARP promoter in the pLew100 vector ( Δfbpase/FBPase-1 ) , and the in-situ re-insertion of one FBPase gene under the control of its endogenous UTRs ( Δfbpase/FBPase-2 ) . Batches of about 50 teneral male tsetse flies were artificially fed through a silicone membrane with cultured PCF parasites of either one of the four cell lines in parallel ( the parental wild-type , the Δfbpase-/+ , the Δfbpase and the Δfbpase/FBPase-1 or -2 cell lines ) . Flies were dissected after one month in order to quantify the infection rates per organ , the parasite densities per organ and the proportion of parasite morphotypes as previously described [28] . A total of 784 flies were dissected with similar results obtained with the two series: 338 flies out of 4 biological replicates with the Δfbpase/FBPase-1 panel ( Fig 8A ) and 446 flies out of 6 replicates with the Δfbpase/FBPase-2 panel ( Fig 8B ) . First , the midgut infection rates were comparable in all conditions ( 33% on average , ranging from 19% ±27% in Fig 8A to 45% ±19% in Fig 8B for the null mutants ) . Second , although the salivary gland infection rates were on average 16% for the wild-type cells , 7% for the Δfbpase-/+ parasites and 3% for the Δfbpase/FBPase rescue cell line , no null mutant parasites were seen in any salivary glands and this observation was statistically significant ( p = 0 . 011 by ANOVA Tukey add-hoc post-test at 95% confidence when comparing %SG+ in Δfbpase versus wild-type ) . This demonstrates that FBPase is essential for the second part of T . brucei cyclical development , i . e . the establishment of mature infections in the insect salivary glands . The lower salivary gland infection rates observed with the Δfbpase/FBPase cell lines could be explained by the additional round of transformation and the longer time spent in culture during the generation of this cell line that would both have affected its virulence . It is also noteworthy that the in-situ rescue approach is apparently more adapted than the use of the pLew100 , probably because the PARP promoter used for driving gene expression in the pLew100 vector may be naturally less efficient in the short epimastigotes , attached epimastigotes and/or in the metacyclic stages during the late steps of the cyclical development as compared to the procyclic and mesocyclic midgut stages . Assuming that ( i ) the parasite densities in the midgut , estimated by microscopic observations , were similar for all the cell lines and that ( ii ) no obvious motility defect was detected among these cell lines neither in vitro nor in vivo , we reasoned that the absence of FBPase may have impaired parasite differentiation , especially the first morphotype switch occurring in proventricular parasites . Therefore , parasites were isolated from the anterior midgut and proventriculus of infected flies , and stained with a DNA marker ( DAPI ) and an axonemal marker Mab25 [29] in order to determine the proportions of cells in each morphotype ( trypomastigote versus epimastigote ) according to the cell lines ( Fig 8C ) . Nevertheless , no difference was detected between cell lines that were all displaying on average 80% of proventricular parasites in the trypomastigote morphotype and 20% of epimastigotes ( 280 to 692 cells out of 2 to 6 flies per cell line ) . This demonstrates again that FBPase is not essential for the trypanosome cyclical development in the midgut , while it is required at least for the initial colonisation of salivary glands . We reasoned that these differences in behaviour could be the consequence of possible low proline concentrations in the oesophagus and/or the salivary glands of the tsetse , which would not be sufficient for feed gluconeogenesis in the absence of FBPase . To test this hypothesis , we estimated the growth of the parental and Δfbpase cell lines as a function of proline concentrations in the SDM79-GlcFree medium , using the Alamar Blue assay . As expected these two cell lines failed to grow under less than 20 μM proline , this amino acid being the primary carbon source used by the parasite to feed its central carbon metabolism ( Fig 8D ) . Growth of the parental cells is considerably improved by addition of at least 20 μM proline , the parental cells growing only 2-times slower in the presence of 0 . 5 mM compared to 5 mM proline ( concentration in the SDM79-GlcFree medium ) . In contrast the Δfbpase cells are barely growing in the presence of 0 . 5 mM proline . This analysis showed that the Δfbpase mutant requires ~30-times more proline to grow at the same speed as the parental cells ( Fig 8D ) , which may provide a rational explanation of the observed difference in behaviour of these two cell lines in vivo if proline concentrations are higher in the midgut compared to the oesophagus and/or the salivary glands . Unfortunately , the concentrations of proline in the tsetse organs are unknown . It is noteworthy that , immunofluorescence analyses using an immune serum against a bona fide glycosomal protein , aldolase , showed no difference in the number of glycosomes per cell between the parental , Δfbpase and Δfbpase/FBPase cell lines ( S2 Fig ) . In order to deepen the study of the in vivo role of FBPase , variations of the FBPase expression during the parasite cycle of wild-type trypanosomes were scrutinised by immunofluorescence analysis with an anti-FBPase antibody ( Fig 9A ) . Assuming that the observed fluorescence is directly correlated to the amount of expressed proteins accessible to the antibody , all stages were seen to express FBPase in their glycosomes , yet in variable amounts ( Fig 9A ) . The maximum signal intensities ( Fig 9B ) and the total amounts of fluorescent signal per surface unit ( Fig 9C ) were measured in a total of 454 individual cells isolated from more than 12 flies , further normalised to the values obtained on PCF cells from the same fly after removal of the background values , and finally plotted by stage in percentage of the fluorescent signals from PCF . The lowest maximum fluorescent signal intensities ( Fig 9B ) and total amounts of fluorescence per surface unit ( Fig 9C ) were observed in proventricular epimastigotes ( DE , LE and SE ) whereas the highest values were detected in salivary glands metacyclic forms ( MT ) , and the differences between the two groups were statistically significant by ANOVA ( p = 0 . 002 , <0 . 001 , <0 . 0001 and <0 . 0001 between MT Vs . DE , LE , SE and AE respectively by Tukey ad-hoc post-test on the maximum intensities , and p = 0 . 002 and <0 . 0001 between MT Vs . LE and AE for the total fluorescence per surface unit ) . The naturally lower FBPase expression in proventricular epimastigotes is in accordance with the absence of phenotype of the Δfbpase null mutants in terms of morphotype switching ability ( Fig 9A ) . The fly passage phenotype observed may be due to an impairment of the parasite developmental program per se or to defects in migration , and/or to mobility defects during parasite migration in the tsetse and/or to non-adapted interactions with the tsetse environments . To distinguish between these possibilities , we tested the impact of the FBPase deletion in an in vitro model mimicking the parasite development upon inducible over-expression of the RNA binding protein 6 ( RBP6 ) , a central regulator of fly-stage differentiation [30] . The stage-specific morphology of induced cells , based on cell size and shape , as well as the relative position of the kinetoplast to the nucleus were monitored over the differentiation kinetics [31] . In PCF trypomastigotes the kinetoplast is at the posterior part of the cell , whereas in epimastigotes ( EMF ) the kinetoplast migrates to the opposite side of the nucleus and is found at the anterior part of the cell or in close proximity to the nucleus . Metacyclic trypomastigotes ( MF ) are smaller than EMF and PCF with the kinetoplast at the very end of the rounded posterior tip and a bloodstream form-like flagellar shape . In addition , we used two distinct stage-specific marker proteins for staging , i . e . EP procyclin ( EP ) for PCF [32] and calflagin for MF [33] . Calflagins are proteins localising to lipid rafts in the flagellar membrane and their amounts in bloodstream forms and MF are ten-fold enriched compared to that in PCF . Four days after the tetracycline-induction of RBP6 over-expression , ~30% of the cells ( RBP6 . i ) had a characteristic MF morphology and after six days MF subpopulation reaches a peak at 55% ( Fig 10A , left ) . Flow cytometry of RBP6 . i cultures showed that EP expression was maintained in all cells during the first two days post induction and only then decreased proportionally to the fraction of the remaining PCF population ( Fig 10A , middle ) . EP expression has previously been described in early salivary gland EMF [34] . The fraction of cells expressing calflagin constantly increased over time and attained the same percentage as the MF subpopulation after four and six days ( Fig 10A , right ) . Previous reports showed that EP was not expressed in MF and that calflagins were detectable at basal levels in PCF but significantly up-regulated in MF [35 , 36] . Additional evidence for developmental changes in these cultures is provided by RBP6 expression levels that decreased over time of the experiment ( Fig 10C ) and by a cell proliferation arrest ( Fig 10D ) in successfully differentiating cultures . The Δfbpase RBP6 . i mutant , in which RBP6 was conditionally over-expressed in the Δfbpase background , was analysed in parallel under exactly comparable conditions ( Fig 10B ) . Most importantly , the RBP6 expression levels as determined by quantitative western blotting ( Fig 10C ) were similar in the Δfbpase and parental backgrounds over the entire time and did not decrease after 8 days . Induced Δfbpase RBP6 . i cells developed into EMF after two days but completely failed to differentiate into MF and consequently the fraction of EMF remained stable at around 50% for the duration of the experiment . EP and calflagin expression levels were not different between cell lines all over the time course and the population growth was maintained ( Fig 10D ) . As the EP-positive population did not decrease , it is likely that the developmental blockage occurred in early EMF . Together , we can conclude that FBPase is absolutely essential for development in culture and hence this strong phenotype is cell autonomous .
The insect stages of T . brucei live in the midgut , the foregut and the salivary glands of the tsetse fly , which are considered glucose-free environment between the insect blood meals . In this context , trypanosomes need to produce G6P , the precursor of essential pathways , from non-glycolytic carbon sources . Production of hexose phosphates from proline through gluconeogenesis has recently been described for the PCF trypanosomes [21] and was consistent with the previously described metabolic switch toward proline when the parasite is incubated in glucose-depleted conditions [4 , 37] , as well as the requirement of an active proline metabolism to support development of trypanosomes in the insect midgut [3] . However , the pathway leading to F6P production as well as possible alternative gluconeogenic carbon sources used by the PCF were not addressed so far . Here , we show that two key steps of gluconeogenesis , i . e . production of PEP and production of F6P , are achieved by redundant reactions in PCF trypanosomes grown in vitro . However , the canonical FBPase gluconeogenic enzyme , is essential for the infection of the tsetse fly salivary glands where the differentiation of the attached epimastigote forms into free metacyclic infective forms occurs [35] . Production of the gluconeogenic precursor PEP from proline is performed by two different and complementary T . brucei glycosomal enzymes , i . e . PPDK and PEPCK . Indeed , abolition of G6P production from proline is only observed in the PPDK/PEPCK null background , but not in the Δppdk and Δpepck single mutants . The implication of both the PEPCK and PPDK in gluconeogenesis has also been observed in Leishmania mexicana , however , the redundant effect was not investigated [20] . This redundancy between PPDK and PEPCK has also recently been observed in Arabidopsis thaliana , in which gluconeogenesis is critical to fuel the transition from seed to seedling [38] . To our knowledge , trypanosomatids and plants are the only eukaryotes known for using two distinct routes for the entry of organic acids into gluconeogenesis . In L . mexicana , PEPCK may participate in the entry of aspartate in promastigotes while PPDK is involved in the entry of alanine in amastigotes [20] . It is noteworthy , that this situation does not occur in T . brucei , since alanine and aspartate are not significantly consumed by the procyclic trypanosomes [4 , 39 , 40] . In A . thaliana PPDK participates in gluconeogenesis by remobilising amino acids such as alanine that give rise to pyruvate , while PEPCK is primarily involved in remobilisation of acetyl-CoA derived from lipids [38] . Interestingly , T . brucei is the only organism known so far to use PPDK in the gluconeogenic or in the glycolytic direction , depending on the carbon source availability . Indeed , we recently showed that in glucose-rich conditions PPDK plays a key role in the maintenance of the glycosomal ATP/ADP balance by converting PEP into pyruvate [12] , while here we report that in the absence of glucose PPDK performs the reverse reaction to feed gluconeogenesis . Important questions remain regarding the maintenance of the glycosomal ADP/ATP and redox balances in the PCF trypanosomes grown in glucose-free conditions . As mentioned above the glycosomal PPDK and PEPCK are critical for the maintenance of the glycosomal ADP/ATP balance by regenerating ATP when the parasite is fed with glucose [12] . However , in the absence of glucose , both enzymes work in the ATP-consuming direction with no glycosomal ATP-generating step in the gluconeogenic pathway . In this context , ATP could be provided to the enzymes involved in the gluconeogenetic pathway by another unknown glycosomal ADP/ATP-dependent enzyme , possibly encoded by one of the many hypothetical genes . Alternatively , glycosomal ATP could be regenerated through a carrier-mediated ADP/ATP exchange with the cytosolic compartment . Such a glycosomal carrier has not been described so far in trypanosomatids , although this hypothesis was previously proposed to explain the maintenance of the glycosomal ADP/ATP balance in the Δppdk/Δpepck mutant grown in glucose-rich conditions [12] . It is important to note that this hypothetical glycosomal ADP/ATP exchange activity should be low enough to prevent the lethal turbo-explosion of the unusual glycolysis developed by trypanosomes [41] . Indeed , trypanosomes lack feedback regulation of the early steps of glycolysis , which is compatible with the glycosomal sequestration of the enzymes within glycosomes . However , partial relocation of the glycosomal enzymes within the cytosol induces accumulation of toxic glycolytic intermediates and ATP depletion by the so-called turbo-explosion . One may consider that a high capacity ATP/ADP exchanger located in the glycosomal membrane would lead to the same negative effect in the presence of glucose , unless the expression of this putative exchanger was controlled by glucose levels . As expected , the T . brucei FBPase is involved in gluconeogenesis , as shown by a 6 . 5-fold reduction of F6P production from proline in the Δfbpase mutant , which was restored by re-expressing an ectopic FBPase copy in the Δfbpase background . However , to our surprise , the Δfbpase and RNAiFBPase . i mutants are viable in the glucose-free conditions , while blocking gluconeogenesis from both proline and glycerol in the Δppdk/Δpepck/RNAiGK . i and Δppdk/Δpepck/RNAiTIM . i cell lines compromises growth in the absence of glucose . These data , which are consistent with a residual production of 13C-enriched G6P from [U-13C]-proline in the Δfbpase cell line , imply the existence of an alternative to FBPase for production of hexose phosphates from gluconeogenic carbon source ( s ) . We must also take into account that the alternative to the FBPase reaction is probably not functional in Leishmania major , since FBPase has been shown to be essential for the development of the intracellular amastigote form as well as for the promastigote form growth in the absence of glucose [19] . Interestingly , the T . brucei gene potentially coding for SBPase is not present in the Leishmania genomes . The residual FBPase activity in the glycosomal fraction of the Δfbpase cell line strongly supports the view that the product of this putative T . brucei SBPase is also involved in gluconeogenesis in the PCF trypanosomes . Our interpretation of the data is consistent with the previously characterisation in Cyanobacteria and in yeast of FBPase/SBPase genes showing a dual-function FBPase/SBPase [42–44] . Whatever this unknown alternative pathway is , it seems to be functional in the absence of FBPase in both in vitro cultured PCF and in vivo during the early cyclical development of midgut trypomastigote parasites . It is however not sufficient to allow Δfbpase null mutants to colonise the tsetse fly vector salivary glands , demonstrating the crucial role of gluconeogenesis during this later part of the trypanosome cyclical development . These in vivo data showed that the rise of epimastigotes in the anterior midgut and proventriculus is not affected by the absence of FBPase . However , the absence of salivary gland infection suggests that either the proventricular epimastigotes have lost their capacity of maturation ( e . g . to attached epimastigotes ) and/or their abilty to migrate/orientate towards the salivary glands through the mouthparts and/or to attach to the salivary gland epithelium . The phenotype obtained in vitro in the RBP6-mediated differentiation model combined to the Δfbpase null mutation is fully consistent with the fly infection data as EMF accumulate and no MF are detectable . Moreover , the development programme arrest in culture rules out that gluconeogenesis is mainly required to meet specific metabolic requirements of the fly alimentary tract micro-environments . A more defined staging of the developmental arrest is not possible in vitro as the EMF accumulating in culture are heterogenous in morphology and do not precisely match the natural morphologies present in the fly . It should also be remembered that the RBP6-induced differentiation is a useful tool yet a clearly “forced” model . Parasite development seems to depend on the expression of an alternative pathway or an enzymatic activity complementing the FBPase gene deletion . Once the identity of this bypass will be unravelled , we will have to test whether this phenotype may be due to a threshold effect of flux through the gluconeogenic pathway . The growth of the PCF is compromised in glucose-free conditions only if triose phosphate production from proline and glycerol metabolism are together abolished in the Δppdk/Δpepck/RNAiGK . i cell line , since the RNAiGK . i and Δppdk/Δpepck mutants are viable in these growth conditions . This implies that the glycerol conversion pathway is sufficient to feed gluconeogenesis . However , glycerol amounts are in the range of 5–10 μM in standard SDM79 medium ( sera contain 50–100 μM glycerol [45 , 46] ) and should be consumed together with glucose by the PCF cells during preparation of the glucose-free medium . In the absence of glycerol in the medium , it could be produced from the medium compounds , such as phospholipids coming from fetal calf serum by the action of trypanosomal phospholipases [47] or from an endogenous carbon source . Interestingly , we previously reported the presence in PCF trypanosomes of a yet unknown endogenous carbon source detectable by excreted end products from its metabolism , acetate and succinate , in the absence of any extracellular carbon source [6 , 23] . We are presently analysing this endogenous carbon source and its possible involvement in gluconeogenesis .
The PCF of T . brucei EATRO1125 . T7T ( TetR-HYG T7RNAPOL-NEO ) and AnTat1 . 1E [26] were cultivated at 27°C in the presence of 5% CO2 in SDM79 medium containing 10% ( v/v ) heat-inactivated fetal calf serum and 3 . 5 mg ml-1 hemin [48] or in a glucose-depleted medium derived from SDM79 , called SDM79-GlcFree . This SDM79-GlcFree medium consists of a glucose-depleted SDM79 medium , containing 20% ( v/v ) heat-inactivated fetal calf serum , in which parental cells were cultured during 72 hours in order to consume the glucose coming from the serum and then diluted with the same volume of glucose-depleted SDM79 medium without serum to finally obtain SDM79-GlcFree . Glucose depletion was verified by NMR spectrometry analyses ( with a detection threshold of glucose ≤ 1 μM ) and to prevent import of residual glucose , a non-metabolised glucose analogue inhibiting glucose transport ( N-acetyl-D-glucosamine , 50 mM ) was added in the medium [21] . The same protocol was followed to produce SDM79-GlcFree and proline-depleted medium , except that the starting glucose-depleted SDM79 medium containing 100 μM proline , instead of 5 mM , was diluted with one volume of proline/glucose-depleted SDM79 . Considering the rate of proline consumption of PCF trypanosomes , we estimated that the proline-depleted SDM79-GlcFree medium contains less than 20 μM proline . To estimate the cell density by the Alamer Blue assay , cells diluted in 200 μl of proline-depleted SDM79-GlcFree medium in which 1 μM to 100 mM proline was added by serial two-fold dilutions at a 2 x 106 cell density , were incubated for 48 h at 27°C in microplates , before adding 20 μl of 0 . 49 mM Alamar Blue ( Resazurin ) . Measurement of fluorescence was performed with the microplate reader Fluostar Optima ( BMG Labtech ) at 550 nm excitation wavelength and 600 nm emission wavelength . The Δppdk mutant cell line was obtained by replacing both alleles of the PPDK gene ( Tb927 . 11 . 3120 , http://www . genedb . org/genedb/tryp/ ) by two different plasmids encoding the hygromycin resistance gene and the T7 RNA polymerase gene for the first one , and the neomycin resistance gene and the tetracycline repressor gene under the control of the T7 RNA polymerase promoter for the second one , as described before [49] . Replacement of the PEPCK gene ( Tb927 . 2 . 4210 ) by the blasticidin ( BSD ) and puromycin ( PAC ) resistance markers via homologous recombination was described before ( Δpepck cell line ) [22] . The Δppdk/Δpepck double mutant was also described before [12] . Replacement of both alleles of the FBPase gene ( Tb927 . 9 . 8720 ) by the blasticidin ( BSD ) and puromycin ( PAC ) resistance markers via homologous recombination was performed with DNA fragments containing a resistance marker gene flanked by the FBPase UTR sequences . Briefly , the pGEMt plasmid was used to clone an HpaI DNA fragment containing the BSD or PAC resistance marker gene preceded by the FBPase 5'-UTR fragment ( 729 bp ) and followed by the FBPase 3'-UTR fragment ( 694 bp ) . The PAC resistance marker replaced one allele of the FBPase gene via homologous recombination and BSD resistance marker replaced the second allele . The EATRO1125 . T7T parental cell line , which constitutively expresses the T7 RNA polymerase gene and the tetracycline repressor under the control of a T7 RNA polymerase promoter for tetracycline inducible expression ( TetR-HYG T7RNAPOL-NEO ) [50] , was used to generate the FBPase knockout cell line . Transfection and selection of drug-resistant clones were performed as reported previously [11] . Transfected cells were selected in SDM79 medium containing hygromycin ( 25 μg ml-1 ) , neomycin ( 10 μg ml-1 ) , blasticidin ( 10 μg ml-1 ) and/or puromycin ( 1 μg ml-1 ) . The selected cell line TetR-HYG T7RNAPOL-NEO Δfbpase::BSD/Δfbpase::PAC is called Δfbpase . To produce the add-back Δfbpase/FBPase rescue cell lines , a FBPase ectopic copy under control of the PARP promoter in the pLew100 expression vector , which contains the phleomycin resistance gene ( kindly provided by E . Wirtz and G . Cross ) [51] , was introduced in the Δfbpase null background . The pLew100-FBPase plasmid was generated by introduction the full length FBPase gene in the HindIII and BamHI restriction sites of the vector ( Genecust ) . The Δfbpase/FBPase cell line was selected in SDM79 medium containing phleomycin ( 5 μg ml-1 ) in addition to the 4 other antibiotics used to select the Δfbpase cell line . For experimental infections of Δtsetse flies , the Δfbpase null mutant was generated in the pleomorphic AnTat1 . 1E PCF strain , using the same approach described above . Then , two distinct strategies were used to produce the add-back Δfbpase/FBPase rescue cell lines in the AnTat1 . 1E Δfbpase null background . The Δfbpase/FBPase-1 cell line expresses a FBPase ectopic copy under control of the PARP promoter in the pLew100 expression vector as described above for the EATRO1125 Δfbpase/FBPase cell line . The Δfbpase/FBPase-2 cell line was generated by in situ re-insertion of one FBPase gene under the control of its endogenous UTRs . The 3 . 5 kb DNA fragment for in situ re-insertion consists on the FBPase coding sequence preceded by its 5' intergenic region ( 722 bp ) and followed by its 3' intergenic region ( 662 bp ) , the phleomycin resistance gene ( 375 bp ) and a second copy of the FBPase 3' intergenic region . The 3 . 5 kb DNA fragment flanked by two HpaI restriction sites was introduced into the pBluescript vector ( Genecust ) and the resulting plasmid was digested by HpaI before transfection of the AnTat1 . 1E Δfbpase cell line . The Δfbpase/FBPase-1 and Δfbpase/FBPase-2 cell lines were selected in SDM79 medium containing blasticidin ( 10 μg ml-1 ) , puromycin ( 1 μg ml-1 ) and phleomycin ( 5 μg ml-1 ) . Accession numbers of genes targeted by RNAi are as follows; fructose-1 , 6-bisphosphatase ( FBPase , Tb927 . 9 . 8720 ) , glycerol kinase ( GK , Tb927 . 9 . 12550-Tb927 . 9 . 12630 ) , glycerol-3-phosphate dehydrogenase ( GPDH , Tb927 . 8 . 3530 ) and triosephosphate isomerase ( TIM , Tb927 . 11 . 5520 ) . RNAi-mediated inhibition of gene expression was performed in the EATRO1125 . T7T PCF by expression of stem-loop “sense-antisense” RNA molecules of the targeted sequences [50 , 52] using the pLew100 expression vector [51] . To inhibit by RNAi the expression of the FBPase gene , a 597-bp fragment of the FBPase gene ( from position 152 to 749 ) was introduced in the pLew100 vector to produce the pLew-FBPase-SAS plasmid . Briefly , a PCR-amplified 716-bp fragment , containing the antisense FBPase sequence with restriction sites added to the primers , was inserted into the HindIII and BamHI restriction sites of the pLew100 plasmid . The separate 615-bp PCR-amplified fragment containing the sense FBPase sequence was then inserted upstream of the antisense sequence , using HindIII and XhoI restriction sites ( XhoI was introduced at the 3'-extremity of the antisense PCR fragment ) . The resulting plasmid pLew-FBPase-SAS contains a sense and antisense version of the targeted gene fragment , separated by a 89-bp fragment , under the control of a PARP promoter linked to a prokaryotic tetracycline operator . The same strategy was used to produce the pLew-GK-SAS , pLew-GPDH-SAS and pLew-TIM-SAS plasmids designed to inhibit the expression of the GK , GPDH and TIM genes , respectively . The size of the GK , GPDH and TIM targeted sequences are 617 bp ( from position 460 to 1077 ) , 622 bp ( from position 218 to 840 ) and 558 bp ( from position 67 to 625 ) , respectively . The RNAiFBPase mutant was generated by transfecting the EATRO1125 . T7T parental cell line with the pLew-FBPase-SAS plasmid and selection in glucose-rich SDM79 medium containing hygromycin ( 25 μg ml-1 ) , neomycin ( 10 μg ml-1 ) and phleomycin ( 5 μg ml-1 ) . The triple mutant cell lines ( Δppdk/Δpepck/RNAiGK , Δppdk/Δpepck/RNAiGPDH and Δppdk/Δpepck/RNAiTIM ) were obtained by introducing the relevant plasmid in the Δppdk/Δpepck cell line , after selection in glucose-rich SDM79 medium containing hygromycin ( 25 μg ml-1 ) , neomycin ( 10 μg ml-1 ) , phleomycin ( 5 μg ml-1 ) , blasticidin ( 10 μg ml-1 ) and puromycin ( 1 μg ml-1 ) . Aliquots were frozen in liquid nitrogen to provide stocks of each line that had not been cultivated long term in medium . Induction of RNAi cell lines was performed by addition of 1 μg ml-1 tetracycline . Subcellular fraction enriched in glycosomes was prepared by differential centrifugation as described in [53] , after homogenising the cells with silicon carbide as grinding material . Briefly , 2 x 109 cells were washed once in 10 ml of STE ( 25 mM Tris , 1 mM EDTA , 250 mM sucrose pH 7 . 8 ) . After centrifugation , the pellet was resuspended in 0 . 15 ml of homogenisation buffer STE ( STE supplemented with ‘Complete EDTA-Free’ protease-inhibitor cocktail , Roche Applied Science , Mannheim , Germany ) and ground in a pre-chilled mortar with 1 . 5 gr of wet-weight silicon carbide per gr of cell pellet . The cells were microscopically checked for at least 90% disruption . The cell lysate was diluted in 7 ml of homogenisation buffer , centrifuged at 1 , 000 g and then at 5 , 000 g for 10 min each , at 4°C . The supernatant was centrifuged at 33 , 000 g for 10 min at 4°C to yield the cytosolic fraction and the glycosome-enriched pellet , which was washed once with 1 ml of STE , centrifuged at 33 , 000 g for 10 min at 4°C before resuspension in 0 . 1 ml of STE . The FBPase activity of aliquots of the glycosomal fractions was measured following the reduction of NADP+ at 340 nm in the presence of 100 mM triethanolamine ( pH 7 . 6 ) , 2 mM MgCl2 , 0 . 1 mM EDTA , 0 . 3 mM NADP+ , 10 mM F1 , 6BP , 25 μg ml-1 glucose-6-phosphate isomerase and 25 μg ml-1 glucose-6-phosphate dehydrogenase . The glycerol kinase activity was measured at 340 nm via oxidation of NADH according to published procedures [54] . Total protein extracts ( 5 x 106 cells ) , or glycosomal and cytosolic fractions , of parental ( AnTat1 . 1E , EATRO1125 . T7T ) or mutant PCF of T . brucei were separated by SDS-PAGE ( 10% ) and immunoblotted on TransBlot Turbo Midi-size PVFD Membranes ( Bio-Rad ) [55] . Immunodetection was performed as described [55 , 56] using as primary antibodies , the rabbit anti-FBPase ( 1:1 , 000 , gift from P . Michels , Edinburgh , UK ) , the rabbit anti-SBPase ( 1:250 , gift from P . Michels , Edinburgh , UK ) , the rabbit anti-GK ( 1:5 , 000 , gift from P . Michels , Edinburgh , UK ) , the rabbit anti-G6PDH ( 1:1 , 000 , gift from P . Michels , Edinburgh , UK ) , the rabbit anti-TIM ( 1:1 , 000 , gift from P . Michels , Edinburgh , UK ) , the rabbit anti-PFR ( 1:10 , 000 ) , the rabbit anti-enolase ( 1:100 , 000 , gift from P . Michels , Edinburgh , UK ) , the rat anti-PEPCK ( 1:1 , 000 , gift from T . Seebeck , Bern , Switzerland ) [22] , the rabbit anti-GPDH ( 1:1 , 000 ) [57] , the rabbit anti-hsp60 ( 1:10 , 000 ) [58] , the rabbit anti-PPDK ( 1:1 , 000 ) [11] , anti-EP mouse monoclonal TBRP1/247 ( 1:500 ) ( Biozol ) , anti-calflagin ( 1:1 , 000 , gift of D . Engman , Chicago , USA [59] ) and the rabbit anti-RBP6 ( 1:1 , 000 , gift from C . Tschudi , New Haven , USA ) [30] . Anti-rabbit or anti-rat IgG conjugated to horseradish peroxidase ( Bio-Rad , 1:5000 dilution ) was used as secondary antibody . Revelation was performed using the Clarity western ECL Substrate as described by the manufacturer ( Bio-Rad ) . Images were acquired and analysed with the ImageQuant LAS 4000 luminescent image analyser . The RBP6 gene was amplified by PCR and cloned via HindIII/BamHI into the pLew100v5b1d vector ( pLew100v5 modified with a blasticidin resistance gene BSD ) . The Δfbpase cell and its parental EATRO1125 . T7T cell line were transfected with the pLew100v5-RBP6 linearised with NotI in pools to generate the Δfbpase RBP6 cell line with the genotype TETR::HYG T7RNAP::NEO Δfbpase::BLAS/Δfbpase::PAC RBP6Ti::BLE and RBP6 , respectively . In vitro differentiation experiments were done as described in [30] in SDM79 medium without glucose in the presence of 50 mM N-acetyl-D-glucosamine and 10% ( v/v ) heat-inactivated fetal calf serum . The faster kinetics of differentiation observed in our experiments as compared to that described in [30] is likely due to the different parental backgrounds ( Lister 427 29:13 versus EATRO1125 . T7T ) or to the different media ( Cunningham’s medium [60] versus modified SDM79 ) that were used in the two labs . 2 × 107 cells were harvested and fixed overnight with 2% paraformaldehyde in PBS at 4°C , washed thrice with PBS and resuspended in 500 μl PBS . Cells were incubated with monoclonal mouse anti-procyclin EP ( TBRP1/247 , 1:500 ) in 1% BSA . For calflagin detection 3 × 107 cells were permeabilised with 0 . 2% NP-40 for 5 min prior incubation with polyclonal mouse α-calflagin ( 1:1 , 000 ) [59] in 1% BSA . Alexa Fluor 488-conjugated goat antibodies were used as secondary antibodies . Samples were analysed with a FACSCalibur cell analyser ( Becton Dickinson ) and data were evaluated with the FlowJo 8 . 8 . 6 software . EATRO1125 . T7T parental and mutant cell lines grown in SDM79 medium were washed twice with PBS and resuspended in PBS containing [U-13C]-proline ( or [U-13C]-glycerol ) with or without the same amount of non-labelled glucose . The cells were incubated for 2 h at 27°C before being collected on filters by fast filtration preparation ( 2 x 107 cells per filter ) for mass spectrometry analysis , as described before [22] . Metabolites were analysed by ion-exchange chromatography coupled with tandem mass spectrometry ( IC-MS/MS ) using the method described by Bolten et al . [61] . Retention time on the column and multiple reactions monitoring ( MRM ) transition of each analysed metabolite are shown in Table 1 of [21] . The 13C mass isotopomer distribution of intracellular metabolites was determined from relevant isotopic clusters in the IC-MS/MS analysis , according to Kiefer et al . [62] . 13C mass isotopomer distribution measurements were performed using a triple quadrupole mass spectrometer ( 4000Qtrap , Applied Biosystems ) . To obtain [13C]-labeling patterns ( 13C isotopologues ) , isotopic clusters were corrected for the natural abundance of isotopes other than 13C , using the in-house software IsoCor ( available at MetaSys ) [63] . 1 x 108 T . brucei PCF were collected by centrifugation at 1 , 400 g for 10 min , washed once with phosphate-buffered saline ( PBS ) and incubated in 5 ml of PBS supplemented with 2 g l-1 NaHCO3 ( pH 7 . 4 ) . Cells were maintained for 6 h at 27°C in incubation buffer containing 20 μmol of D-glucose or [U-13C]-glycerol ( 4 mM ) . The integrity of the cells during the incubation was checked by microscopic observation . The supernatant was collected and 50 μl of maleate solution in D2O ( 10 mM ) was added as internal reference . H-NMR spectra were performed at 500 . 19 MHz on a Bruker Avance III 500 HD spectrometer equipped with a 5 mm cryoprobe Prodigy . Measurements were recorded at 25° . Acquisition conditions were as follows: 90° flip angle , 5 , 000 Hz spectral width , 32 K memory size , and 9 . 3 sec total recycle time . Measurements were performed with 64 scans for a total time close to 10 min 30 sec . Resonances of the obtained spectra were integrated and metabolites concentrations were calculated using the ERETIC2 NMR quantification Bruker program . Tsetse flies ( Glossina morsitans morsitans ) were maintained , infected and dissected at the Institut Pasteur as previously described [28] . Teneral males were collected 24 h to 48 h post-eclosion and artificially fed through a silicone membrane with 6–9 x 106 parasites ml-1 in SDM79 medium supplemented with 10% FCS for their first meal . Flies were then maintained in Roubaud cages for one month at 26°C and 60% hygrometry and fed twice a week with mechanically defibrinated sheep blood . Flies were starved for at least 48 h before being individually dissected 28 days after ingestion of the infected meal . Salivary glands were first rapidly dissected into a drop of PBS . The whole tsetse alimentary tract was then dissected and arranged lengthways for assessment of parasite presence . The proventriculus and anterior midgut were physically separated from the posterior midgut in a distinct PBS drop . Tissues were dilacerated to allow parasites to spread in PBS , parasites were recovered and treated for further experiments no more than 15 min after dissection . For immunofluorescence , parasites were rapidly allowed to settle on poly-lysine coated slides until drying . Cells were fixed for 10 sec in methanol at -20°C and re-hydrated in PBS during 10 min . Slides were then incubated for 45 min at 37°C with the anti-FBPase primary antibody diluted at 1:10 in PBS with 0 , 1% bovine serum albumin . Slides were washed in PBS and incubated for 30 min at 37°C with the appropriate anti-rabbit secondary antibody coupled to an Alexa 488 fluorophore ( Invitrogen ) . Slides were then stained with DAPI for visualisation of their kinetoplast and nuclear DNA contents and mounted under coverslip with Prolong antifade reagent ( Invitrogen ) . IFA experiments were repeated on trypanosomes issued from 2 to 4 flies and from 6 distinct experimental infections . Slides were finally observed with a DMI4000 epifluorescence microscope ( Leica ) and images were captured with a Horca 03G5 camera ( Hamamastu ) . Normalisation of signals was carried out by parallel manipulation of min/max signals in ImageJ ( NIH ) . Statistical analyses and plots were performed with XLSTAT 2015 . 4 . 01 ( Addinsoft ) and Excel 2011 ( Microsoft ) , respectively . Statistical analyses include ANOVA with Tukey ad-hoc post-tests for inter-group comparison with 95% confidence .
|
Trypanosoma brucei , the parasite responsible for sleeping sickness in humans , is transmitted by the tsetse fly that primarily uses amino acids for its energy production . In the glucose-free environment encountered between the insect blood meals , T . brucei needs to produce through gluconeogenesis glucose 6-phosphate , a key precursor for several essential metabolic pathways . We have shown here that two key gluconeogenic steps , which produce phosphoenolpyruvate and fructose 6-phosphate , respectively , are performed by redundant enzymes ( PPDK and PEPCK for phosphoenolpyruvate production; FBPase and a yet unknown enzyme for fructose 6-phosphate production ) , which highlights the importance of this metabolic pathway for the insect stages of the parasite . Interestingly , deletion of the parasite FBPase gene abolished both the colonisation of the insect salivary glands and the in vitro differentiation of the epimastigote forms into the mammalian infective form of the parasite . Altogether , these data demonstrate for the first time that gluconeogenesis is essential for development of T . brucei in its insect vector and that early development stages of the parasite present in the tsetse midgut are not affected by the absence of FBPase , probably by developing an alternative yet unknown approach to produce fructose 6-phosphate .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"carbohydrate",
"metabolism",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"monomers",
"phosphates",
"carbohydrates",
"glucose",
"metabolism",
"organic",
"compounds",
"parasitic",
"protozoans",
"glucose",
"parasitic",
"diseases",
"protozoans",
"amino",
"acids",
"digestive",
"system",
"polymer",
"chemistry",
"proteins",
"chemistry",
"cyclic",
"amino",
"acids",
"exocrine",
"glands",
"proline",
"biochemistry",
"trypanosoma",
"eukaryota",
"organic",
"chemistry",
"anatomy",
"glycerol",
"monosaccharides",
"salivary",
"glands",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"metabolism",
"organisms"
] |
2018
|
Gluconeogenesis is essential for trypanosome development in the tsetse fly vector
|
The nonsense-mediated decay ( NMD ) pathway subjects mRNAs with premature termination codons ( PTCs ) to rapid decay . The conserved Upf1–3 complex interacts with the eukaryotic translation release factors , eRF3 and eRF1 , and triggers NMD when translation termination takes place at a PTC . Contrasting models postulate central roles in PTC-recognition for the exon junction complex in mammals versus the cytoplasmic poly ( A ) -binding protein ( PABP ) in other eukaryotes . Here we present evidence for a unified model for NMD , in which PTC recognition in human cells is mediated by a competition between 3′ UTR–associated factors that stimulate or antagonize recruitment of the Upf complex to the terminating ribosome . We identify cytoplasmic PABP as a human NMD antagonizing factor , which inhibits the interaction between eRF3 and Upf1 in vitro and prevents NMD in cells when positioned in proximity to the termination codon . Surprisingly , only when an extended 3′ UTR places cytoplasmic PABP distally to the termination codon does a downstream exon junction complex enhance NMD , likely through increasing the affinity of Upf proteins for the 3′ UTR . Interestingly , while an artificial 3′ UTR of >420 nucleotides triggers NMD , a large subset of human mRNAs contain longer 3′ UTRs but evade NMD . We speculate that these have evolved to concentrate NMD-inhibiting factors , such as PABP , in spatial proximity of the termination codon .
The process of nonsense-mediated decay ( NMD ) subjects mRNAs with premature termination codons ( PTCs ) to rapid decay . This helps rid the cell of aberrant mRNAs that have acquired PTCs through mutation or faulty processing [1–3] . Moreover , several lines of evidence suggest that NMD is also used as a posttranscriptional mechanism of normal gene regulation [4] . The NMD pathway employs a set of factors that are conserved amongst eukaryotes . Central to the NMD pathway is the Upf complex , which consists of the proteins Upf1 , Upf2 , and Upf3 [1–3] . The Upf complex interacts with the eukaryotic translation release factors , eRF3 and eRF1 , and triggers NMD when translation termination takes place at a PTC [1–3] . In addition , the Smg proteins , which are conserved in metazoans , regulate Upf1 function by phosphorylation and dephosphorylation [2 , 3] . A fundamental question is how mRNAs with PTCs are distinguished from those with normal termination codons . Despite the conservation of core NMD factors , contrasting models have been proposed in mammalian cells as opposed to other eukaryotes . Evidence in Saccharomyces cerevisiae and in cell lines from Drosophila melanogaster suggests that termination codons are recognized as PTCs when positioned too far upstream of the poly ( A ) tail [5–7] . This is thought to be a consequence of an impaired interaction between eRF3 at the terminating ribosome and factors associated with the normal 3′ UTR , including cytoplasmic poly ( A ) -binding protein ( PABP ) [1 , 5 , 7] , which on mRNAs with regular stop codons ( proximal to the poly ( A ) tail ) stimulates normal translation termination [8] . Consistent with this model for NMD , termed the “faux 3′ UTR” model [1 , 7] , 3′ UTRs of S . cerevisiae and D . melanogaster mRNAs are generally short , on average ∼100 and ∼330 nucleotides in length , respectively [9 , 10] . Interestingly , recent observations show evidence that cytoplasmic PABP is not required for the discrimination of normal termination codons from PTCs in S . cerevisiae [11] . Thus , cytoplasmic PABP may function redundantly with other 3′ UTR–associated factors to antagonize NMD . 3′ UTRs of human mRNAs are on average longer ( ∼750–800 nucleotides [12] ) than those of S . cerevisiae and D . melanogaster , and current models for NMD in mammalian cells do not involve the length of the 3′ UTR . Rather , the exon junction complex ( EJC ) , which is deposited 20–25 nucleotides upstream of mRNA exon-exon junctions after pre-mRNA splicing [13] , is thought to play a central role . A termination event more than ∼30 nucleotides upstream of one or more EJCs is thought to trigger NMD through EJC-mediated recruitment of the Upf complex [2 , 3] . This is consistent with observed direct interactions between EJC components and Upf3 proteins [14–18] . However , the EJC plays no apparent role in NMD in D . melanogaster [19] or in Caenorhabditis elegans [20] and no evidence for the existence of an EJC has been reported in yeast . Nevertheless , a conceptually similar model to the EJC model was proposed earlier for NMD of the PGK1 mRNA in yeast , in which a “downstream sequence element” ( DSE ) , when present downstream of a termination codon , promotes NMD through recruitment of the protein Hrp1p , which interacts with Upf proteins [21 , 22] . A fundamental difference between the faux 3′ UTR and the EJC/DSE models for NMD is that the EJC/DSE models propose that NMD-stimulating factors ( the EJC and Hrp1p , respectively ) trigger NMD when positioned downstream of a termination codon , whereas the faux 3′ UTR model postulates that NMD is caused instead by the absence of NMD-antagonizing factors , such as cytoplasmic PABP , which normally positively influence translation termination and mRNA stability . Here , we present evidence for a merged model for NMD in human cells , which likely can be extended to other eukaryotes . According to this model , PTC recognition is determined by a competition between 3′ UTR–associated factors , which stimulate ( including the EJC ) or antagonize ( including cytoplasmic PABP ) the recruitment of the Upf complex to the terminating ribosome . Our observations suggest that the fundamental principles of the NMD pathway are much more conserved between mammals and other eukaryotes than previously anticipated .
The EJC model for human NMD postulates that any translation termination event taking place >50–55 nucleotides upstream of an exon-exon junction should result in NMD . However , during our studies of the human NMD pathway , we observed that a β-globin mRNA , in which the adenovirus major late ( AdML ) intron was inserted into the 3′ UTR 175 nucleotides downstream of the normal β-globin mRNA translation termination codon , did not show enhanced mRNA decay as compared to the wild-type β-globin mRNA in human HeLa Tet-off cells ( compare Figure 1A and 1B ) . Moreover , in contrast to a well-characterized β-globin NMD substrate , which contains a premature termination codon at position 39 ( β39; Figure 1C ) , the AdML intron containing β-globin mRNA is not stabilized when the central NMD factor hUpf1 is knocked down ( Figure 1A , middle panel and Figure S1 ) or when a point mutation causes translation termination to take place downstream of the inserted intron ( Figure 1A , bottom panel ) . Sequencing of cDNAs derived from the mRNAs in Figure 1A revealed the expected splicing patterns ( unpublished data ) . Thus , in contrast to the prediction from the EJC model for human NMD , the AdML intron is not sufficient for triggering NMD when positioned in the 3′ UTR of β-globin mRNA , even though the AdML intron has been previously demonstrated to recruit an EJC [13] . The observation in Figure 1A was surprising , because a β-globin mRNA in which the MINX-intron had been placed in the 3′ UTR was previously found to cause reduced mRNA steady-state levels [23] . Therefore , to rule out the possibility that the observations in Figure 1A represent an unusual property of the specific mRNA reporter , we tested two other substrates . As seen in Figure 1D and 1E , insertion of the triosephosphate isomerase ( TPI ) mRNA intron 6 or the AdML intron , 140 or 149 nucleotides downstream of the termination codons of β-globin or TPI mRNAs , respectively , failed to cause hUpf1-dependent mRNA decay , despite the previously demonstrated ability of each of these introns to recruit an EJC [13] . Sequencing of cDNAs derived from the tested mRNAs revealed the expected splicing patterns ( unpublished data ) , although a minor fraction of the βTPIi6 mRNA fails to remove the TPI intron ( see asterisk in Figure 1D ) . We attempted to test the β-globin mRNA with the MINX-intron in the 3′ UTR , which was previously found to accumulate at reduced steady-state levels as compared to wild-type β-globin mRNA [23] ( construct generously provided by A . Kulozik and M . Hentze ) . However , the MINX intron ( as well as a number of other introns tested in this study ) failed to be spliced out of the β-globin mRNA 3′ UTR in the HeLa Tet-off cells used here ( unpublished data ) . We conclude that a 3′ UTR intron is not sufficient to trigger NMD in human cells . This contradicts an EJC-centric model for human NMD . Our observation that 3′ UTR introns are not sufficient for triggering human NMD spurred us to test whether cytoplasmic PABP may antagonize NMD in human cells , as it does in S . cerevisiae and D . melanogaster . We therefore first manipulated the position of the poly ( A ) tail relative to the termination codons of β-globin and TPI reporter mRNAs and tested the effect on mRNA decay . As seen in Figure 2A and 2B and Figure S2 , artificial extension of the 3′ UTRs of β-globin or TPI mRNAs ( from 292 and 447 nucleotides , respectively , to 846–1 , 112 nucleotides ) through insertion of fragments of glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) ( Figures 2A and S2 ) or green fluorescent protein ( GFP ) ( Figure 2B ) mRNAs results in mRNA destabilization ( compare with Figures 1B and S2 , top panel ) . This is due to NMD because depletion of the NMD factors hUpf1 or hUpf2 stabilizes the mRNAs ( Figure 2A and 2B and Figures S1 and S2 ) . Moreover , the introduction of single point mutations in the termination codons that results in termination on the same mRNAs in proximity ( 180–357 nucleotides upstream ) of the poly ( A ) tail , results in mRNA stabilization ( Figure 2A and 2B and S2 , bottom panels ) . Even though sequencing of cDNAs derived from the tested mRNAs revealed no cryptic splicing in the extended 3′ UTRs ( unpublished data ) , depletion of the central EJC component eIF4AIII results in stabilization of the βGAP mRNA ( Figure S3 ) , possibly reflecting the ability of the EJCs in the β-globin mRNA open reading frame to stimulate translation as has been previously observed [24 , 25] . Successive shortening of the 3′ UTR of the βGAP mRNA revealed that a 3′ UTR as short as 422 nucleotides can trigger NMD ( Figure S4 ) . This is surprising because a large fraction of human mRNAs contain 3′ UTRs longer than 422 nucleotides [12] . We conclude that artificially extended 3′ UTRs trigger NMD in human cells . This is consistent with recent reports in which steady-state levels of PTCs containing TPI , β-globin , and Ig-μ reporter mRNAs lacking 3′ UTR introns were measured [26–28] , and with observations using unspliced Rous sarcoma virus RNAs in chicken cells [29] . Thus 3′ UTR introns are neither necessary ( Figure 2 ) nor sufficient ( Figure 1 ) for human NMD . Having observed that 3′ UTR introns are not required for NMD , we asked whether a completely intron-less mRNA can undergo NMD . It was observed previously that introduction of PTCs in the naturally intron-less Hsp70 and histone H2A mRNAs does not result in their decreased steady-state levels , which led to speculations that intron-less mRNAs are immune to NMD [30] . However , it has been pointed out that wild-type Hsp70 and histone H2A mRNAs are both highly unstable and may thus not be further destabilized by a PTC [5] . We therefore tested the stability of three naturally occurring intron-less mRNAs ( encoding eRF3b , SFN , and TBCC ) and found that both wild-type and PTC containing versions of these mRNAs were unstable ( ∼100- to 150-min half-lives , unpublished data ) . Thus , mRNA instability may be a general feature of natural intron-less mRNAs . However , when the only intron in the Glutathione Peroxidase 1 ( GPx1 ) mRNA is removed , introduction of a PTC triggers NMD , although not as efficiently as in the presence of the intron ( Figure 2C , compare lower and upper panels ) . Thus , neither 3′ UTR introns nor internal introns are essential for human NMD . However , similarly to a previous report [30] we have not been able to identify a natural human intron-less mRNA for which NMD could be observed , perhaps due to the observed inherent instabilities of the tested mRNAs . To more directly test whether cytoplasmic PABP antagonizes NMD in human cells , we examined the effect of positioning cytoplasmic PABP in proximity of a PTC using two different approaches . First , as seen in Figure 3A and 3B , when an MS2-PABPC1 fusion protein ( PABPC1 is one of five human cytoplasmic PABPs [31] ) is artificially tethered downstream of a PTC in two different β-globin NMD reporter mRNAs , a partial rescue of NMD is observed . This rescue is due to tethered PABPC1 , because similar levels of unfused PABPC1 ( Figure 3A ) or MS2 coat protein ( Figure 3A and 3B ) do not stabilize the mRNAs . Moreover , tethering of the nuclear poly ( A ) -binding protein PABPN1 does not rescue NMD ( Figure 3A ) even though it is expressed at levels similar to MS2-PABPC1 ( Figure S5 ) . The efficiency of the rescue from NMD by tethered PABPC1 decreases as the MS2 binding sites are moved more distal to the PTC in the β-globin PTC-39 mRNA ( Figure S6 ) . As a second independent approach to ask whether PABP can antagonize human NMD , we tested the effect of inserting a binding site for PABP downstream of the PTCs . As seen in Figure 3C and 3D , inclusion of a poly-A30-stretch , but not that of a random 30-nucleotide stretch , 182 or 230 nucleotides downstream of the PTC of two different NMD reporter mRNAs , results in partial rescue of NMD . Thus , similarly to S . cerevisiae [7] and D . melanogaster [5] , cytoplasmic PABP can antagonize NMD in human cells when placed in proximity of a PTC . Recent observations suggest that while cytoplasmic PABP can antagonize NMD in S . cerevisiae , it is not required for discriminating a normal mRNA from an NMD substrate [11] . Attempts at testing whether PABPC1 is required for preventing NMD in human cells failed because HeLa Tet-off cells became detached from plates upon short interfering RNA ( siRNA ) -mediated PABPC1 depletion ( unpublished data ) . Our observations raise the question of whether naturally occurring mammalian mRNAs with long 3′ UTRs , which can be several kilobases in length , are normal targets of NMD or whether they have evolved mechanisms to evade the NMD pathway . We noted that mRNAs identified by microarray assays to be upregulated upon hUpf1 knockdown in HeLa cells [32] contain on average significantly longer 3′ UTRs than those mRNAs unaffected by hUpf1 knockdown ( Figure S7 ) . Moreover , the majority of these 3′ UTRs ( 75% ) are longer than the ∼420 nucleotides observed here to trigger NMD in the βGAP reporter mRNA ( Figures S4 and S7 ) . It is possible that at least a subset of these transcripts undergo NMD due to an increased distance between the termination codon and the poly ( A ) tail . Indeed , when the 1 , 342-nucleotide 3′ UTR of one of these mRNAs , encoding hSmg5 , is replaced for the β-globin 3′ UTR ( βSmg5 ) , the chimeric mRNA undergoes NMD ( Figure 4 , top two panels ) . Thus , the Smg5 3′ UTR stimulates NMD , and a subset of mRNAs may have evolved long 3′ UTRs to be regulated by the NMD pathway . However , numerous mRNAs with long 3′ UTRs are not upregulated upon hUpf1 knockdown [32] ( Figure S7 ) . When the 3′ UTRs from two such mRNAs , Cript1 and Tram1 , were inserted into the β-globin mRNA , no NMD was observed ( Figure 4 , bottom panels ) . This is in sharp contrast to the observations using artificial long 3′ UTRs ( compare to Figures 1 and S4 ) and suggests that the ability of a subset of endogenous long 3′ UTRs to evade NMD is an acquired property ( see Discussion ) . Our observations that 3′ UTR introns are neither necessary ( Figure 2 ) nor sufficient ( Figure 1 ) for human NMD raises the question of whether introns play any role in human NMD . We therefore tested the effect of inserting the AdML intron into the 3′ UTR of an mRNA , which already undergoes NMD due to an extended 3′ UTR . Interestingly , insertion of the AdML intron downstream of the termination codon of βGAP mRNA results in enhanced mRNA decay ( Figure 5; βGAP-AdML mRNA ) . This effect is only observed when the intron is positioned in the 3′ UTR , as insertion of the same intron upstream of the termination codon ( without disrupting the open reading frame ) does not enhance mRNA decay ( Figure 5; βAdML-GAP mRNA ) . Thus , while a downstream intron is neither sufficient nor necessary for triggering NMD in human HeLa cells ( Figures 1 and 2 ) [33] , it can enhance the degradation of an mRNA that is already a target of NMD due to an extended 3′ UTR ( Figure 5A and 5B ) . Consistent with this , the presence of an intron appears to also stimulate NMD of a PTC containing GPx1 mRNA ( Figure 2C ) . How does cytoplasmic PABP antagonize NMD when positioned in proximity of the termination codon ? Both cytoplasmic PABP and Upf1 have been previously observed to stimulate translation termination in yeast cells [8 , 34] and to associate with translation release factor eRF3 [35–40] . This raised the possibility that cytoplasmic PABP inhibits NMD by preventing Upf1 from interacting with eRF3 and the terminating ribosome . As seen in the co-immunoprecipitation ( co-IP ) assays in Figure 6A , endogenous hUpf1 and PABPC1 can both be observed in complex with eRF3 in RNase-treated HeLa cell extracts . However , PABPC1 co-IPs much more efficiently than hUpf1 with eRF3 ( Figure 6A ) , in spite of comparable estimated number of molecules of cytoplasmic PABP ( ∼8 × 106/cell ) and hUpf1 ( ∼3 × 106/cell ) in HeLa cells [41 , 42] . Consistent with this , bacterially expressed GST-tagged eRF3 was found to associate much more efficiently with rabbit reticulocyte-lysate-translated PABPC1 ( Kd ∼ 5 nM ) than hUpf1 ( Kd > 1 μM ) ( unpublished data ) . To test whether PABPC1 can antagonize the interaction between eRF3 and hUpf1 in vitro , we immunopurified transiently expressed epitope-tagged eRF3 , PABPC1 , and hUpf1 proteins from HEK 293T cells and tested the ability of hUpf1 to associate with eRF3 in the presence of increasing amounts of PABPC1 . As seen in Figure 6B , in contrast to the negative control protein hnRNP A1 , increasing amounts of PABPC1 efficiently prevent the interaction between hUpf1 and eRF3 , even when hUpf1 is present in 10- to 40-fold excess over PABPC1 ( Figure 6B , compare lanes 5 and 6 with lanes 4 and 7 ) . Thus , PABPC1 can antagonize the interaction between hUpf1 and eRF3 in vitro . However , no reduction in the co-IP efficiency between hUpf1 and eRF3 was observed upon transient over-expression of FLAG-tagged PABPC1 in HeLa or HEK 293T cells ( unpublished data ) . Thus , either exogenous PABPC1 failed to express at adequate levels to observe a competition in cells , or the relation between hUpf1 , PABPC1 , and eRF3 is more complex in cells than it is in vitro . To test whether amino acid residues of eRF3 , which are important for cytoplasmic PABP interaction , are also important for the interaction with hUpf1 , we constructed a eRF3 protein ( eRF3 KAKA ) mutated in four N-terminal residues that are conserved between cytoplasmic PABP-binding proteins [38 , 43] . As seen in the co-IP assays in Figure 6C , the exogenously expressed eRF3 KAKA mutant protein is equally impaired in interaction with PABPC1 and hUpf1 ( Figure 6C , compare lanes 5 and 6 with lanes 2 and 3 ) . As a control , the mutant eRF3 KAKA protein associates with eRF1 with similar affinity as wild-type eRF3 , suggesting that these mutations do not cause gross structural alterations , although local changes cannot be ruled out . Thus , hUpf1 and PABPC1 interact with a similar , though not necessarily overlapping region of eRF3 . The ability of PABPC1 to antagonize the association between hUpf1 and eRF3 in vitro could therefore be a result of a direct competition for eRF3 binding , or of a local structural alteration of eRF3 upon PABPC1 binding , which prevents hUpf1 association .
Previous contrasting models for PTC-recognition in NMD invoke either 3′ UTR–associated factors that stimulate NMD , the EJC in human cells [2 , 44] , and DSE-binding proteins in yeast [21] , or factors that stimulate normal translation termination and antagonize NMD [1 , 45] . Our observations , together with the observations in the paper by Eberle et al . [46] , are consistent with a unified model for human NMD , in which the balance between NMD-antagonizing ( such as PABPC1 ) and NMD-stimulating ( such as the EJC ) factor ( s ) that are associated with the mRNA 3′ UTR , determines whether termination is considered normal or premature ( Figure 7A ) . According to this model , a translation termination event proximal to cytoplasmic PABP ( Figure 3 ) , or other unknown NMD-antagonizing factors , precludes the interaction of hUpf1 with eRF3 ( Figure 6C ) and thus prohibits NMD ( Figure 7A , top ) . By contrast , if hUpf1 associates with eRF3 , NMD ensues ( Figure 7A , bottom ) . This occurs when cytoplasmic PABP , or other inhibitory factors , are spatially distant from the termination event ( Figure 2 ) and is enhanced when a splicing event downstream of a termination codon results in deposition of an EJC ( Figure 5 ) , which provides higher affinity for the hUpf complex ( Figure 7A , bottom ) . However , an exon-exon junction in the 3′ UTR is not sufficient for NMD ( Figure 1 ) . This suggests that a proximal cytoplasmic PABP is dominant over 3′ UTR exon-exon junctions , which is consistent with the observation that the affinity of PABPC1 for eRF3 appears to be several orders of magnitude higher than that of hUpf1 ( Figure 6 and unpublished data ) . However , while introns are observed to only stimulate NMD of the substrates tested in this study , it cannot be ruled out that a subset of human mRNAs requires downstream introns for NMD . Previous experiments , in which EJC or hUpf proteins tethered to an mRNA 3′ UTR were observed to trigger NMD , may have been assisted by the extended 3′ UTRs resulting from insertion of multiple tethering sites and/or by the recruitment of multiple NMD-promoting factors [15 , 17 , 47–49] . The model depicted in Figure 7A may be extended to eukaryotes other than mammals and is consistent with the observation in Drosophila S2 cells that the decay of an NMD reporter mRNA is inhibited upon cytoplasmic PABP depletion [5] . In this case it is predicted that a large subset of normally stable endogenous mRNAs become NMD substrates , thus out-titrating the NMD pathway . How does cytoplasmic PABP antagonize NMD ? While PABPC1 can out-compete the association of hUpf1 with eRF3 in vitro ( Figure 6B ) , a more complex relationship may exist between these proteins in the cell . For example , we failed to observe exogenously expressed PABPC1 out-compete the co-IP of endogenous hUpf1 with eRF3 ( unpublished data ) . Moreover , in S . cerevisiae , cytoplasmic PABP truncated of its C-terminal eRF3-interaction region was capable of suppressing NMD when tethered in proximity of a PTC [7] . However , we found no loss of eRF3-association of a similarly truncated PABPC1 in co-IP assays between exogenously expressed human proteins ( unpublished data ) , suggesting that eRF3 may form a complex with PABPC1 through additional regions . Understanding the specific mechanism by which NMD is antagonized by cytoplasmic PABP , and likely other 3′ UTR–associated factors , is an important goal for future studies and could involve both direct competition with the Upf complex as well as modulation of the translation termination event that excludes Upf complex recruitment in a more indirect manner . Another open question is how the interplay between eRF3 , PABP , and the Upf complex influences events downstream of translation termination . Interestingly , it was previously observed that the interaction between eRF3 and cytoplasmic PABP stimulates mRNA deadenylation in yeast [50] , and that deadenylation can be an early step in NMD [51–53] . Clearly , a great deal remains to be learned about the relationship between eRF3 , the Upf complex , and cytoplasmic PABP and how it controls the fates of mRNAs after translation termination . It is likely that 3′ UTR–associated factors ( indicated by a question mark in Figure 7A ) other than cytoplasmic PABP can antagonize NMD . This hypothesis is consistent with the observation that in yeast cells , cytoplasmic PABP is not required for discriminating tested NMD substrates from their normal counterparts [11] . An excellent candidate for such an activity is the yeast protein Pub1p , which has been identified as a factor that binds downstream of upstream open reading frames ( uORFs ) in GCN4 and YAP1 mRNAs to prevent NMD [54] . It is possible that Pub1p and factors with similar activities are found in a subset of normal 3′ UTRs . It remains to be tested whether Pub1p acts on the terminating ribosome in a manner similar to cytoplasmic PABP . Similarly , factors other than the EJC could provide an enhanced affinity for the Upf complex and stimulate NMD . For example , the protein Hrp1p appears to serve such a role in the yeast PGK1 NMD substrate [21] . Moreover , human Staufen1 and histone mRNA stem loop binding protein have been shown to recruit hUpf1 to the 3′ UTR of specific mRNAs to trigger NMD-like mRNA decay [55 , 56] . Thus , our observations suggest that the NMD pathway is much more conserved between mammals and other eukaryotes than previously appreciated . Nevertheless , there is evidence that differences exist between yeast and mammalian cells as to which round of translation can stimulate NMD [28 , 57–59] . Our observations suggest that while artificial long 3′ UTRs trigger NMD ( Figure 2 ) , a subset of mRNAs containing long 3′ UTRs have evolved mechanisms to evade NMD ( Figure 4 ) . Future studies should reveal the mechanism by which this is accomplished . This could conceivably be achieved by ( i ) induced looping of the 3′ UTR , thus placing the poly ( A ) tail and cytoplasmic PABP in close spatial proximity to the translation termination event ( Figure 7B , top ) , or ( ii ) by recruitment of factors that antagonize NMD ( such as PABPC1 or Pub1p ) to the 3′ UTR in proximity to the termination codon ( Figure 7B , bottom ) . The observation that cytoplasmic PABP alleviates NMD when placed in the vicinity of a PTC ( Figure 3 ) [5 , 7 , 46] and the finding in the paper by Eberle et al . that artificially induced 3′ UTR looping rescues reporter mRNAs with extended 3′ UTRs from NMD [46] , provides proof-of-principle evidence for each of these models . The mechanism by which specific mRNAs evade the NMD pathway is an important subject for future investigation and is likely to vary between individual mRNAs . After the submission of this paper , we have become aware of two other studies reporting that cytoplasmic PABP antagonizes human NMD when placed in proximity to a PTC [60 , 61] .
All plasmid sequences are available upon request . Plasmids expressing different β-globin reporter mRNAs were derived from the pcTET2-βwt plasmid that was constructed by inserting the human β-globin gene between HindIII and ApaI sites of a pcDNA3-based plasmid containing six copies of the Tet-operator sequences upstream of the TATA box . For extended 3′ UTR constructs , parts of the GAPDH mRNA coding sequence and the entire GAPDH 3′ UTR ( pcTET2-βGAP ) or the GFP ORF ( pcTET2-βGFP ) were inserted between NotI and XbaI sites of the pcTET2-βwt plasmid , thus replacing the β-globin 3′ UTR . The β-globin stop codon was mutated to UAC by site-directed mutagenesis to generate pcTET2-βGAP-UAC and pcTET2-βGFP-UAC . Plasmids expressing βGAP-UAC-696 , βGAP-UAC-485 , and βGAP-UAC-422 mRNAs were generated by site-directed mutagenesis of pcTET2-βGAP-UAC to introduce a stop codon ( UAA ) in the GAPDH sequence , respectively , 696 , 485 , or 422 nucleotides upstream of the polyadenylation site . The plasmid expressing βwt mRNA was described earlier [47] . To construct plasmids expressing β39–2xMS2-Ex2 , β39–2xMS2-Ex3 , and β39–2xMS2-3UTR , the 2xMS2 cassette from the previously described plasmid pcβ-2bs [47] was inserted into the BamHI , EcoRI , or NotI sites , respectively , of the pPC-β39 plasmid described earlier [62] . A stretch of A30 ( pPC-β39-A30 ) or N30 ( pPC-β39-N30 ) was inserted into the BamHI site of pPC-β39 plasmid using annealed DNA oligos . Similarly , A30 ( pcTET2-βGAP-A30 ) or N30 ( pcTET2-βGAP-N30 ) was inserted into the XbaI site of the pcTET2-βGAP plasmid . To construct the pcTET2-βGAP-4xMS2 plasmid , four MS2 binding sites were amplified from a previously described plasmid pcβ-4bs [47] and inserted into the XbaI site of pcTET2-βGAP . Plasmids expressing βAdML , βAdML-UAC , and βTPIi6 mRNAs were constructed by cloning the AdML intron or TPI intron 6 ( TPIi6 ) and flanking exon sequences into the XbaI site in pcTET2-βwt or pcTET2-βwt-UAC plasmids . βGAP-AdML and βAdML-GAP mRNA– expressing plasmids were constructed by inserting the same AdML intron into XbaI and EcoRI sites , respectively , in the pcTET2-βwtGAP plasmid . Plasmids expressing chimeric β-globin mRNAs with 3′ UTRs from Smg5 , Cript1 , and Tram1 genes , the respective 3′ UTRs , were cloned into the NotI-XbaI sites of pcTET2-βwt . Plasmids expressing TPI reporter mRNAs were constructed by inserting the entire human TPI gene between HindIII and XbaI sites of the pcTET2 plasmid . A NotI site was inserted into exon 6 ( in a manner that preserved the encoded protein ) by site-directed mutagenesis . Codon 189 was mutated to TGA using site-directed mutagenesis to generate pcTET2-TPI-189 . To remove intron 6 , a NotI-XbaI fragment containing exon6-intron6-exon7 was replaced by the same region amplified from TPI cDNA , to generate pcTET2-TPIΔi6–189 . To extend the TPI 3′ UTR , a fragment containing part of the GAPDH coding region and 3′ UTR was inserted into the NotI site of pcTET2-TPIΔi6–189 to generate pcTET2-TPIΔi6–189-GAP , or into the NotI site of pcTET2-TPIΔi6 to give rise to pcTET2-TPIΔi6-GAP . TPI-AdML mRNA–expressing plasmid was constructed by inserting the AdML intron and flanking exonic sequences into the XbaI site of pcTET2-TPI . The plasmid expressing intron-containing GPx1 mRNA with a PTC ( pPC-GPx1-UAA ) was described earlier [18] . GPx1 cDNA ( HindIII-XbaI ) sequence replaced the intron-containing sequence in pPC-GPx1Δi-UAA . The constructs for knockdowns were based on the pSHAG plasmid ( a gift from Dr . G . Hannon ) and contained inserts expressing precursors to hUpf1 , hUpf2 , or eIF4AIII siRNAs described earlier [63 , 64] . Plasmids expressing FLAG-hUpf1 , FLAG-PABPC1 , FLAG-hnRNP A1 , and Myc-hnRNP A1 were described earlier [47 , 62] . pcDNA3-Myc-eRF3 was constructed by inserting the ORF of eRF3 ( longer isoform ) between BamHI and NotI sites of the pcDNA3-Myc vector previously described [65] . pcDNA3-Myc-eRF3 KAKA was prepared using site-directed mutagenesis ( the mutations are: L66K , N69A , A70K , F73A ) . pcDNA3-MS2-FLAG-PABPC1 or pcDNA3-MS2-FLAG-PABPN1 were obtained by inserting PABPC1 and PABPN1 cDNAs , respectively , into BamHI-NotI sites of pcDNA3-MS2-FLAG described previously [62] . NMD factor knockdowns were performed by co-transfecting cells with reporter mRNA plasmids and plasmids encoding small hairpin ( sh ) RNAs targeting hUpf1 , hUp2 , or eIF4AIII , 60 h before pulse-chase mRNA decay assays were carried out . mRNA decay assays were performed in HeLa Tet-off cells in DMEM/10% FBS/tetracycline ( 50 ng/ml ) transfected with β-globin mRNA expression plasmids . For each 2-cm well of HeLa Tet-off cells , 10 ng of pcβG or pcβwt ( as an internal control ) and 0 . 2 μg of tetracycline-regulated reporter mRNA expression plasmids were co-transfected using TransIT HeLa Monster reagent ( Mirus ) . For knockdowns , 0 . 5 μg of pSHAG plasmids were co-transfected . In each transfection , empty pcDNA3 vector was added to 1 μg of total plasmid . 36–40 h after transfection , or approximately 60 h in the case of knockdowns , transcription of reporter mRNAs was induced by removal of tetracycline through washing cells with 1 ml of phosphate-buffered saline ( PBS ) and adding DMEM/10% FBS . 6 h later , transcription was shut off by adding tetracycline to a final concentration of 1 μg/ml . Cells were washed with 1 ml PBS and taken up in 500 μl of TRIzol ( Invitrogen ) starting 30 min after tetracycline addition ( 0 min time point ) , and subsequently at time points indicated in each figure . For analysis of knockdown of endogenous hUpf1 , hUpf2 , and eIF4AIII , 0 . 2 μg of the plasmid pSUPERpuro was co-transfected instead of the plasmids expressing β-globin mRNA , and cells were treated and harvested as described earlier [66] . Total cellular RNA was isolated and analyzed by Northern blots as described earlier [47] . The anti-sense RNA probe used for β-globin mRNA detection was described earlier [47] . Northern blots for exogenously expressed TPI mRNAs were probed using UltraHyb reagent following the manufacturer's protocol ( Ambion ) , with a short anti-sense RNA probe complementary to the bovine growth hormone 3′ UTR sequence encoded from the pcDNA3 plasmid . GPx1 mRNAs were probed as described earlier [18] . Rabbit polyclonal anti-sera raised against eIF4AIII ( amino acids 1–41 ) , hUpf1 ( amino acids 1–416 ) , hUpf2 ( C-terminal 206 amino acids ) , and hUpf3b ( full-length ) were described earlier [18 , 47] . Monoclonal mouse antibodies were commercially obtained ( anti-FLAG M2 , Sigma; anti-Myc 9B11 , Cell Signaling ) . Monoclonal mouse anti-HuR antibodies were described earlier [67] . Rabbit polyclonal eRF3 ( #ab-49878 ) and mouse monoclonal PABPC1 ( #ab-6125–100 ) antibodies were from Abcam . In immunoprecipitations shown in Figure 6B , HEK 293T cells were transiently transfected in 3 . 5-cm plates with plasmids expressing FLAG-hUpf1 ( 0 . 4 μg ) , FLAG-PABP1 ( 0 . 5 μg ) , or FLAG-MS2 ( 0 . 5 μg ) , 0 . 5 μg of plasmid expressing wild-type or mutant Myc-eRF3 and 0 . 1 μg of pcDNA3-Myc-hnRNP A1 . Empty pcDNA3 plasmid was added to each transfection to a total of 2 μg . 36–40 h post-transfection , cells were lysed in 400 μl of hypotonic gentle lysis buffer ( 10 mM Tris-HCl [pH 7 . 5] , 10 mM NaCl , 2 mM EDTA , 0 . 5% Triton X-100 , 1 . 0 mM phenylmethylsulfonyl fluoride , 1 μg/ml of aprotinin , and 1 μg/ml of leupeptin ) for 10 min on ice . NaCl was added to 150 mM , and RNase A was added to 125 μg/ml . The extracts were incubated on ice for 5 min and cell debris was removed by centrifugation . RNase-treated lysed cell extracts were incubated for 2 h at 4 °C with 40 μl anti-FLAG M2 agarose ( Sigma ) . The beads were washed eight times with NET-2 ( 50 mM Tris-HCl [pH 7 . 5] , 150 mM NaCl , 0 . 05% Triton X-100 ) and the FLAG-tagged protein was eluted off the beads by gently shaking the beads for 2 h at 4 °C in 20 μl of NET-2 containing 200 μg/ml of FLAG peptide . Immunoprecipitates separated by SDS-PAGE were probed with anti-Myc 9B11 monoclonal antibody ( Cell Signaling ) at a 1:1 , 000 dilution . Co-IPs between wild-type or KAKA-mutant eRF3 and eRF1 were performed as described above from the cells co-transfected with 0 . 5 μg of plasmids expressing FLAG-tagged proteins ( eRF3 , eRF3-KAKA , or MS2 as control ) , 0 . 5 μg of plasmids expressing Myc-eRF1 , and 0 . 1 μg of Myc-hnRNP A1 expressing plasmid . Endogenous eRF3 IPs ( Figure 6A ) were performed as described above except that ∼2 . 5 × 107 HeLa cells were lysed in 1 ml hypotonic gentle lysis buffer , and the lysates were incubated with 10 μg of anti-eRF3 rabbit polyclonal antibody ( Abcam ) , or rabbit pre-immune serum as control , pre-conjugated to 5 mg of protein-A sepharose beads ( GE Healthcare ) . Approximately 107 HEK293T cells from a 10-cm plate expressing Myc-eRF3 , or Myc-peptide as a negative control , were lysed in 1 ml hypotonic gentle lysis buffer as described above . The RNase A–treated , cleared extracts were subsequently incubated with 40 μl anti-Myc resin ( Sigma ) at 4 °C for 2–3 h , following which the beads were washed eight times with 1 ml of NET-2 buffer . The beads were divided into eight equal parts , and indicated amounts of FLAG-hUpf1 , FLAG-PABP1 , or FLAG-hnRNP A1 proteins , which had each been affinity-purified from RNase A–treated HEK293T cell extracts ( protein concentrations estimated by comparison in anti-FLAG Western blot to a GST-FLAG fusion protein of known concentration ) , were incubated in 50 μl of NET-2 supplemented with 0 . 1 mg/ml BSA and 0 . 2 mg/ml FLAG peptide . The reactions were gently shaken at 4 °C for 2–3 h following which the beads were washed eight times with 1 ml of NET-2 buffer . The beads were resuspended in 25 μl of SDS-loading buffer ( 10 mM Tris-HCl [pH 6 . 8] , 2% SDS , 10% glycerol , 0 . 5% bromophenol blue , and 50 mM DTT ) , and 10 μl of the protein sample was resolved on SDS-PAGE followed by Western blot analysis using anti-FLAG M2 antibody ( Sigma , 1:1 , 000 dilution ) .
|
The nonsense-mediated mRNA decay pathway is responsible for rapidly degrading mRNAs with premature termination codons . This is important because it prevents the production of potentially deleterious truncated proteins from aberrant mRNAs , such as those that have undergone erroneous processing . How does the cell discriminate aberrant mRNAs from those that are normal ? Here we present evidence that in human cells , the targeting of an mRNA to nonsense-mediated mRNA decay depends on a competition between proteins associated with the mRNA 3′ UTR that stimulate or antagonize mRNA decay . We show that cytoplasmic poly ( A ) -binding protein , a protein associated with the mRNA 3′ end poly ( A ) tail , antagonizes mRNA decay . By contrast , a protein complex deposited onto mRNAs upon pre-mRNA splicing , called the exon junction complex , stimulates mRNA decay . Our observations suggest that the competition between these proteins , and probably other unknown proteins with similar activities , determines whether a key protein complex in the pathway , the Upf complex , is recruited to the mRNA upon translation termination , which leads to mRNA decay .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"biology"
] |
2008
|
A Competition between Stimulators and Antagonists of Upf Complex Recruitment Governs Human Nonsense-Mediated mRNA Decay
|
Glycosylation of viral envelope proteins is important for infectivity and interaction with host immunity , however , our current knowledge of the functions of glycosylation is largely limited to N-glycosylation because it is difficult to predict and identify site-specific O-glycosylation . Here , we present a novel proteome-wide discovery strategy for O-glycosylation sites on viral envelope proteins using herpes simplex virus type 1 ( HSV-1 ) as a model . We identified 74 O-linked glycosylation sites on 8 out of the 12 HSV-1 envelope proteins . Two of the identified glycosites found in glycoprotein B were previously implicated in virus attachment to immune cells . We show that HSV-1 infection distorts the secretory pathway and that infected cells accumulate glycoproteins with truncated O-glycans , nonetheless retaining the ability to elongate most of the surface glycans . With the use of precise gene editing , we further demonstrate that elongated O-glycans are essential for HSV-1 in human HaCaT keratinocytes , where HSV-1 produced markedly lower viral titers in HaCaT with abrogated O-glycans compared to the isogenic counterpart with normal O-glycans . The roles of O-linked glycosylation for viral entry , formation , secretion , and immune recognition are poorly understood , and the O-glycoproteomics strategy presented here now opens for unbiased discovery on all enveloped viruses .
Enveloped viruses contain one or more membrane proteins important for adhesion and entry to host cells [1] . The majority of envelope membrane proteins are predicted or confirmed to be covered with glycans with important functions in protein folding , transport , formation of infectious particles , entry into host cells , and shielding from the host’s immune system [2–7] . Numerous studies have addressed the structures and functions of N-linked glycans on membrane glycoproteins from different viruses [8–13] , and N-glycosylation has attracted particular attention for the human immunodeficiency virus ( HIV ) , where a cluster of N-glycans constitute the epitope for the 2G12 and other antibodies with broadly neutralizing function [14 , 15] . In striking contrast , information on O-linked glycans and , in particular , where O-glycans are found is generally missing , which leaves a void in knowledge of the biological functions of O-glycosylation . This is in spite of substantial evidence suggesting that O-glycosylation is important for viral infectivity and virus-induced immunomodulation for several viruses [4 , 7 , 16–18] . Viral proteins destined for the virion surface travel through the host’s secretory pathway where they hijack the host cell’s glycosylation machinery and get decorated with glycans [19] . Protein glycosylation is controlled by hundreds of glycosyltransferases that reside in the secretory pathway and that , in a non-template fashion , orchestrate the diversity of glycan structures found on proteins [20] . There is substantial evidence that many viral membrane proteins are N-glycosylated , although there is surprisingly limited experimental evidence for actual glycosylation sites for many viruses with few exceptions [21 , 22] . However , to a large extent the consensus sequence motif NXS/T ( X—all amino acids except P ) enables reliable prediction of N-glycosites [23] . There is less evidence for the presence of O-glycosylation ( GalNAc-type ) on virus membrane glycoproteins , and this largely relies on the presence of mucin-like sequence motifs with high density of PST residues . Such are found in e . g . HSV-1 gC [24] and Ebola virus glycoprotein [25] , but recent studies suggest that O-glycosylation is more prevalent in non-mucin-like regions and often exist as isolated sites or in small clusters [26] . Site-specific O-glycosylation in such isolated or clustered positions may exert co-regulatory functions of basic processes such as pro-protein processing and ectodomain shedding [27] , which may affect viral fusion protein activation and function [28 , 29] . In contrast to N-linked glycosylation that can be predicted with reasonable certainty our knowledge of O-glycosylation is hampered by lack of simple consensus motifs for prediction of O-glycosites . O-glycosylation is unique in being controlled by 20 polypeptide GalNAc-transferases ( GalNAc-Ts ) that transfer GalNAc to select Ser , Thr and , possibly , Tyr residues [30] . The initial GalNAc residues are further elongated , branched , and capped by a large number of different glycosyltransferases in subsequent processing steps . The large number of GalNAc-T isoenzymes with distinct peptide substrate specificities and cell expression patterns provides a high degree of differential regulation of O-glycosylation capacity directed by the repertoire of GalNAc-Ts in a given cell . This unprecedented complexity of protein glycosylation adds to the need for direct experimental analysis of O-glycosylation in the appropriate cellular context to probe biological functions . It is therefore essential to develop strategies to enable characterization of the O-glycoproteomes of viruses produced in representative host cells during virus infection . To address this need , we used herpes simplex virus type 1 ( HSV-1 ) as a model to develop a comprehensive viral O-glycoproteomics strategy . We first determined the major O-glycan structures produced during virus infection , and used this to design a two-step sequential lectin enrichment strategy for capture of desialylated O-glycopeptides in total proteolytic digests of infected cells ( Fig 1A ) . The strategy is based on our recent “SimpleCell” approach for O-glycoproteomics [26 , 31] , but extended to enable sensitive mapping of O-glycosites in cells with the common sialylated core 1 O-glycosylation capacity such as found in human embryonic lung ( HEL ) fibroblasts . Applied to HSV-1 as a proof-of-concept , we provide the first comprehensive HSV-1 O-glycoproteome with identification of 8 of the 12 HSV-1 envelope proteins as O-glycoproteins with a total of 74 unique O-glycosites . We further took advantage of an isogenic cell model in the human keratinocyte ( HaCaT ) cell line in which productive HSV-1 infection can be established , and provide evidence that O-glycan elongation has functional consequences for virus production and infectivity . The strategies and findings presented may have important bearings for vaccine design .
We glycoprofiled mock or HSV-1 infected HEL fibroblasts using a panel of well characterized monoclonal antibodies to core 1 O-glycans . HEL fibroblasts infected with HSV-1 predominantly expressed sialylated core 1 O-glycan structure ST ( Neu5Acα2-3Galβ1-3GalNAcα1-O-Ser/Thr ) and truncated O-glycan structure Tn ( GalNAcα1-O-Ser/Thr ) ( Fig 1B and 1C ) . In order to have a more comprehensive view of O-glycan repertoire we also performed chemical glycan release by reductive β-elimination and analyzed the native glycans by direct infusion nanoESI-MS . Glycomic analysis in negative polarity ( S4A and S4C Fig ) identified the majority of O-glycans as mono- or disialylated T structures in mock- and HSV-1-infected fibroblasts . Due to a potential tendency of sialylated glycans to be ionized better at the negative polarity compared to neutral ones , we have analyzed the same samples at the positive ion mode as well . As it is shown in S4B and S4D Fig , sialylated core 1 O-glycan structures represented the most abundant class of O-glycans in both mock- and HSV-1 infected fibroblasts . Non-sialylated T structures ( Galβ1-3GalNAc1α-O-Ser/Thr ) and various core 2 O-glycan structures ( non- , mono- , and disialylated ) were also present at lower levels . In conclusion , mock- and HSV-1-infected fibroblasts exhibited similar O-glycan profiles with predominantly sialylated core 1 O-glycan structures . Given the finding that infected HEL cells expressed both ST/T and Tn glycans , we developed a two-step lectin enrichment strategy to enable identification of O-glycosites . Enrichment of glycopeptides in total protease digest of complex mixtures of proteins is essential for sensitive detection . The strategy to identify O-glycosylation sites in HSV-1 is depicted in Fig 1A . Cell lysates from HSV-1 infected cells and released virions were digested sequentially with trypsin and neuraminidase . We first employed peanut agglutinin ( PNA ) Lectin Weak Affinity Chromatography ( LWAC ) to capture T-glycopeptides and the flow through of this step was further subjected to Vicia villosa lectin ( VVA ) LWAC to capture Tn-glycopeptides [32] ( Fig 1A ) . The elution fractions of PNA and VVA LWACs were analyzed by tandem mass spectrometry equipped with ETD fragmentation to identify O-linked glycosylation sites on HSV-1 envelope glycoproteins . By using this strategy we identified eight out of the 12 HSV-1 envelope glycoproteins and a total of 74 unique O-glycosylation sites . Nearly all of these O-glycosites ( 72 sites ) were identified in total lysates of infected cells , while direct analysis of released virions resulted in identification of 20 O-glycosites of which only two were not found in the lysate . Comparing identifications from PNA and VVA LWACs , fewer glycosites ( 43 sites ) were identified from PNA LWAC than from VVA ( 58 sites ) ( Tables 1 and S1 ) . However , at least one third of the sites identified with PNA LWAC were non-redundant with VVA LWAC identified sites . The direct analysis of virions yielded markedly lower number of O-glycosylation sites ( 10 sites with each VVA and PNA LWAC ) ( Table 1 ) . Fig 2 presents a graphic depiction of the HSV-1 O-glycoproteome . A total of 34 out of 74 identified O-glycosylation sites were localized on the four HSV-1 membrane proteins , gB , gD , gH and gL , which are all essential for viral infectivity in vitro [33–36] ( Table 2 ) . Twenty-one glycosylation sites were identified in gB , which is essential for fusion with host cell membrane [33] . The identified glycosylation sites include two positions , T53 and T480 ( S1 Table ) , which have previously been proposed to be important for the interaction with the paired immunoglobulin-like type 2 receptor α based on the finding that Ala substitutions resulted in loss of interaction [37] . In addition , O-linked glycans were found throughout the ectodomain and localized to both ordered and unstructured regions of the molecule [38] ( Fig 2 ) . Interestingly , several gB O-glycosylation sites were highly conserved between 8 members of the human herpesviruses ( Figs 3 and S2 ) . Glycoprotein D , which is necessary for virus entry into host cells [34] , was glycosylated at five sites ( Fig 2 ) . Two of the identified O-glycosites , S93 and S100 , were located within the Ig-like core of the molecule . Interestingly , two more glycans ( T255 and Y259/S260 ) were situated on the Ig-core-flanking functional alpha-helix important for maintaining the unliganded conformation of the molecule as well as interaction with nectin-1 [39–41] . One additional site , S33 , was located within the N-terminal motile region involved in the interaction with the entry receptor HVEM [39] . In glycoprotein H , which is also required for HSV-1 entry into permissive cells [35] , we found four ambiguous O-glycosylation sites at the N-terminus ( Fig 2 ) . Three of the four sites were situated within a disordered region between two structural subdomains [42] . The peripheral membrane protein gL , that forms a heterodimer with gH , carried four O-glycans all of which were located within a poorly structured region of the molecule [42] ( Fig 2 ) . No sites were found within the protein-protein interaction regions of the two proteins . The remaining 40 O-glycosites were distributed among four HSV-1 glycoproteins ( gC , gE , gI and gG ) , which are all important for virus-host interaction and modulation of the host immune response ( Table 2 , Fig 2 ) . Glycoprotein C is involved in initial attachment to heparan sulphate proteoglycans as well as immune evasion by acting as a complement receptor [43 , 44] , and is known to contain a glycosylated mucin-like tandem repeat region [45] . Accordingly , 9 of the identified sites were localized within the mucin-like region , while 3 sites were found outside of the tandem repeat region ( Fig 2 ) . Glycoprotein E forms an Fc receptor together with glycoprotein I and is known to facilitate cell-to-cell spread [46 , 47] . Interestingly , we identified 16 O-glycosylation sites , of which many densely covered the N-terminal domain of the molecule , whereas the Fc-binding domain did not carry any O-glycans [48] ( Fig 2 ) . Four O-glycosylation sites were identified on glycoprotein I ( Fig 2 ) , one of which was situated within the region required for the Fc receptor function [49] . Unfortunately , we were not able to identify any of the sites , which are known to be glycosylated within the tandem repeat region of gI [50] . It is known that glycans within this mucin-like region of gI are much more closely spaced as compared to gC , thus it is very likely that trypsin digestion is inefficient within the very tightly glycosylated region of gI . Finally , we detected eight O-glycosylation sites on the chemokine-binding [51] glycoprotein G ( Fig 2 ) . HEL fibroblasts normally produce complete O-glycans with fully sialylated core 1 structures ( Fig 1B and 1C ) , but as shown here , infection with HSV-1 resulted in marked intracellular expression of truncated O-glycans ( Tn ) , which was confirmed by our two-step O-glycoproteomics strategy where a substantial number of truncated Tn-glycopeptides were identified in mixture with T-glycopeptides ( Fig 1 , Tables 1 and S1 ) . This prompted us to further investigate the effect of HSV-1 infection on O-glycan synthesis in more detail . Immunofluorescent staining for the truncated Tn O-glycan structure in permeabilized HSV-1-infected cells showed a Golgi-like staining pattern with numerous dispersed vesicle-like structures throughout the cytoplasm . The intracellular Tn expression increased along the course of infection with a complete dispersal of Tn staining throughout the cell after 9 hours ( Fig 4A and 4C , HPA ) . Tn expression also partially co-localized with gC , implying that envelope glycoproteins are indeed O-glycosylated in the fragmented Golgi ( S3 Fig ) . Despite the high expression of Tn inside the infected cells , we only detected small amounts of Tn on the surface as evaluated without permeabilization ( Fig 4D and 4E ) . Both HSV-1 and mock-infected cells predominantly expressed elongated and sialylated O-glycans on the surface ( Fig 4D and 4E ) . Co-localization studies showed that the intracellular Tn-positive structures in HSV-1-infected cells were highly correlated with the Golgi marker giantin during early stages of infection ( Fig 4A: HSV-1 5 hpi ) , whereas lower degree of co-localization was observed late in infection ( Fig 4A: HSV-1 9 hpi ) . There was no correlation between expression of Tn and the ER marker GRP94 in either mock- or HSV-1-infected cells ( Fig 4B ) . However , partial co-localization was observed with trans-Golgi network marker TGN46 in heavily infected cells ( Fig 4C: HSV-1 9 hpi ) . To further evaluate how the infection impacted the organization of Golgi apparatus , we next investigated the relative localization of cis- and trans-Golgi markers during the course of HSV-1 infection . In HEL fibroblasts the classical Golgi markers GM130 , giantin and β4Gal-T1 were redistributed into several distinct punctuate vesicular-like structures most likely representing remnants of Golgi structures as previously described [52] . Both cis-Golgi-resident GM130 and cis-/medial-Golgi-specific giantin were detected in close proximity within infection-induced Golgi fragments ( Fig 4F: HSV-1 5 hpi , 9 hpi ) . Similarly , giantin and trans-Golgi-resident β4Gal-T1 were highly correlated within discrete Golgi fragments of infected cells ( Fig 4G: HSV-1 5 hpi , 9 hpi ) . These findings suggest that the individual vesicular-like structures in HSV-1-infected HEL fibroblasts mirror the composition of an intact Golgi apparatus and potentially contain all the glycosyltransferases required for O-glycan synthesis and elongation . Given our findings that most HSV-1 membrane proteins are O-glycosylated , and that O-glycans are speculated to play important roles in viral infectivity [37] , we wanted to analyze whether O-glycan structures were important for virus production and infectivity . In the past , several studies have used inhibitors of glycosylation that are known to disturb protein trafficking , inhibit growth , and even cause cell death [53–55] . We recently produced an isogenic cell model based on the non-tumorigenic human epidermal keratinocyte cell line , HaCaT [56] . Wild-type HaCaT cells produce mainly core 1 mature ST O-glycans similar to HEL fibroblasts , while HaCaT with COSMC knockout , also designated SimpleCells ( SC ) , express glycoproteins with homogenous truncated Tn and STn O-glycans [56] . The isogenic HaCaT cells therefore provide a unique well-defined cellular system to study the effect of truncated O-glycosylation on viral production and infectivity . We infected HaCaT WT and SC in parallel with 10 PFU/cell of HSV-1 Syn17+ produced in HaCaT WT and evaluated viral titers produced in the media . Virus produced in HaCaT SC compared to WT exhibited severely reduced titers as evaluated by plaque assay ( on average 10-fold at 12 h and 24-fold at 20 h after infection ) ( Fig 5A ) . To evaluate whether this effect was due to production of viral particles we analyzed viral DNA in the media , which showed a substantial reduction in viral DNA ( 4- to 8-fold at 20 h after infection ) detected in the medium from HaCaT SC compared to WT ( Fig 5B ) . These results suggest that truncated O-glycans per se pose problems with viral particle formation or early entry events .
Here we provided a strategy for comprehensive characterization of O-glycoproteomes of any virus produced in infected host cells . We demonstrate with the complex model virus HSV-1 that the envelope proteins are heavily O-glycosylated with at least eight membrane proteins being O-glycoproteins . Most of the HSV-1 membrane proteins are also predicted to be N-glycosylated ( 37 predicted NXS/T sites in total on 11 proteins ) , although actual N-glycosites in the majority of cases are unknown . In contrast to N-glycosylation , there is a particular need for experimental identification of O-glycosites arising from a lack of simple predictive consensus sequence motifs and the necessity of taking the O-glycosylation capacity of the host cell into account . The key step for sensitive glycoproteomics is enrichment of glycopeptides [57] . For N-glycoproteomics the common N-glycan core structure enables efficient capture of most N-glycopeptides with a mixture of lectins [58 , 59] , but this is not the case for O-glycans where there is no mixture of lectins available that can encompass all O-glycan structures . Our O-glycoproteomics strategy is therefore versatile for host cells producing core 1 O-glycan structures , but not yet fully applicable to host cells producing more complex O-glycans . However , we show that HSV-1 infection causes an accumulation of truncated O-glycans as well as elongated core 1 structures that can be captured with the available VVA and PNA lectins . The O-glycoproteomic strategy is applicable to any virus produced in infected host cells , which should enable wide application for even highly infectious viruses such as HIV and Ebola . We chose HSV-1 as a model system because of its complex envelope proteome . Whereas most enveloped viruses in general encode only one or two membrane glycoproteins , the human herpes viruses , including HSV-1 , express more than ten glycoproteins located in the viral envelope and various membranes of the infected cells . Human herpes viruses are widely spread pathogens known to establish latency in various cell types enabling recurrent disease by reactivation [60] . HSV-1 is a large DNA virus of high complexity and one of the most prevalent herpes viruses infecting up to 80% of the world’s population [61 , 62] . Mature viral particles consist of an icosahedral capsid containing the viral genome , a second less structured protein layer called the tegument , and the surface envelope with at least 12 viral proteins [63] . A number of previous studies have addressed the structure and function of N-linked glycans on HSV-1 glycoproteins [3 , 64–66] , and there have been attempts to identify and characterize O-glycosylation of proteins as well [45 , 67–70] . Although several HSV-1 proteins were previously found to be O-glycosylated , studies of actual O-glycosites are generally missing . We identified multiple O-linked glycosylation sites on HSV-1 proteins important for attachment and entry into target cells ( gB , gC , gD , gH , gL ) by interactions with host cell receptors such as herpes virus entry mediator ( HVEM ) [71] , nectin-1 [72] , 3-O-sulfated heparan sulfate [73] , 4-O-sulfated chondroitin sulfate [74] , as well as the paired immunoglobulin-like type 2 receptor α ( PILRα ) [75] and αVβ6/αVβ8 integrins [76] . Other of the identified O-glycans were localized to HSV-1 envelope glycoproteins involved in virus spread or immune modulation ( gE , gI , gG ) [46 , 47 , 51] . Of particular interest , we provided confirmation of the glycosylation of the previously identified T53 and T480 sites on gB essential for virus entry in host cells [4 , 37 , 77] . Based on mutational studies , O-linked glycans at these sites have been specifically implied in gB binding to PILRα [37] . Furthermore , studies in mice indicate that glycosylation at these sites promotes development of keratitis and neuroinvasion [37] . Three of the newly identified sites on gB , T169 and two sites within the peptide stretch 265-YGTT-268 , are situated in close proximity to hydrophobic loop regions that are predicted to be involved in fusion with the host membrane , suggesting that the O-glycans could influence the interaction with the host cell [38] . Interestingly , these three sites in gB are found within highly conserved gB regions between the Herpesviridae family members of at least seven out of the eight human herpesviruses . Furthermore , a recent study reported that mutational insertion of a fluorescent protein at position 241 , which we found to be O-glycosylated , resulted in loss of fusogenic gB function [78] . Another indication that specific O-glycans could be important for interaction between HSV and the host cell was the identification of O-glycosylation sites on glycoprotein D , both within the flexible N-terminus of the molecule that is forming a hairpin upon binding to HVEM ( S33 ) [39 , 79] , and within the α helix that is part of the interaction surface with the adhesive protein nectin-1 [41] ( T255 and Y259/S260 ) . We also identified a cluster of O-glycosites in the HSV-1 glycoprotein C mucin-like domain , which contributes to interaction with glycosaminoglycans on host cells [80 , 81] . We did not , however , identify all expected sites in the mucin-like sequences that are notoriously difficult regions to analyze by MS sequencing strategies . Similarly , we did not detect O-glycans in the mucin-like tandem repeat region of gI , which has been shown in vitro to accommodate a high level of glycosylation [50] . The glycoproteomic strategy used is based on direct protease digests of virus-infected cells followed by lectin enrichment of O-glycopeptides and ETD-based 'bottom-up' tandem mass spectrometry . With this approach it is possible to address O-glycosylation of viral proteins in a global proteome manner as glycosylated by infected host cells . Since capacity for O-glycosylation varies among cell types , the O-glycoproteome determined in representative infected host cells may guide selection of host cells for recombinant expression of vaccines based on viral membrane proteins . This should be especially important for viruses with high number of O-glycosylation sites such as Ebola virus , Marburg virus , and Crimean-congo hemorrhagic fever virus [25 , 82 , 83] . It should be noted that the current MS sequencing strategy has some limitations with particularly dense O-glycopeptides with abundant Pro residues . The problem is partly due to difficulties in protease digestion and partly due to insufficient glycopeptide fragmentation in MSn . While this clearly is a limitation , such clustered regions with mucin-like sequence containing high density of PST residues may be reliably predicted to be O-glycosylated in many cell types . In the intact Golgi apparatus the topology of glycosyltransferases is well organized in different Golgi stacks and TGN in an ordered fashion somewhat reflecting the step-wise biosynthetic pathways of glycosylation [20] . Viral infection is known to induce changes in organization of the Golgi in agreement with the findings of accumulation of truncated O-glycoforms throughout the cytoplasm in the present study [52] . We thus characterized the micro-organization of HSV-1-induced Golgi fragments with respect to different Golgi-compartment resident proteins . Surprisingly , confocal microscopy suggested that the individual Golgi fragments contained the structural components of cis , medial , and trans-Golgi , as demonstrated by highly correlated localization of GM130/giantin and giantin/β4Gal-T1 upon Golgi fragmentation . The existence of cis , medial , and trans-Golgi enzymes within the same fragments would allow sequential O-glycan processing despite Golgi fragmentation , and could explain why we found that infected cells retain the ability to sialylate most of the surface O-linked glycans , regardless of massive amounts of newly synthesized proteins trafficked through a fragmented Golgi apparatus . A major function of glycosylation of viral envelope glycoproteins appears to be shielding from host immunity [5–7] . The shielding function is well documented for N-glycans [5 , 6] but presumably O-glycans serve similar functions [7] . However , the expression of immature truncated O-glycans in the context of virus glycoproteins may have immunostimulatory effects . In contrast to N-glycans with their common large core structure that is highly conserved throughout evolution , the most immature truncated O-glycans are highly immunogenic and may be accommodated together with a short peptide backbone in the binding pocket of an antibody [84] . We and others have previously shown how truncated O-glycopeptides may serve as immunodominant antibody epitopes , which are useful for development of cancer-specific immunotherapeutic intervention and as biomarkers for cancer [85] . In this context , we recently screened a library of short Tn O-glycopeptides covering gG of HSV-2 for the presence of immunodominant O-glycopeptide IgG antibody epitopes in HSV-1 and -2 infected individuals . Interestingly , we did identify one O-glycopeptide epitope to which IgG antibodies were present in HSV-2 , but not HSV-1 infected individuals providing a potential diagnostic biomarker [86] . Moreover , the serum IgG antibodies reacted with several glycan structures on the same peptide including truncated ( Tn ) or elongated ( ST ) O-glycan , suggesting that these antibodies participate in immunity to viral glycoproteins [86] . The existence of antibodies recognizing O-glycopeptide epitopes suggests that O-glycosylation both with respect to sites and structures should be considered for vaccine design and production . This is especially appealing in relation to targeting patches of O-glycans in mucin domains contained in herpes viruses as well as several human pathogenic virus species , including the deadly Ebola and Marburg viruses . Currently , there are no effective HSV vaccines despite extensive efforts and a better understanding of the O-glycans of the viral glycoproteins may lead to novel approaches for vaccine development . The widespread nature of O-glycosylation of the HSV-1 envelope proteins prompted us to address the question whether elongated O-glycans are important . We exploited our recently produced isogenic cell model based on the human epidermal keratinocyte cell line HaCaT [56] . Wild-type HaCaT cells express mature core 1 O-glycans while HaCaT SimpleCells ( SC ) express homogenous truncated Tn and STn O-glycans due to knockout of the private chaperone of the core 1 synthase , C1Gal-T . We used this isogenic cell system to demonstrate that viral propagation and titers in HaCaT SC with truncated O-glycans were severely hampered . Thus , elongated O-glycans are functionally relevant and it is likely that these functions are directed by O-glycans at specific sites in the HSV-1 O-glycoproteome . It should be noted , however , that loss of O-linked glycan elongation has multiple cellular consequences , and further experimentation is required to define the molecular mechanisms behind the observed effect . For this purpose , the HaCaT cell model can now be further explored with glycosyltransferase gene targeted isogenic cell pairs to dissect requirements for particular GalNAc-T repertoire and/or O-glycan structures for HSV-1 viral propagation . It is also conceivable that the dependence on intact O-linked glycosylation for virus generation is not unique to HSV-1 , but we anticipate that the described strategy can be used to test the importance of O-glycosylation for other enveloped viruses . A similar genetic deconstruction approach has previously been used with great success for mapping Lassa virus binding to α-dystroglycan and cellular entry [87] , and this should greatly advance our understanding of the role of glycosylation in virology . In summary , we have mapped the O-glycosylation sites on HSV-1 and shown that elongation of O-linked glycosylation is important for HSV-1 biology . Further studies are now possible to decipher the exact mechanism responsible for the observed effects . The glycoproteomics workflow developed should be widely applicable to enveloped viruses with the potential to consider the natural O-glycan coat in the design of antiviral vaccines and drugs .
The wild-type HSV-1 virus Syn17+ [88] was used throughout the study , and the virus titers were determined by plaque titration on Green monkey kidney ( GMK , obtained from the Swedish Institute for Infectious Disease Control , Stockholm ) cells as previously described [89] . HSV-1 Syn17+ virus was cultivated in HaCaT wild type keratinocytes or GMK cells depending on downstream application and the titers were determined as mentioned above . Diploid human embryonic lung fibroblasts [70] ( HEL , obtained from the cell culture collection at the Sahlgrenska University Hospital , department of Clinical Microbiology , Gothenburg ) at a low passage level were cultivated in Eagle’s MEM ( Gibco , Life Technologies ) with 10% FCS ( Sigma ) , 100 IU/mL penicillin , 100 μg/mL streptomycin ( Gibco , Life Technologies ) and 2 mM L-glutamine . HaCaT wild type [90] and HaCaT COSMC-/- [56] keratinocytes were grown in DMEM ( Gibco , Life Technologies ) , supplemented with 10% FCS ( HyClone ) , 100 IU/mL penicillin and 100 μg/mL streptomycin ( Gibco , Life Technologies ) . HaCaT COSMC-/- clone D5 harbors a 10 bp deletion at the zink finger nuclease target site of COSMC gene , whereas clone E5 harbors a combined 12 bp deletion and a 2 bp insertion . Both genetic alerations result in introduction of STOP codons due to frameshift mutations [56] . Monoclonal mouse to Tn ( 5F4 , IgM ) , mouse to T ( 3C9 , IgM ) , mouse to STn ( 3F1 , IgG ) , mouse to GalNAc-T2 ( 4C4 , IgG ) mouse to β4Gal-T1 ( 2F5 , IgG ) and polyclonal rabbit to gC-1 ( KF922 , 1:700 ) antibodies were produced as previously described [24 , 91] . Rabbit anti-giantin ( 1:500 ) and rat anti-GRP94 ( 1:50 ) were purchased from Abcam . FITC-conjugated HPA ( Helix pomatia agglutinin , 1:2000 ) was from Invitrogen . FITC-conjugated polyclonal rabbit anti-HSV-1 antibody was purchased from DAKO ( 1:40 ) . Alexa Fluor 488 F ( ab' ) 2 fragment of Goat anti-Mouse IgG ( H+L ) ( 1:500 ) , Alexa Fluor 546 Goat anti-Mouse IgM ( μ chain ) ( 1:500 ) were from Life Technologies . FITC-conjugated polyclonal Goat anti-Mouse antibody ( 1:100 ) and TRITC-conjugated Swine anti-Rabbit antibody ( 1:200 ) were from DAKO . Alexa Fluor 647 Goat anti-Mouse IgM ( μ chain ) was purchased from Life Technologies . For glycoproteomic analysis , GMK-produced HSV-1 at a multiplicity of infection ( MOI ) of 3 plaque-forming units ( PFU ) per cell was added to HEL fibroblasts in roller bottles ( 34 × 106 cells/bottle ) . The viral particles were allowed to attach to the cells for 1 h at 37°C and 5% CO2 before the inoculum was removed and new growth medium was added . The cells and medium were harvested after most of the cells exhibited cytopathic effects of infection ( ~20 h ) . The cells from 3 roller bottles were harvested by scraping with a rubber policeman in ice-cold PBS . The viral particles from the medium were harvested by ultracentrifugation at 100 , 000 × g for 1 hour at 4°C using 25 × 89 mm ultracentrifuge tubes ( Beckman Coulter , Brea , CA ) and a Ti70 . 1-rotor ( Beckman Coulter ) . For glycoprofiling by reductive β-elimination ( see S1 Text ) , confluent HEL fibroblast monolayers ( ~6 × 106 cells ) were infected with MOI of 3 PFU/cell of GMK-produced HSV-1 and harvested at ~23 h post-infection as described above . Medium without serum was used throughout the infection to avoid serum glycan contamination . For immunofluorescence staining , HEL cells were grown either on teflon-coated glass slides or on glass cover slips . Confluent monolayers were infected with GMK-produced HSV-1 at a MOI of 10 PFU/cell as described above . The cells were harvested at either 4 and 8 or 5 and 9 hours post-infection . For infection of keratinocytes , confluent HaCaT wild type or COSMC-/- cell monolayers in 6-wells were infected with HaCaT wild type-produced HSV-1 Syn17+ at a MOI of 10 PFU/cell as described above . The growth medium was harvested at 12 and 20 hours after infection . Infected HEL cell pellet and ultracentrifuged HSV-1 pellet from the growth medium were processed in parallel . The lysates were prepared as previously described [31] with several modifications . Briefly , cell or virus pellet was resuspended in 0 . 05% RapiGest ( Waters ) in 50 mM ammonium bicarbonate and lysed using a sonic probe . Cleared cell and virus lysates were reduced and alkylated as described [31] and then treated with 5 U and 1 U , respectively , of PNGase F ( Roche ) over night at 37°C , followed by digestion with 30 μg/7 μg of trypsin ( Roche ) for 12 h at 37°C . The PNGase F treatment was then repeated followed by 2 h incubation with 10 μg/3 μg of trypsin . The samples were then treated with concentrated trifluoracetic acid ( 8 μL/sample , 20 min at 37°C ) and cleared by centrifugation ( 10 , 000 × g 10 min ) . The cleared digests were purified on C18 Sep-Pak ( Waters ) and treated with 100 U of neuraminidase ( P0720 , New England Biolabs ) in 50 mM sodium citrate pH 6 . 0 at 37°C for 2 h . T and Tn glycopeptides were sequentially enriched using PNA and VVA LWAC as previously described [32] and as described in detail in S1 Text . LWAC fractions from total cell lysate digests were screened by preliminary LC-MS for glycopeptide content , and those most enriched in glycopeptides were pooled together and further fractionated by isoelectric focusing as previously described [92] . Mass spectrometry analysis was performed on an EASY-nLC 1000 UHPLC ( Thermo Scientific ) interfaced via nanoSpray Flex ion source to an LTQ-Orbitrap Velos Pro spectrometer ( Thermo Scientific ) as previously described [56] with minor changes and as described in detail in S1 Text . Data processing was performed using Proteome Discoverer 1 . 4 software ( Thermo Scientific ) as previously described with small changes [31] . Due to the high speed of data processing Sequest HT node was used instead of Sequest . All spectra were initially searched with the full cleavage specificity , filtered according to the confidence level ( medium , low and unassigned ) and further searched with the semi-specific enzymatic cleavage . In all cases the precursor mass tolerance was set to 6 ppm and fragment ion mass tolerance to 50 mmu . Carbamidomethylation on cysteine residues was used as a fixed modification . Methionine oxidation and HexNAc and HexHexNAc attachment to serine , threonine and tyrosine were used as variable modifications for ETD MS2 . All HCD MS2 were pre-processed as described [31] and searched under the same conditions mentioned above using only methionine oxidation as variable modification . All spectra were searched against a concatenated forward/reverse human-specific database ( UniProt , January 2013 , containing 20 , 232 canonical entries . In addition , another 251 common contaminants and 3187 entries of viruses known to infect humans were included in the search ) using a target false discovery rate ( FDR ) of 1% . FDR was calculated using target decoy PSM validator node , a part of the Proteome Discoverer workflow . The resulting list was filtered to include only peptides with glycosylation as a modification . This resulted in a final glycoprotein list identified by at least one unique glycopeptide . ETD MS2 data were used for unambiguous site assignment . HCD MS2 data were used for unambiguous site assignment only if the number of GalNAc residues was equal to the number of potential sites on the peptide . Teflon-coated glass slides were washed 3 times in PBS followed by 5 min fixation in ice-cold acetone and allowed to air-dry . For cell surface staining , HEL cells grown on cover slips were washed with Hank’s balanced salt solution and fixed with warm 4% paraformaldehyde in PBS for 10 min . Control cells were permeabilized with 0 . 3% Triton-X100 for 1 . 5 min . For neuraminidase treatment , teflon-coated slides/cover slips were incubated with 0 . 1 U/mL Clostridium perfringens neuraminidase ( Sigma-Aldrich ) in 0 . 05 M sodium acetate pH 5 . 5 at 37°C for 1 h . Teflon-coated slides/cover slips were incubated with primary antibodies at 4°C over night , washed 3 times with PBS and incubated with secondary antibodies ( diluted in 2 . 5% bovine serum albumin ( BSA ) in PBS , 0 . 03% azide ) for 45 min in RT . After 3 washes with PBS , the specimens were mounted using ProLong® Gold antifade mounting reagent with 4’ , 6-diamidino-2-phenylindole ( DAPI ) ( Life Technologies ) . Immunofluorescent staining was inspected using a Zeiss Axioskop 2 microscope equipped with AxioCam MR3 digital camera . For the co-localization staining using three fluorophores , the teflon-coated glass slides were blocked in 3% BSA in PBS for 30 minutes followed by incubation with primary antibodies and lectins at 4°C over night . Slides were washed 3 times in PBS and once in distilled water followed by incubation with secondary antibodies at 37°C for 45 minutes . Finally the glass slides were washed as described above , air dried and mounted with Prolong Gold Anti-fade reagent containing DAPI ( Life Technologies ) . Triple immunofluorescence staining was analyzed using a Zeiss LSM 510 Meta confocal microscope ( Carl Zeiss AG , Oberkochen , Germany ) equipped with a Plan-Apochromat 63x objective in oil immersion . Images were edited in Adobe Photoshop CS6 . Confluent HEL cells grown in 6-well plates were infected with HSV-1 at a MOI of 10 or mock infected as described above for indicated time points . The cells were harvested by trypsinization ( TrypLE , Life Technologies ) , washed in 10 mL ice-cold PBS and fixed in 0 . 1% paraformaldehyde for 24 h at 4°C . After fixation the cells were washed in ice-cold PBS as described above and divided into 100 μL samples with 5 × 105 cells per sample . Half of the samples were permeabilized after fixation by addition of 1x Perm/Wash solution ( BD Biosciences ) according to the manufacturer’s instructions . After permeabilization , a portion of the HEL cells samples were washed in PBS and treated with 100 μL Clostridium perfringens neuraminidase ( 0 . 1 U/mL ) ( Sigma-Aldrich ) in 0 . 05 M sodium acetate pH 5 . 5 for 40 minutes at 37°C . Thereafter the samples were washed two times in 1x Perm/Wash solution or PBS and incubated with primary antibodies or lectins for 30 min at 4°C . Subsequently the cells were washed as described above and then incubated with secondary antibodies for 30 min at 4°C in the dark followed by washing as described . The cells were analyzed using a Cube8 instrument ( Partec Nordic AB ) and FlowJo software . Titers of virus produced in HaCaT wild type or COSMC-/- keratinocytes were determined on Green monkey kidney ( GMK ) cells . Cell monolayers were infected with serial dilutions of virus and allowed to attach . After 1 h the inoculum was removed and the cells overlaid with medium containing 1 . 5% methylcellulose ( Sigma-Aldrich ) , 2 . 5% FCS , 100 IU/mL penicillin and 100 μg/mL streptomycin ( in HBSS ( Sigma-Aldrich ) + DMEM ( Gibco , Life Technologies ) at a ratio of 1:1 ) . After 48 h incubation , the overlay medium was removed , the cells fixed with 1% crystal violet ( in 70% EtOH:37% formaldehyde:acetic acid 20:2:1 ) , washed three times with water and allowed to dry . The resulting plaques were inspected and counted using a light microscope ( Olympus IMT-2 ) . The samples were diluted 1:1000 in ice cold PBS and the total DNA content of each diluted fraction was extracted in a MagNa Pure LC robot ( Roche Diagnostics , Mannheim , Germany ) using a MagNa Pure DNA isolation kit ( Roche Diagnostics Scandinavia AB , Stockholm , Sweden ) , according to the manufacturer’s instructions . The input and the output volumes were adjusted to 200 μL and 100 μL respectively . For assessing the DNA copy number of HSV-1 , a 118-nucleotide segment of the gB-1 region was amplified with primers described in [93] . The PCR reaction volume was set to 50 μL and contained 25 μL TaqMan® 2x PCR Master Mix ( Roche Diagnostics , Branchburg , NJ ) , 15 μL primer/probe mix ( forward primer at 0 . 9 μM , reverse primer at 0 . 9 μM and probe at 0 . 2 μM concentrations ) , and 10 μl of total DNA sample . Amplification of the target sequence was performed using the ABI Prism 7900 system ( Applied Biosystems , Foster City , CA ) . The reaction conditions were set to 2 min at 50°C followed by incubation for 10 min at 95°C and finally 45 PCR cycles of two-step amplification ( 15 sec at 95°C and 60 sec at 58°C ) . HSV-1 Forward 5’-GCAGTTTACGTACAACCACATACAGC-3’; HSV-1 Reverse 5’-AGCTTGCGGGCCTCGTT-3’; HSV-1 Probe FAM-5’-CGGCCCAACATATCGTTGACATGGC-3’-TAMRA . The efficiency of each round of PCR was determined using 10-fold dilutions of Topo TA plasmids ( Invitrogen AB , Stockholm , Sweden ) with insert of respective amplicon created according to the manufacturer’s instructions . Statistical analysis was performed using GraphPad Prism 6 software .
|
Information on site-specific O-glycosylation of viral envelope glycoproteins is generally very limited despite important functions . We present a powerful mass-spectrometry based strategy to globally identify O-glycosylation sites on viral envelope proteins of a given virus in the context of a productive infection . We successfully utilized the strategy to map O-linked glycosylation sites on the complex HSV-1 virus demonstrating that O-glycosylation is widely distributed on most envelope proteins . Moreover , we used genetically engineered keratinocytes lacking O-glycan elongation capacity to demonstrate that O-linked glycans are indeed important for HSV-1 biology as HSV-1 particles produced in these cells had significantly lower titers compared to wild-type keratinocytes . These tools enable wider discovery and detailed analysis of the role of site-specific O-glycosylation in virology .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
A Strategy for O-Glycoproteomics of Enveloped Viruses—the O-Glycoproteome of Herpes Simplex Virus Type 1
|
Cerebral malaria ( CM ) is a severe complication of Plasmodium falciparum infection that results in thousands of deaths each year , mostly in African children . The in vivo mechanisms underlying this fatal condition are not entirely understood . Using the animal model of experimental cerebral malaria ( ECM ) , we sought mechanistic insights into the pathogenesis of CM . Fatal disease was associated with alterations in tight junction proteins , vascular breakdown in the meninges / parenchyma , edema , and ultimately neuronal cell death in the brainstem , which is consistent with cerebral herniation as a cause of death . At the peak of ECM , we revealed using intravital two-photon microscopy that myelomonocytic cells and parasite-specific CD8+ T cells associated primarily with the luminal surface of CNS blood vessels . Myelomonocytic cells participated in the removal of parasitized red blood cells ( pRBCs ) from cerebral blood vessels , but were not required for the disease . Interestingly , the majority of disease-inducing parasite-specific CD8+ T cells interacted with the lumen of brain vascular endothelial cells ( ECs ) , where they were observed surveying , dividing , and arresting in a cognate peptide-MHC I dependent manner . These activities were critically dependent on IFN-γ , which was responsible for activating cerebrovascular ECs to upregulate adhesion and antigen-presenting molecules . Importantly , parasite-specific CD8+ T cell interactions with cerebral vessels were impaired in chimeric mice rendered unable to present EC antigens on MHC I , and these mice were in turn resistant to fatal brainstem pathology . Moreover , anti-adhesion molecule ( LFA-1 / VLA-4 ) therapy prevented fatal disease by rapidly displacing luminal CD8+ T cells from cerebrovascular ECs without affecting extravascular T cells . These in vivo data demonstrate that parasite-specific CD8+ T cell-induced fatal vascular breakdown and subsequent neuronal death during ECM is associated with luminal , antigen-dependent interactions with cerebrovasculature .
Malaria , a disease caused by protozoan parasites of the genus Plasmodium , is a leading cause of morbidity and mortality in the developing world . Of the 627 , 000 annual deaths due to malaria , the vast majority are caused by Plasmodium falciparum infections [1] . Human cerebral malaria ( HCM ) is one of several clinical manifestations of severe P . falciparum infection and is diagnosed by coma and parasitemia in the absence of meningitis , hyperglycemia , and postictal state [2] . HCM is fatal in 15–30% of affected individuals [3 , 4] , while an additional 10% of survivors suffer long-term neurological sequelae such as ataxia , hemiplegia , and cognitive impairment [5] . Yet , the underlying cause of HCM remains unknown . Several characteristic pathologies are observed in the brains of patients suffering from HCM including vascular hemorrhage [6] , breakdown of the blood brain barrier ( BBB ) [7 , 8] , and edema [9 , 10] . At the cellular and molecular level , HCM is associated with an increase in systemic pro-inflammatory cytokines [11 , 12] , endothelial cell ( EC ) activation [13] , and sequestration of parasite-infected red blood cells ( iRBCs ) [2] and leukocytes [8] within the brain vasculature . These conditions are hypothesized to contribute to the observed BBB disruption and cerebral edema as well as ischemia throughout the CNS [14] . However , interpretations of these HCM data are limited by the fact that most information about CNS pathology and the cellular response to P . falciparum is derived from post-mortem analyses . Although real-time in vivo imaging techniques such as MRI [9] and ophthalmoscopy [15] have been used in patients suffering from HCM , they lack the resolution needed to observe cellular dynamics in the CNS and have been used mostly to improve the fidelity of CM diagnoses . Examination of cellular dynamics in animal model systems is therefore needed to uncover mechanistic insights into HCM . Infection of mice with Plasmodium berghei ANKA ( PbA ) induces a neurological disease called experimental cerebral malaria ( ECM ) that mirrors many of the pathological features observed in HCM . These include increased pro-inflammatory cytokines , vascular pathology , disruption of the BBB , and cerebral edema [16–19] . ECM in mice is a widely used model of HCM and provides a valuable tool for elucidating the mechanisms involved in CM pathogenesis and identifying cellular and molecular targets for adjunctive therapy [20] . Leukocytes have been shown to accumulate in the brains of mice throughout the course of ECM [21 , 22] . In addition , mice that are genetically deficient in peripheral leukocyte chemokine receptors such as CCR5 and CXCR3 ( or , their ligands ) are resistant to ECM [23–26] . Many individual immune cell populations including neutrophils [27 , 28] , macrophages/monocytes [18 , 29] , NK cells [30] , and CD4+ T cells [31 , 32] have been implicated in the pathogenesis of this disease . However , other studies have shown that neither antibody-mediated nor genetic depletion of these cells affected the accumulation of iRBCs in the CNS [33] or the ability of mice to develop ECM [21 , 34–36] . Therefore , the contribution of these immune cell subsets to ECM is still a matter of debate . In contrast , the role of CD8+ T cells in ECM is unequivocal . Numerous studies have demonstrated that CD8+ T cell depletion [21 , 31 , 33 , 34] or ablation of effector functions [22 , 37] completely abrogates this disease . Furthermore , parasite-specific CD8+ T cells can mediate ECM in the absence of bystander T cells [38] . Despite the critical role played by CD8+ T cells during ECM , little is known about the dynamics , kinetics , anatomical localization , and function of these cells in vivo . One recent intravital study demonstrated that perivascular CD8+ T cell arrest is a signature of the disease [39] , but it is unclear how these cells cause neurological symptoms or why mice succumb to ECM . Using several techniques , including intravital imaging , we set out in this study to conduct an unbiased examination of how innate and adaptive immune cells contribute to cerebrovascular perturbations during ECM and to identify the cause of this fatal disease . We also sought in vivo insights into how cerebrovascular ECs respond to ECM and whether immune interactions with these cells could be therapeutically manipulated to ameliorate disease pathogenesis . Our data demonstrate that CD8+ T cells drive fatal vascular leakage during ECM , and they accomplish this by interacting in an antigen-dependent manner with cerebrovascular ECs . This results in profound BBB dysfunction and secondary death of neurons , most notably in the brainstem , which likely gives rise to autonomic dysfunction and death . ECM can be prevented by eliminating antigen presentation in cerebrovascular endothelial cells or by displacing parasite-specific CD8+ T cells from CNS blood vessels using an anti-adhesion molecule therapy .
Human cerebral malaria is associated with several hallmark pathologies in the brain parenchyma including vascular hemorrhaging , breakdown of the BBB , and cerebral edema . Similarly , we observed that mice infected with PbA were highly parasitemic , moribund , and showed evidence of profound BBB breakdown and edema at day 6 p . i . ( S1A–S1E Fig ) . We also observed evidence of significant vascular hemorrhaging in the brain parenchyma ( S1F Fig ) that was almost always associated with iRBCs ( S1G Fig ) . Interestingly , evidence of vascular pathology was also found in the meninges , which has not been reported previously ( Fig 1A ) . To gain insights into the immunopathogenesis of ECM , we conducted a temporal analysis of the immune subsets that arrive in the CNS as PbA-infected mice develop neurological symptoms . We used a clinical scoring system of 0 ( moribund ) to 20 ( asymptomatic ) developed by Carroll et al . [40] to assess disease severity ( S1A Fig ) . We separately evaluated the meninges and brains of naïve , d5 p . i . ( parasitemic but asymptomatic ) , and d6 p . i . ( highly parasitemic and symptomatic ) mice . This study revealed a significant increase in CD8+ T cells and Ly6Chi inflammatory macrophages/monocytes in the brains and meninges as mice developed symptoms on d6 p . i . ( Fig 1B and 1C ) . To monitor PbA-specific CD8+ T cells , we generated pentamers consisting of H-2Db loaded with an immunodominant PbA epitope ( SQLLNAKYL ) described by Howland et al . [41] . Interestingly , PbA-specific CD8+ T cells increased in frequency and number in the spleen , meninges , and brain as mice progressed from asymptomatic ( d5 p . i . ) to symptomatic ( d6 p . i . ) ( Fig 1D and 1E ) . Thus , neurological symptoms during ECM are associated with the migration of Ly6Chi monocytes and parasite-specific CD8+ T cells to the CNS . Because myeloid cells are recruited to the CNS during ECM , we examined the dynamics and anatomical distribution of these cells using intravital two-photon microscopy ( TPM ) through a thinned skull window as described [42 , 43] . We monitored the dynamics of myelomonocytic cells ( monocytes and neutrophils ) using lysozyme M-GFP ( LysMgfp/+ ) reporter mice [44] . Consistent with our flow cytometric data Fig 1B and 1C we observed a significant increase in the number of myelomonocytic cells in symptomatic mice at d6 p . i . compared to d5 p . i . and uninfected animals . Interestingly , the myelomonocytic cells were contained almost entirely within the vasculature and were found in both the brain and periphery ( ear ) indicating systemic inflammation ( Fig 2A ) ( S1–S3 Movies ) . These cells were also occasionally associated with vascular leakage in the brain ( S2 Movie ) . To determine the anatomical relationship between myelomonocytic cells and iRBCs during the development of ECM , we infected LysMgfp/+ mice with recombinant PbA expressing mCherry and OVA ( PbA-OVA-mCherry ) [45] . In symptomatic mice at d6 p . i . , we observed fluorescent iRBCs adherent to cerebral vasculature ( Fig 2B; S3 Movie ) , similar to what is found in human CM patients . We also visualized myelomonocytic cells actively phagocytosing these iRBCs while patrolling the vascular lumen ( Fig 2B; S3 Movie ) . In fact , we found that on average nearly 60% of the fluorescent iRBCs were associated with LysM-GFP signal at any one time . In addition to the vascular lumen , some fluorescent iRBCs localized to the perivascular spaces of intact vessels , suggesting extravasation ( S4 Movie ) . These parasites were rapidly acquired by perivascular macrophages visualized in CX3CR1gfp/+ mice ( S4 Movie ) . These data indicate iRBCs were positioned on the luminal and abluminal surface of cerebral blood vessels during the development of ECM . The contribution of innate immune cells to the pathogenesis of ECM is still a matter of contention . Previous studies have either implicated monocytes and neutrophils in the pathogenesis of ECM [18 , 27–29] or shown them to be irrelevant [21 , 33 , 35 , 36] . Our flow cytometric ( Fig 1 ) and imaging ( Fig 2 ) data suggested that these cells might be involved in disease pathogenesis . To determine if myelomonocytic cells contribute to BBB breakdown or mortality during ECM , we depleted neutrophils with anti-Ly6G antibodies in CCR2-/- mice ( S2A Fig , which are deficient in circulating monocytes [46 , 47] . This resulted in an 85% reduction of the total circulating myelomonocytic compartment . When compared to PbA-infected controls , mice lacking monocytes and neutrophils showed no preservation of BBB integrity or reduced mortality rate during ECM ( Fig 2C–2E ) . In addition , depletion of myelomonocytic cells did not improve clinical scores or alter parasitemia levels ( S2B and S2C Fig ) . Based on these results , myelomonocytic cells do not appear to play a significant role in the pathogenesis of ECM . Having ruled out myelomonocytic cells as the cause of fatal pathology during ECM , we focused on the adaptive immune response , with a specific emphasis on T cells given that mice lacking B cells are still susceptible to ECM [34] . We initiated this line investigation by conducting a series a T cell depletion experiments . On d4 p . i . PbA-infected mice were administered antibodies specific for CD8+ or CD4+ T cells ( S3A Fig ) and then monitored for development of ECM relative to untreated control mice ( Fig 3A ) . Although depletion of CD4+ T cells had no effect on survival , CD8+ T cell depleted mice were completely resistant to ECM ( Fig 3A ) despite having levels of parasitemia comparable to control mice ( S3B Fig ) . Depletion of CD8+ T cells also prevented the vascular hemorrhaging ( S3C Fig ) , BBB breakdown ( S3D and S3E Fig ) , and edema ( S3F Fig ) normally associated with ECM . To determine the anatomical localization of activated PbA-specific CD8+ T cells within the brain at the peak of ECM , naïve B6 mice were seeded i . v . with 104 naïve mCerulean+ OT-I T cell receptor transgenic CD8+ T cells and then infected with PbA-OVA-mCherry . Symptomatic mice were imaged 6 days later by TPM . These intravital imaging studies revealed that nearly all PbA-specific CD8+ T cells in symptomatic mice were arrested on or slowly crawling along the luminal and extravascular surfaces of cerebral blood vessels and were often associated with significant vascular breakdown . ( Fig 3B; S5–S7 Movies ) . Interestingly , this activity was specific to the brain , as PbA- specific CD8+ T cells were not observed arresting along the vasculature in a peripheral tissue ( ear ) within the same mouse ( Fig 3B; S5 Movie ) . Although PbA-specific CD8 T cells in the CNS were observed within perivascular spaces and the parenchyma , the majority appeared to be interacting with the luminal surface of blood vessels . To quantify the percentage of luminal vs . extravascular PbA-specific CD8 T cells , we created volumetric masks corresponding to the Evans blue signal in each blood vessel ( Fig 3C , S8 Movie ) . After applying the mask , all visible PbA-specific CD8+ T cells , including those located on the surface blood vessels , were counted as extravascular , whereas cells obscured by the mask were considered luminal ( Fig 3C , S8 Movie ) . We found that the vast majority of PbA-specific CD8 T cells were associated with the luminal surface of CNS blood vessels ( Fig 3D ) . Collectively , these data indicate that CD8+ T cells arrest along the cerebral vasculature during ECM and are responsible for the vascular pathology . Because PbA-specific CD8+ T cells were intimately associated with cerebral vasculature , we postulated that cytotoxic lymphocyte-mediated killing of vascular ECs might serve as the cause of BBB breakdown and death during ECM . To simultaneously assess cell death and vascular leakage in vivo , we injected naïve and symptomatic mice at d6 p . i . intravenously with propidium iodide ( PI ) ( to label dead cells ) and Evans blue ( to assess vascular leakage ) . This assay revealed a striking pattern of pathology in all mice succumbing to ECM ( Fig 4A ) . Whereas evidence of cell death was observed in multiple brain regions ( e . g . olfactory bulb , cortex , cerebellum , brainstem , choroid plexus ) , the brainstem and olfactory bulb pathology were particularly severe ( Fig 4A ) . Both brain regions showed evidence of profound vascular leakage and cell death . The brainstem pathology is consistent with cerebral herniation [48] and would likely give rise to autonomic dysfunction . To determine if the PI+ cells were in fact vascular ECs , we co-stained sagittal brain sections with anti-CD31 antibodies and performed quantitative analyses ( Fig 4B and 4C ) . Although EC death was observed in multiple brain regions during ECM , only a small fraction of the total ECs was PI+ ( Fig 4B and 4C ) . Thus , it is unlikely that EC death is the cause of fatal disease in mice with ECM . We found a small number of PI+ ECs in the brain during ECM , but the vast majority of the dead cells were unknown . Based on cellular morphology and anatomical location , we hypothesized that at least some of the PI+ cells were neurons . Interestingly , co-staining with anti-NeuN antibodies revealed that nearly all of the PI+ cells in the brainstem were neurons—a pattern of cell death that was unique to this brain region ( Fig 4D and 4E ) . Because the brainstem controls vital functions such as the cardiovascular and respiratory systems , it is likely that mice succumb to ECM due to the widespread neuronal death observed in this brain region . We observed evidence of profound vascular leakage in several brain regions , including the brainstem; however , previous studies have suggested that this is not due to reduced expression of endothelial tight junction proteins [18] . We hypothesized that this negative result might be explained by a failure to directly compare tight junction ( TJ ) protein expression in leaking versus intact cerebral vasculature during the development of ECM . To test this hypothesis , we injected uninfected or symptomatic mice at d6 p . i . with Evans blue to locate areas of vascular leakage within the brain ( Fig 5A ) . Next , we stained sagittal brain sections with antibodies against CD31 and claudin-5 to identify ECs and TJs , respectively ( Fig 5B ) . Thick sections were used in order to generate volumetric 3D masks of individual blood vessels in the frontal cortex , cerebellum , and brainstem ( Fig 5B ) . This method provides a more accurate representation of claudin-5 staining over an entire blood vessel than would be obtained by performing 2D analyses on thin sections . When we quantified the intensity of claudin-5 staining on 3D reconstructed blood vessels in various regions of the brain , we consistently found reduced expression in areas of Evans blue+ vascular leakage within symptomatic mice when compared to the same areas in uninfected mice ( Fig 5C ) . Furthermore , we found that claudin-5 levels in brain regions of symptomatic mice where there was no vascular leakage were comparable to the levels observed on naïve blood vessels ( Fig 5C ) . We were unable to find areas without vascular leakage in the brainstem due to the extensive amount of pathology in this brain region . Thus , by comparing to leaking to intact cerebral blood vessels , we uncovered that vascular leakage is indeed associated with reduced TJ expression . We observed the PbA-specific CD8+ T cells interact heavily with cerebral vasculature during ECM , but the vast majority of ECs survive these engagements . To better understand the mechanisms guiding these interactions , we conducted flow cytometric analyses of cerebral ECs during the development of ECM and compared them to ECs extracted from a peripheral tissue ( ear ) ( Fig 6A–6C ) . By gating on live , CD45-CD31+ cells ( S4A Fig ) , we noted that adhesion ( ICAM-1 , VCAM-1 ) and antigen-presenting ( I-Ab , Db , Kb ) molecules were all significantly increased on ECs from symptomatic mice at d6 p . i . ( Fig 6A and 6B ) . Elevated expression of these molecules appeared as early as d4 p . i . and usually increased further as the disease progressed ( Fig 6B ) . ECs extracted from meningeal blood vessels were similarly activated ( S4B Fig ) . Interestingly , this EC activation phenotype was unique to the CNS , as ECs extracted from a peripheral site ( ear ) showed a significantly reduced expression level of adhesion and antigen presentation molecules ( Fig 6C ) . This is consistent with our intravital imaging data showing slow crawling and arrest of PbA-specific CD8+ T cells along cerebral , but not ear vasculature ( Fig 3B; S5 Movie ) . This finding also suggests that the increase in myelomonocytic cells observed in the ear vasculature of PbA-infected mice did not induce EC activation . We next sought insights into the mechanism underlying cerebral EC activation during ECM . Previous studies have shown that IFNγ-deficient mice are protected from ECM [34 , 49 , 50] . We confirmed these findings ( S4C–S4E Fig ) and hypothesized that IFNγ produced by lymphocytes recruited to the CNS might be responsible for cerebral endothelial cell activation during ECM ( Fig 6B ) . A link has been established between IFNγ and ICAM-1 expression in the brain during ECM , but the cell type ( s ) affected by the absence of IFNγ was not determined [51 , 52] . To address the role of IFNγ in EC activation during ECM , we infected wild type and IFNγ-/- mice with PbA and examined EC phenotype on day 6 . ECs from IFNγ-/- mice had significantly reduced levels of adhesion and antigen presentation molecules ( Fig 6D ) . In fact , with the exception of ICAM-1 and H-2Db , the loss of IFNγ reduced the activation state of brain ECs to uninfected levels . These data correlated with a severe reduction in PbA-specific CD8 T cell accumulation in the brain , despite normal peripheral expansion and migration ( S4F Fig ) . Furthermore , PbA-infected IFNγ-/- mice lacked the severe vascular breakdown and brainstem neuronal cell death observed in wild type mice ( Fig 6E and 6F ) . These data indicate that IFNγ is responsible for cerebral EC activation during ECM and that IFNγ-deficiency likely protects mice in part by keeping ECs in a naïve state . Previous studies using CBA/J mice have demonstrated that antibody blockade of LFA-1 during ECM is highly efficacious at preventing disease [17 , 53 , 54] . However , we have found that this treatment is ineffective in PbA-infected C57BL/6 mice . Activated CD8+ T cells express multiple adhesion molecules , including LFA-1 and VLA-4 , which are ligands for ICAM-1 and VCAM-1 , respectively . Given that brain ECs upregulate both ICAM-1 and VCAM-1 during ECM ( Fig 6B ) , we hypothesized that administration of a combination of anti-LFA-1 and anti-VLA-4 antibodies ( anti-LFA-1/VLA-4 ) could be used to therapeutically displace PbA-specific CD8+ T cells from cerebral vasculature and prevent fatal disease . We administered this therapy on day 5 . 5 post-infection to avoid interfering with T cell priming . Mice at this time point were parasitemic ( S5B Fig ) and had symptoms associated with ECM ( S5C Fig ) . Importantly , treatment with anti-LFA-1/VLA-4 completely reversed these symptoms and prevented death in PbA-infected mice ( S5A and S5C Fig ) . This treatment had no effect on PbA-specific CD8+ T cell expansion ( S5D and S5E Fig ) . To assess the impact of anti-LFA-1/VLA-4 therapy on PbA-specific CD8+ T cell dynamics in the brain , we performed a series intravital imaging studies . At day 6 following infection with PbA-OVA , we imaged the dynamics of mCerulean+ OT-I T cells in the brain for 30 min by TPM . This was followed by intravenous administration of anti-LFA-1/VLA-4 therapy and an additional 30 min of imaging in the same anatomical location . Before antibody treatment , PbA-specific CD8+ T cells were observed slowly crawling along and arresting on cerebral blood vessels ( Fig 7A; S9 Movie ) . In contrast , anti-adhesion antibody treatment resulted in an immediate displacement of PbA-specific CD8+ T cells from the vasculature ( Fig 7A; S9 Movie ) . A significant reduction in the frequency of PbA-specific CD8+ T cells associated with brain vasculature was observed for the entire viewing period after anti-adhesion therapy ( Fig 7B ) , which was not seen in mice treated with istoype control antibodies ( Fig 7C ) . Interestingly , the frequency of PbA-specific CD8+ T cells in the parenchyma and perivascular spaces was not affected by blocking adhesion molecules ( Fig 7D ) , suggesting that intravascular CD8+ T cell interactions are the ones responsible for fatal pathology during ECM . Furthermore , disruption of luminal CD8 T cell interactions with brain ECs via adhesion molecule blockade also prevented the death of brainstem neurons observed in isotype control treated mice ( Fig 7E and 7F ) . In concert , these data show that blocking adhesion to ICAM-1 and VCAM-1 , which are highly expressed on brain ECs during ECM , prevents PbA-specific CD8+ T cells from arresting along cerebral vasculature and rescues mice from fatal pathology and disease . During the development of ECM , we routinely observed PbA-specific CD8+ T cells dividing following arrest on the luminal surface of cerebral vasculature ( Fig 8A; S10 Movie ) . Because cognate peptide-MHC I interactions can advance the cell cycle program of effector CD8+ T cells [55] , we hypothesized that PbA-specific CD8+ T cells interact with the brain vasculature in an antigen dependent manner . To address this hypothesis , we first compared the dynamics of PbA-specific vs . bystander CD8+ T cells of an irrelevant specificity in cerebral blood vessels . Mice were seeded with mCerulean+ OT-I T cells and then infected with PbA-OVA . When these mice became symptomatic on day 6 post-infection , we intravenously injected yellow fluorescent protein ( YFP ) + DbGP33-41 CD8+ T cells ( YFP+ P14 ) purified from the spleens of a separate group of mice infected 8 days earlier with lymphocytic choriomeningitis virus ( LCMV ) . YFP+ P14 cells were used as activated bystander CD8+ T cells because they are specific to the LCMV glycoprotein ( GP ) , not PbA . TPM imaging and subsequent analysis of these two CD8+ T cell populations revealed that PbA-specific CD8+ T cells moved at a significantly slower speed ( Fig 8B ) and spent more time arrested along cerebral vasculature ( Fig 8C and S6A Fig ) than the bystander CD8+ T cells . These results demonstrate that antigen-specificity dictates the interaction between CD8+ T cells and brain microvasculature during ECM . To further demonstrate the specificity of PbA-specific CD8+ T cell interactions , we used TPM to evaluate the dynamics of these cells following injection of an anti-peptide MHC I blocking antibody . PbA-OVA infected mice seeded mCerulean+ OT-I cells were imaged by TPM on day 6 post-infection . Midway through the imaging session , mice were injected i . v . with anti-Kb-SIINFEKL ( the peptide MHC complex recognized by OT-I cells ) or isotype control antibodies . Injection of anti-Kb-SIINFEKL , but not isotype , control antibodies significantly elevated the velocity PbA-specific CD8+ T cells ( Fig 8D ) , further supporting that the interactions with cerebral ECs are antigen-specific . Next , we set out to determine the functional importance of PbA-specific CD8+ T cell interactions with cerebral ECs during the development of ECM . This was accomplished by generating bone marrow ( BM ) chimeras in which MHC I deficient hosts ( Kb-/-Db-/- mice ) were reconstituted with wild type bone marrow ( Fig 8E ) . These mice were incapable of presenting MHC I peptides on ECs and other stromal cells while maintaining normal hematopoietic presentation . Irradiated wild type mice receiving wild type bone marrow served as a control for this experiment . Interestingly , the Kb-/-Db-/- chimeras were nearly all resistant to fatal ECM , whereas the wild type controls succumbed to disease as expected ( Fig 8F and S6C Fig ) . Protection from ECM was observed in Kb-/-Db-/- mice despite normal parasitemia levels ( S6B Fig ) and generation of an equal , if not greater , PbA-specific CD8+ T cell response relative to the wild type controls ( Fig 8G and 8H ) . To evaluate how MHC I deficiency affected PbA-specific CD8 T cell interactions with the brain vasculature during ECM , we seeded wild type and Kb-/-Db-/- BM chimeras with mCerulean+ OT-I cells and then used TPM to monitor their intravascular motility 6 days following infection with PbA-Ova . Quantification of PbA-specific CD8+ T cells in the cerebral vasculature of the Kb-/-Db-/- BM chimeras revealed significantly increased velocities ( Fig 8I ) and reduced arrest within the lumen of brain blood vessels ( Fig 8J and S6D Fig ) relative to the wild type controls . Because there are no other radio-resistant cells within the vascular lumen , these data suggest that PbA-specific CD8+ T cells engage cerebral ECs in an antigen-dependent manner during ECM . Because PbA-infected Kb-/-Db-/- BM chimeras were resistant to ECM , we set out to determine whether they were also free from brainstem pathology normally associated with this disease ( Fig 4A , 4D and 4E ) . Brains from wild type BM chimeras injected with PI and Evans blue at the peak of disease revealed extensive vascular leakage and cell death in the brainstem ( Fig 9A ) . However , Kb-/-Db-/- BM chimeras were free of pathology at this same time point ( Fig 9A ) . Nearly all of the dead cells in the brainstems of wild type BM chimeras were neurons , whereas brainstem neurons in Kb-/-Db-/- BM chimeras were unaffected ( Fig 9B and 9C ) . These results suggest that antigen presentation by brain ECs , which fosters increased interactions with PbA-specific CD8 T cells , leads to severe brainstem pathology during ECM .
The pathological mechanisms underlying HCM are not entirely understood . Because ECM shares many of the pathological features of HCM , we set out to uncover novel mechanistic insights into the immunopathogenesis of this disorder . We made several important observations that significantly advance our understanding of cerebral malaria . Studies have shown that CM in humans and rodents is associated with BBB breakdown , edema , and hemorrhaging . During the peak of ECM , we noted that the meninges in addition to the brain parenchyma show evidence of profound vascular pathology , and this was associated a reduction in tight junction protein expression . Importantly , death from ECM is linked to marked vascular leakage and neuronal cell death in the brainstem , which is consistent with the edema and subsequent cerebral herniation recently observed in children with HCM [9] . Mechanistically , we uncovered that this fatal disease is caused by the activities of parasite-specific CD8+ T cells operating along cerebral blood vessels . As the disease developed , cerebrovascular ECs were highly activated by IFNγ , which promoted induction of cell adhesion and antigen presenting molecules . This in turn facilitated cognate peptide-MHC I dependent engagement by parasite-specific CD8+ T cells primarily on the luminal surfaces of cerebral blood vessels . The pathological significance of these interactions was demonstrated in mice rendered genetically deficient in their ability to present antigen in MHC I on ECs . These mice had reduced cerebrovascular engagement by CD8+ T cells and were resistant to fatal disease . Lastly , therapeutic administration of antibodies specific for VLA-4 and LFA-1 rapidly displaced CD8+ T cells from cerebral blood vessels and promoted survival , thus providing a simple yet effective means to treat this disease . One of the most interesting findings in our study is the pathology observed in mice with severe ECM . By simultaneously injecting Evans blue and propidium iodide , we were able to evaluate the relationship between vascular leakage and cell death in mice succumbing to ECM . While vascular leakage was notable through the brain and meninges , the most striking areas of pathology were the olfactory bulb and brainstem . Vascular pathology was previously reported in the olfactory bulb of mice with ECM and linked to a decline in their sense of smell [56] . The severe pathology observed in the brainstem , however , is more relevant from the perspective of survival . Profound vascular leakage and neuronal cell death was seen in all mice succumbing to ECM and was associated with reduced expression of the tight junction protein , claudin-5 . Sudden neuronal depolarization and death in this brain region would cause cardiorespiratory failure as observed in other neurological disorders [57 , 58] . Importantly , a recent magnetic resonance imaging ( MRI ) study revealed evidence of severe brain swelling in 84% of children with HCM [9] . Brainstem herniation was also demonstrated in fatal instances of this disease . Our ECM results are consistent with brainstem herniation and pathology being the cause of death , which is supported by a MRI study showing significant displacement of the cerebellum and brainstem in mice with fatal ECM [48] . Therefore , increased intracranial pressure leading to cerebral herniation is the likely cause of death in rodents and children with CM . To gain insights into the mechanisms that give rise to fatal edema during CM , we examined the activities of innate and adaptive immune cells . Our flow cytometric and two-photon imaging data revealed that ECM is associated with recruitment of innate and adaptive immune cells to blood vessels in the meninges and brain parenchyma . A strong consensus exists in the literature among several studies showing that CD8+ T cells play an essential role in ECM pathogenesis [21 , 22 , 31 , 33 , 34 , 37 , 38] , and our data support this conclusion . There is , however , some debate regarding the role of innate immune cells , such as monocytes and neutrophils , in this disease . As mice developed ECM , our intravital time lapses revealed myelomonocytic cells migrating along cerebral blood vessels and acquiring adherent iRBCs . These cells were associated on occasion with vascular leakage; however , depletion had no impact on BBB breakdown , edema , or survival . This is different from the significant vascular disruption induced by synchronously extravasating myelomonocytic cells during fatal viral meningitis [42] . Some studies have implicated myelomonocytic cells in the pathogenesis of ECM [18 , 27–29] , whereas others have not [21 , 33 , 35 , 36] . The difference in the outcome of these studies could be linked to many different variables , including strain of mice , depletion strategy , inadvertent blockade of T cells , and genetic variation in the strain of Plasmodium used . Regardless of the explanation , our findings suggest that while myelomonocytic cells may contribute to disease , they are not a dominant participant like CD8+ T cells . At the peak of ECM , we observed parasite-specific CD8+ T cells slowly rolling , arresting , and dividing within the lumen of CNS vasculature . These interactions were highly associated with breakdown of the BBB and the flow of vascular contents into the meninges and parenchyma . Several lines of investigation have suggested that CTL-mediated killing of cerebrovascular ECs might be the cause of CNS vascular breakdown during ECM . Mice deficient in CTL effector pathways such as perforin and granzyme B are resistant to ECM [22 , 37] , apoptotic ECs have been identified in the retina [59] and brain [60 , 61] during disease development , and CD8+ T cells isolated from ECM mice kill parasite-loaded ECs in vitro [62] . However , using a sensitive in vivo approach to quantify cell death , we were unable to demonstrate evidence of widespread EC death in the brains of highly symptomatic mice , but were able to show that Evans blue leakage was associated with reduced expression of tight junction proteins . This is consistent with other studies showing minimal EC death during ECM [17 , 39] . Therefore , it is unlikely that CD8 T cells mediate ECM by directly killing brain ECs . Alterations in tight junction protein expression are the most likely explanation for vascular leakage in this model , although the low percentage of EC death we observed could certainly contribute to vascular leakage and surrounding brain pathology . From the standpoint of vascular pathogenesis , we favor a mechanism whereby CD8+ T cells induce reversible alterations in EC tight junctions ( such as claudin-5 ) that cause cerebrovascular leakage during ECM . A previous study demonstrated that CD8+ T cells can traverse the BBB following recognition of cognate peptide MHC I complexes on the lumen of cerebral ECs [63] . It was discovered more recently that CD8+ T cells can actually use granzyme B in a nontraditional manner to cleave vascular basement membrane [64] . This mechanism allows CD8+ T cells to extravasate across vasculature . Considering that we and others [39] have found parasite-specific CD8+ T cells along CNS vasculature during ECM , it is possible that the cumulative vascular breaks associated with CD8+ T cell extravasation contribute to global breakdown of the BBB . This event would be mitigated in granzyme B knockout mice , which are resistant to ECM [37] . CD8+ T cells could also use IFNγ to remodel the CNS vasculature . Studies using in vitro cultured ECs have shown that IFNγ causes cytoskeletal rearrangement and decreased barrier integrity [65 , 66] . During HSV-2 infection , T cell-derived IFNγ has been shown to open the BBB , facilitating antibody access to the CNS [67] . Therefore , PbA-specific CD8+ T cells engaging CNS ECs in an antigen-specific manner could induce barrier openings through IFNγ release . In addition , IFNγ can promote cross presentation of antigen by cerebral ECs [62] as well as production of CXCL9 and 10 [68] , chemokines known to attract CD8+ T cells . Thus , it is conceivable that engagement of peptide MHC I complexes by CD8+ T cells on cerebral ECs results in IFNγ release that opens tight junctions and recruits additional parasite-specific T cells . This type of amplification loop localizing primarily to cerebral vasculature would result in a rapid deterioration of the CNS barrier system . Consistent with this theory , deletion of IFNγ significantly reduced accumulation of parasite-specific CD8+ T cells along cerebral vasculature and completely eliminated fatal brainstem pathology . The precise localization of parasite-specific CD8+ T cells to cerebrovascular ECs led us to consider whether this pattern was unique to the CNS . To address this question , we examined the vasculature of the ear skin as a representative peripheral location . Interestingly , our intravital imaging studies revealed that parasite-specific CD8+ T cells localized along brain but not ear vasculature . Moreover , vascular ECs extracted from the ear showed a significantly reduced expression of adhesion ( ICAM-1 and VCAM-1 ) and antigen presenting ( MHC I , MHC II ) molecules when compared to cerebral ECs . This is an intriguing observation given that malaria is a systemic disease characterized by circulating iRBCs and inflammatory cytokines . Thus , vascular ECs throughout the body should have access to the same inflammatory mediators . The specific changes in cerebral ECs are best explained by parasite-specific T cell engagement following local antigen presentation . We observed upregulation of MHC I and II molecules on cerebral ECs , which would allow direct engagement by CD8+ and CD4+ T cells , respectively . These two T cell subsets are known to secrete IFNγ upon engagement and amplify the activities of one another during ECM [32] . Although we and others have shown that activation of the CNS endothelium occurs during ECM ( Fig 6C ) [69] , we have provided direct evidence that this is driven by IFNγ ( Fig 6D ) . Because antigen recognition occurs on cerebral ECs , it is likely that T cell-derived IFNγ plays a major role in the activation of the CNS vasculature . However , it remains to be determined why antigen presentation and T cell engagement occurs primarily on cerebral blood vessels and not vasculature residing in a peripheral site such as the ear . Features of the infecting parasite are also likely to contribute to severe disease . In children , one particular var gene product expressed on the iRBC surface to endothelial protein C receptor ( EPCR ) is associated with CM [70 , 71] , and brain autopsies of Malawian children who died from CM showed loss of EPCR at sites of sequestered iRBCs [72] . In mice , ECM-inducing PbA differs by only 18 non-synonymous mutations in open reading frames from the NK65 parasite that does not cause ECM , suggesting that these parasite genes may contribute to the unique brain pathology during ECM . The localization of parasite-specific CD8+ T cells along cerebral vessels [39] does not alone prove that these cells are engaged in cognate peptide MHC I interactions or that these interactions are important for disease development . To demonstrate the specificity of these interactions , we conducted several lines of investigation . We demonstrated previously that CTL division could be advanced by peptide MHC I interactions in the virally infected meninges [55] . Interestingly , we observed a similar pattern of parasite-specific CD8+ T cell division in the lumen of cerebral blood vessels during ECM , suggesting antigen-specific interactions . The specificity of the vascular CD8+ T cell interactions was proven by injecting anti-Kb-SIINFEKL antibodies intravenously during ECM . This increased the velocity of luminal parasite-specific CD8+ T cells , indicating that the interactions with ECs were in fact antigen specific . This conclusion was further supported by the preferential arrest of parasite-specific CD8+ T cells on cerebral vasculature when compared to bystander CD8+ T cells of an irrelevant specificity . Lastly , the functional importance of the interactions was demonstrated by removing MHC I from the stromal compartment through the generation of bone marrow chimeras . Expression of MHC I on the hematopoietic system but not stromal cells such as vascular endothelium significantly reduced parasite-specific CD8+ T cell arrest on cerebral ECs and promoted survival . In vitro studies have shown that human and murine brain ECs have the ability to internalize and cross-present Plasmodium antigen to CD8+ T cells [41 , 62] . In fact , ex vivo cultured brain ECs from mice with ECM have the capacity to stimulate CD8+ T cells in an antigen-specific manner . These data collectively demonstrate that CD8+ T cells mediate fatal vascular pathology during ECM via antigen-dependent interactions with cerebrovascular ECs . While murine and human CM share similar pathologies , the role of CD8+ T cells in HCM remains unknown . Less attention is given to CD8+ T cells during HCM because they are difficult to find in human post-mortem brain samples [73] . It should be noted , however , that we and others [21] have also found it difficult to identify intravascular CD8+ T cells in post-mortem brain samples from mice with ECM . An inability to observe an abundance of these cells histologically does not negate their involvement in HCM . It is unfortunately not possible to examine parasite-specific CD8+ T cells intravitally during HCM ( as we have done in mice ) ; however , genetic studies offer some clues regarding their involvement in this disease . For example , resistant vs . susceptibility to HCM has been linked to specific human leukocyte antigen class I alleles [74 , 75] . In addition , the CD8 T cell chemoattractant , CXCL10 , is a strong biomarker for HCM [76–78] , and a genetic polymorphism that elevates CXCL10 levels was linked to an increased incidence of HCM [79] . Importantly , CXCL10 blockade or deficiency is partially protective against development of ECM [25 , 26] . It remains to be determined whether CD8+ T cells are definitively involved in the pathogenesis of HCM , but further studies are warranted given the similarities between HCM and ECM . Without knowing the specificity of disease-inducing parasite-specific T cells , it would be difficult to therapeutically target these cells during CM . Thus , we surmised that a non-specific displacement of T cells from the cerebral vasculature would provide a more effective means to thwart this disease . We demonstrated that late blockade with antibodies against LFA-1 and VLA-4 completely prevented development of ECM , which is consistent with other studies showing the effectiveness of adhesion molecule blockade in this model [17 , 53 , 54] . However , the mechanism by which this therapy prevented disease was unclear . Using intravital imaging , we observed that anti-adhesion molecule therapy completely disengages CD8+ T cells from the CNS vasculature at the peak of disease without affecting the number of extravascular CD8+ T cells . These data indicate that this disease can be treated by interfering with luminal interactions between CD8+ T cells and cerebrovascular ECs . Natalizumab ( anti-VLA-4 ) and Efaluzimab ( anti-LFA-1 ) are both FDA approved drugs used to treat human inflammatory diseases [80–82] , opening the possibility of an expedited intervention in patients with HCM . It would be important to begin treatment in patients with edema prior to the development of cerebral herniation , which would result in irreversible brainstem pathology . In conclusion , we have provided the first in vivo evidence detailing how CD8+ T cells cause ECM . Parasite-specific CD8+ T cells engage CNS ECs in an antigen-dependent manner , which leads to profound vascular breakdown ( associated with alterations in tight junction protein expression ) , edema , loss of brainstem neurons , and subsequent death . Despite CTL engagement , we observed little evidence of EC death during ECM . We therefore hypothesize that CD8+ T cells use a noncytopathic mechanism to disrupt cerebrovascular EC tight junctions . This signifies that the disease is reversible up to the point when severe edema gives rise to cerebral herniation and death of brainstem neurons ( an irreversible event ) . The reversibility of the disease is demonstrated by the effectiveness of late anti-adhesion molecule blockade . Because many patients with CM likely arrive in the hospital with active BBB breakdown , effective therapies need to target parasite acquisition by ECs and subsequent CD8+ T cell engagement . In addition to anti-malarial drugs and supportive care , consideration should also be given to therapeutics that interfere with T cell function / metabolism [83 , 84] or that temporarily displace these cells from cerebral vasculature ( e . g . anti-VLA-4 / LFA-1 ) . A reduction of parasite burden in combination with a transient disruption of T cell activity should give the BBB sufficient time to repair and for neurological disease to subside .
C57BL/6J ( B6 ) , B6 . 129S7-Ifntm1Ts/J ( IFNγ-/- ) , B6 . 129S4-Ccr2tm1Ifc/J ( CCR2-/- ) B6 . 129P-Cx3cr1tm1Litt/J ( CX3CR1gfp/gfp ) [85] , B6 . SJL-Ptprca Pepcb/BoyJ ( Ly5 . 1 ) , and BL/6-Tg ( TcraTcrb ) 1100Mjb/J ( OT-I ) mice were purchased from The Jackson Laboratory ( Bar Harbor , ME ) . CX3CR1gfp/gfp , OT-I , and Ly5 . 1 were then bred and maintained under specific pathogen free conditions at the National Institute of Health ( NIH ) . B6 DbGP33–41 TCR-tg ( P14 ) [86] , B6 LysMgfp/+ [44] , B6 H-2Kb-/-Db-/- , B6 actin-mCerulean , and B6 actin-YFP were also bred and maintained at the NIH . CX3CR1gfp/+ mice were generated by crossing B6 mice with CX3CR1gfp/gfp mice . Actin-mCerulean and actin-YFP mice were made as described below . YFP+ P14 and mCerulean+ OT-I mice were derived from the following F1 crosses: actin-YFP x P14 and actin-mCerulean x OT-I , respectively . All mice bred in house were confirmed to be on a pure C57BL/6J background by SNP analysis ( Charles River ) . All transgenic mice were generated by the National Institute of Mental Health ( NIMH ) Transgenic Core Facility . Transgenic mice expressing monomeric cerulean ( mCerulean ) under the chicken β-actin promoter were generated as described [87] by first PCR amplifying the entire 717 bp mCerulean coding region from the pCMV-mCerulean vector ( Addgene ) using primers ( fwd: 5’ATATATGAATTCGCCACCATGGTGAGCAAGGGCGAGG3’; rev: 5'ATATATCTC GAGTTACTTGTACAGCTCGTCCATG3' ) containing EcoRI and XhoI restriction sites . This cDNA was cloned into the same sites following removal of the GFP sequence from the pCAG-GFP vector ( Addgene ) . The resultant plasmid was digested with Sal I / HindIII , and a 2925 bp fragment containing the CMV early enhancer/chicken β-actin ( CAG ) promoter , mCerulean , and PolyA sequence was prepared for microinjection into the pronuclei of fertilized mouse eggs . Mice expressing the Venus variant of yellow fluorescent protein ( YFP ) under the CAG promoter were generated in a similar manner . The entire 717 bp coding sequence for YFP was PCR amplified using the aforementioned primers and cloned into pCAG-GFP following removal of GFP with EcoRI and XhoI . A 2925 bp fragment consisting of the CAG promoter , YFP , and a PolyA sequence cut with Sal I / HindIII and prepared for microinjection . To generate all transgenic mice , linearized constructs were injected into C57BL/6J eggs . Following selection of transgene positive founder lines , all mice were backcrossed one additional generation onto C57BL/6J background before intercrossing . PbA was maintained and used as previously reported [83] . PbA . OVA::mCherryhsp70 ( PbA-OVA-mCherry ) parasites were kindly provided by C . Janse and S . Khan [45] . PbTg ( PbA-OVA-GFP ) parasites were kindly provided by W . Heath [38] . All mice were infected intraperitoneally ( i . p . ) with 106 parasitized red blood cells ( pRBCs ) . Parasitemia was determined by Giemsa-stained thin blood smear or by flow cytometry as described [88] . Mice were monitored daily for neurological symptoms of ECM using a quantitative scale as described [40] . For survival studies , mice were euthanized upon reaching a score of 0 . For LCMV infections , mice were injected with 105 plaque forming units of LCMV Armstrong 53b strain i . p . Mice were i . v . injected with 20mg Evans blue ( Sigma ) per kg body weight at the indicated time points p . i . After 4 hours , brains were extracted following saline perfusion and then processed as described previously [42] . Brain water content was measured as previously described [19] . Weights of brains removed at indicated time points were compared to the dry weight after overnight incubation at 80°C . Mice were anesthetized with chloral hydrate and perfused with 4% paraformaldehyde in PBS for H&E analysis , 10% neutral buffered formalin ( NBF ) for meningeal whole mounts , or 2% NBF in PBS for all other immunohistochemical stains . For meningeal whole mounts , skull caps were removed and incubated overnight in 10% NFB . Following a brief wash , meninges ( including dura , arachnoid , and partial pia layers ) were carefully removed from the bone with fine-tipped forceps and placed in PBS + 2% fetal bovine serum ( FACS buffer ) at 4°C and stained with primary antibody overnight . After a brief wash , the meninges were stained with secondary antibodies for 1 hr at room temperature , followed by another brief wash and then a DAPI stain for 5 min . The meninges were placed in mounting medium on a glass slide , spread out and flattened with forceps , and cover-slipped . Brain tissues were extracted following perfusion and incubated either 2 hours for TJ stains or overnight for all other stains in the same solution with which they were perfused . After a brief wash in PBS , brain tissues were incubated in 30% sucrose in PBS for 24 hours . H&E staining was performed by HistoServe , Inc . For all other stains , tissues were frozen in Tissue-Tek optimal cutting media ( Thermo Fisher Scientific ) . A Leica CM1850 cryostat was used to cut 40 μm thick tissue sections for TJ staining or 20 μm sections for all other stains . Primary stains were performed overnight at 4°C , and secondary and tertiary stains were performed for 1 hr at room temperature with 3 x 5 min washes after each stain . DAPI ( Sigma ) was added for 5 min at RT to label cell nuclei . Following staining , 1 drop of FluorSave Reagent ( Calbiochem ) was added to each section before addition of a coverslip . The following primary antibodies were used: anti-Claudin-5 Alexa Fluor 488 ( 4C3C2 ) ( Thermo Fisher Scientific ) ( 1:100 ) , polyclonal anti-laminin ( Abcam ) ( 1:200 ) , anti-NeuN ( A60 ) ( EMD Millipore ) ( 1:250 ) , polyclonal anti-PECAM-1 ( CD31 ) ( EMD Millipore ) ( 1:40 ) , and anti-TER119 ( Biolegend ) ( 1:200 ) . All secondary antibodies and staining reagents used for immunohistochemistry were purchased from Jackson ImmunoResearch including anti-rat Alexa Fluor 647 , anti-goat Rhodamine Red-X , biotin anti-goat , biotin anti-rabbit , streptavidin Alexa Fluor 488 , and streptavidin Rhodamine Red-X . All secondary antibodies were used at a concentration of 1:500 with the exception of streptavidin conjugates , which were used at a concentration of 1:1000 . H&E images were acquired using a Nikon Eclipse Ci microscope with 4x/0 . 2 NA and 40x/0 . 75 objectives . Fluorescent images were acquired using an Olympus FV1200 laser scanning confocal microscope equipped with 405 , 458 , 488 , 515 , 559 , and 635 laser lines , 4 side window PMTs for simultaneous 4 channel acquisition , and 4x/0 . 16 NA , 10x/0 . 4 NA , 20x/0 . 75 , 40x/0 . 95 , and chromatic aberration corrected 60x/1 . 4 NA objectives . For cell death analyses , 10–12 0 . 4 mm2 field images were collected from the brain sections of each mouse . Each field was chosen at random within one of four brain regions: olfactory bulb , cerebrum , cerebellum , and brainstem . Images were analyzed using Imaris 7 . 6 . 4 software . Only cells with PI staining throughout the nucleus were counted as dead . To determine co-localization of PI and NeuN , PI+ nuclei were first identified and labeled using the spots function with Imaris software . Each spot was then masked in the NeuN channel to reveal all the NeuN+PI+ co-staining . For tight junction analysis , areas of vascular hemorrhaging were first identified within brain sections of symptomatic mice at d6 p . i . by Evans blue staining . We focused specifically on the frontal cortex , brainstem , and cerebellum . Within those regions , xyz images ( each 808 , 992 μm3 ) were captured by confocal microscopy from each brain region per mouse . Three to four z-stacks were captured from areas with and without Evans blue leakage . Images from the corresponding brain regions of uninfected mice were also collected . Within each group , we then quantified the intensity of claudin-5 expression on 30–32 blood vessels as described previously [89] . Briefly , using Imaris 7 . 6 . 4 software , contours were generated around individual blood vessels based on CD31 staining and used to create a 3D surface . The claudin-5 intensity per unit area of an individual blood vessel within this surface was then calculated as follows: ( total # voxels x claudin-5 MFI ) / total surface area of vessel . Mice were injected with 200μg Evans blue dye i . v . for one hour and 50μg propidium iodide ( Invitrogen ) i . v . for 30 minutes . Mice were anesthetized , perfused , and brain tissue was processed as described above . Anesthetized mice received an intracardiac perfusion with PBS to remove all blood leukocytes other non-adhered cells . Single-cell spleen suspensions were prepared by mechanical disruption through a 100 mm mesh barrier followed by RBC lysis with Ack lysis buffer ( 0 . 15M NH4Cl , 10mM KHCO3 , 0 . 1mM EDTA ) . Leukocytes were isolated from the meninges , using forceps to gently separate them from the underside of skull cap ( exactly as performed for immunohistochemistry whole mounts ) followed by enzymatic digestion in 2mg/ml collagenase D ( Roche ) + 50 mg/ml DNase ( Roche ) in RPMI for 30min at 37°C . Leukocytes were isolated from the brain as described [90] . EC isolation was adapted from [91] . Briefly , whole meninges and minced brains from perfused mice were enzymatically digested in 0 . 8mg/ml Dispase II + 0 . 2mg/ml Collagenase P + 0 . 1mg/ml DNase ( all Roche ) in RPMI at 37°C for 15 min with gentle shaking followed by 15 min of mechanical comminution at 37°C . Following digestion , supernatants were isolated and washed . Meningeal ECs were ready for staining but brain EC preps were resuspended in a 40% Percoll ( GE Healthcare ) gradient in HBSS and centrifuged to remove excess myelin . Surface staining and Fc blocking was performed as described [90] . Dead cells were excluded from the analysis by using the LIVE/DEAD fixable Blue Cell Staining kit ( Invitrogen ) . The following antibodies and reagents from BioLegend were used: CD4 PE ( RM4-5 ) , CD8a APC ( 53–6 . 7 ) , CD8β . 2 FITC ( 53–5 . 8 ) , CD11b Brilliant Violet 605 ( M1/70 ) , CD31 PE/Cy7 ( 390 ) , CD45 . 1 AlexaFluor 647 ( A20 ) , CD45 . 2 Alexa Fluor 700 ( 104 ) , CD45 . 2 FITC ( 104 ) , CD54 Alexa Fluor 488 ( YN1/1 . 7 . 4 ) , CD106 biotin ( 429 ) , Gr-1 PE ( RB6-8CJ ) , H-2Db Alexa Fluor 647 ( KH95 ) , H-2Kb PerCP/Cy5 . 5 ( AF6-88 . 5 ) , IAb/IEb Pacific Blue ( M5/114 . 15 . 2 ) , Ly6C APC/Cy7 ( HK1 . 4 ) , Ly6G PE ( 1A8 ) , Streptavidin Brilliant Violet 605 , Thy1 . 2 Alexa Fluor 700 ( 30-H12 ) , and corresponding isotype controls . To identify PbA-specific CD8+ T cells , single cell suspensions were first stained with H-2Db-SQLLNAKYL+ pentamers ( ProImmune ) in PBS + 10% BSA for 10 min at room temperature before surface staining as described above . PbA-specific CD8+ T cells were also identified by IFNγ production following in vitro stimulation with SQLLNAKYL peptide . Briefly , 2 x 106 splenocytes from the denoted mice were incubated with 1μg/ml SQLLNAKYL peptide ( AnaSpec ) , 100U/ml IL-2 ( NIH ) , and 10μg/ml BFA ( Sigma ) in RPMI complete media at 37°C for five hours . To detect intracellular IFNγ production , single cell suspensions were surface stained as described above , treated with cytofix/cytoperm ( BD ) , and then stained intracellularly with anti-IFNγPE/Cy7 ( XMG1 . 2 ) ( BioLegend ) . Samples were acquired using an LSRII flow cytometer ( BD ) , and data were analyzed using FlowJo software version 9 . 7 . 2 ( Tree Star ) . All antibodies used for cell depletion and blocking assays were purchased from BioXcell . Mice were depleted of neutrophils by injecting 500 μg of anti-Ly6G ( clone 1A8 ) i . p . on days -1 and 3 p . i . Mice were depleted of CD4 or CD8+ T cells by i . p . injection of 500 μg ( clone GK1 . 5 ) or 200 μg anti-CD8 ( clone YTS 169 . 4 ) , respectively , on d4 p . i . Cell depletion efficiency was calculated in the blood using the following formula: 100 – ( ( frequency of targeted cell population of a mouse / the average frequency of the targeted cell population of the 4–5 untreated mice ) x 100 ) . For adhesion molecule blocking assays mice were treated with a combination of 500 μg anti-LFA-1 ( clone M17/4 ) and 500 μg anti-VLA-4 ( clone PS/2 ) i . v . at the indicated time points . For cognate peptide-MHC blocking assays , mice were treated with 680 μg anti-Kb-SIINFEKL ( clone 25-D1 . 16 ) or mIgG1 isotype control ( clone MOPC-21 ) antibodies i . v . at the indicated time points . Mice were seeded i . v . with 104 mCerulean+ OT-I or YFP+ P14 CD8+ T cells purified from the splenocytes of naïve transgenic mice using a CD8 negative selection kit ( Stem Cell Technologies ) . To compare activated PbA-specific and non-specific CD8+ T cell responses in the same mice , activated YFP+ P14s were first purified from the splenocytes of previously seeded mice using a CD8 positive selection kit ( Stem Cell Technologies ) , 8 days following infection with LCMV . Symptomatic PbA-infected mice previously seeded with mCer OT-1 cells were then i . v . injected with 5x106 activated YFP-P14s i . v . Mice were anesthetized and thin skull windows were prepared as previously described [42 , 43] . Mice were injected with 5 μl Qdot655 ( BD ) or 50 μg Evans blue where indicated to visualize blood vessels . 4D datasets were acquired using an SP5 two-photon microscope ( Leica ) equipped with two Mai Tai HP DeepSee lasers ( SpectraPhysics ) , an 8 , 000-Hz resonant scanner , a 20×/1 . 0 NA objective , an NDD4 detector array , and a custom-environment chamber . Simultaneous excitation and detection of multiple fluorophores was achieved using custom dichroic mirrors ( Semrock ) and by tuning one laser to 905 nm and the other to 990 nm . All imaging studies focused on the meninges and superficial neocortex . Imaging data were processed with Imaris 7 . 6 . 4 software . A surface was created for each CNS blood vessel and PbA-specific CD8+ T cells located within that surface ( vessel lumen ) were quantified , whereas T cells residing outside the surface ( perivascular or parenchymal ) were quantified separately . Mean track velocities ( μm/min ) , arrest coefficients ( proportion of time a cell spent arrested <2 μm/min ) , and arrest duration ( the total amount of time a cell spent arrested <2 μm/min ) were calculated for all individual luminal PbA-specific CD8+ T cell tracks using Imaris 7 . 6 . 4 and T Cell Analyzer software ( TCA 1 . 7 . 0; Strathclyde Institute of Pharmacy and Biomedical Sciences ) . All time lapses used for these analyses were at least 40 minutes in length . BM was harvested from femurs and tibias of Ly5 . 1 mice and 5x106 BM cells were i . v . injected into recipient mice following a lethal irradiation dose of 900 RAD . Mice received antibiotics in drinking water for 4 weeks following irradiation and were allowed 8 weeks to fully reconstitute bone marrow and donor peripheral cells . Statistical analyses for data were performed using a Student’s t test ( two groups ) or ANOVA ( more than two groups ) in Prism 6 ( GraphPad Software ) . Groups were considered statistically different at a p value of <0 . 05 . All data are displayed as the mean ± SD . This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the NINDS Animal Care and Use Committee ( Protocol Number: 1295–14 ) .
|
Cerebral malaria ( CM ) is a severe and potentially fatal complication of malaria in humans that results in swelling and bleeding within the brain . The mechanisms that cause this fatal disease in humans are not completely understood . We studied an animal model known as experimental cerebral malaria to learn more about the factors that drive this disease process . Using a technique referred to as intravital microscopy , we captured movies of immune cells operating in the living brain as the disease developed . At the peak of disease , we observed evidence of immune cells interacting with and aggregating along blood vessels throughout the brain . These interactions were directly associated vascular leakage . This caused the brain to swell , which gave rise to an unsustainable pressure that ultimately killed neurons responsible for heart and lung function . The fatal swelling was induced by immune cells ( referred to as T cells ) interacting with bits of parasite presented by blood vessels in the brain . Removal of this parasite presentation protected the mice from fatal disease . We also evaluated a straightforward therapy that involved intravenous administration of antibodies that interfered with T cell sticking to blood vessels . Our movies revealed that this therapeutic approach rapidly displaced T cells from the blood vessels in the brain and prevented fatal disease .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"cardiovascular",
"anatomy",
"nervous",
"system",
"immunology",
"tropical",
"diseases",
"brain",
"cell",
"processes",
"parasitic",
"diseases",
"neuronal",
"death",
"cytotoxic",
"t",
"cells",
"white",
"blood",
"cells",
"animal",
"cells",
"blood",
"vessels",
"t",
"cells",
"cerebral",
"malaria",
"brainstem",
"cell",
"biology",
"anatomy",
"central",
"nervous",
"system",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"malaria"
] |
2016
|
CD8+ T Cells Induce Fatal Brainstem Pathology during Cerebral Malaria via Luminal Antigen-Specific Engagement of Brain Vasculature
|
Chronic kidney disease ( CKD ) is a worldwide public health problem that is associated with substantial morbidity and mortality . To search for sequence variants that associate with CKD , we conducted a genome-wide association study ( GWAS ) that included a total of 3 , 203 Icelandic cases and 38 , 782 controls . We observed an association between CKD and a variant with 80% population frequency , rs4293393-T , positioned next to the UMOD gene ( GeneID: 7369 ) on chromosome 16p12 ( OR = 1 . 25 , P = 4 . 1×10−10 ) . This gene encodes uromodulin ( Tamm-Horsfall protein ) , the most abundant protein in mammalian urine . The variant also associates significantly with serum creatinine concentration ( SCr ) in Icelandic subjects ( N = 24 , 635 , P = 1 . 3×10−23 ) but not in a smaller set of healthy Dutch controls ( N = 1 , 819 , P = 0 . 39 ) . Our findings validate the association between the UMOD variant and both CKD and SCr recently discovered in a large GWAS . In the Icelandic dataset , we demonstrate that the effect on SCr increases substantially with both age ( P = 3 . 0×10−17 ) and number of comorbid diseases ( P = 0 . 008 ) . The association with CKD is also stronger in the older age groups . These results suggest that the UMOD variant may influence the adaptation of the kidney to age-related risk factors of kidney disease such as hypertension and diabetes . The variant also associates with serum urea ( P = 1 . 0×10−6 ) , uric acid ( P = 0 . 0064 ) , and suggestively with gout . In contrast to CKD , the UMOD variant confers protection against kidney stones when studied in 3 , 617 Icelandic and Dutch kidney stone cases and 43 , 201 controls ( OR = 0 . 88 , P = 5 . 7×10−5 ) .
Chronic kidney disease ( CKD ) is a common disorder that can progress to kidney failure and is associated with an increased risk of cardiovascular disease and mortality [1] , [2] . The cause of CKD is not always known and frequently appears multifactorial with hypertension ( HTN ) and diabetes mellitus ( DM ) being the most important causes [3]–[6] . Other causes include intrinsic kidney disorders , atherosclerosis and nephrotoxic drugs [7] , [8] . Studies also indicate a dramatic increase in the prevalence of CKD with advancing age [9] , [10] . With greater lifespan , the burden of CKD is thus steadily rising in the Western world [11] , resulting in a substantial impact on the health care system [12] . Previous studies have suggested a genetic contribution to the risk of kidney disease . Heritability estimates of serum creatinine ( SCr ) and estimated glomerular filtration rate based on SCr ( eGFRcrea ) , both common measures of kidney function , have been reported as 29% and 33% , respectively [13] . Recently , Köttgen et al . [14] published the first genome-wide association study ( GWAS ) on eGFRcrea , eGFR based on cystatin C ( eGFRcys ) , another measure of kidney function , and CKD , reporting significant association with eGFRcrea at three loci ( UMOD , SHROOM3 ( GeneID: 57619 ) and GATM-SPATA5L1 ( GeneIDs: 2628 and 79029 ) ) , with eGFRcys at two loci ( CST3-CST9 ( GeneIDs: 1471 and 128822 ) and STC1 ( GeneID: 6781 ) ) and with CKD at one locus ( UMOD ) [14] . With the aim of discovering sequence variants that associate with kidney function , we conducted a GWAS in a total of 3 , 203 Icelanders with CKD and 38 , 782 controls and in 24 , 635 Icelandic and 1 , 819 Dutch subjects with SCr information . We found a sequence variant at the UMOD locus that associates with both CKD and SCr at a genome wide-significant ( GWS ) level , providing an independent replication of the result by Köttgen et al [14] . We also show that this variant interacts with age-related increase in SCr levels with little or no effect on SCr before the age of 50 years , followed by a rapidly growing effect with increasing age . We demonstrate that this variant associates significantly with serum urea , uric acid and suggestively with gout . In contrast to the deleterious effect on kidney function , the variant confers protection against kidney stone disease .
A GWAS of 2 . 5 million SNPs , either directly genotyped ( Illumina HumanHap300 or HumanHapCNV370 bead chips ) or imputed based on the HapMap CEU samples [15] , was performed on a sample set of 2 , 903 Icelanders with CKD ( see Materials and Methods for sample set description ) and 35 , 818 controls and also on 22 , 256 Icelandic subjects with SCr information ( See QQ-plots in Figure S1 and Figure S2 ) . The Icelandic SCr measurements came from two laboratories; the Laboratory in Mjodd , a private outpatient laboratory , and the Clinical Biochemistry Laboratory of Landspitali University Hospital ( LUH ) , serving both inpatients and outpatients . These subjects had 5 . 2 SCr measurements on average ( geometric mean ) and we used the median SCr value for each individual in the subsequent analysis . The SCr values from the two Icelandic laboratories showed similar dependence on age and sex but there was clearly a trend towards higher SCr in the hospital laboratory compared with the outpatient laboratory ( Figure S3 ) . The GWAS on CKD and SCr both yielded several SNPs in high linkage disequilbrium ( LD ) on chromosome 16p12 with GWS ( P<5×10−8 ) association to increased risk of CKD and elevated SCr . For both phenotypes , this signal is tagged by rs4293393-T ( Table 1 and Table 2 ) . For CKD , the odds ratio ( OR ) conferred by rs4293393-T was 1 . 25 ( 95% CI = 1 . 16–1 . 34 ) with a corresponding P value of 6 . 2×10−9 . In an attempt to replicate this finding , rs4293393 was typed in additional 300 Icelandic subjects with CKD and 2 , 964 controls . The association was nominally significant in the replication sample set and the effect size consistent with the initial observation ( Table 1 ) . The combined OR for rs4293393-T in the two Icelandic CKD sample sets was 1 . 25 ( 95% CI = 1 . 17–1 . 35 ) and P = 4 . 1×10−10 . The association between SCr and rs4293393-T on 16p12 was very strong with an effect of 1 . 86 µmol/L per allele carried and P = 6 . 7×10−20 ( Table 2 ) . To follow up on these results , we genotyped rs4293393 in 2 , 379 additional Icelanders with SCr information , significantly replicating the initially observed effect ( P = 1 . 4×10−5 , Table 2 ) . Analysis of the combined Icelandic datasets showed a strong GWS association between rs4293393-T and elevated SCr ( effect = 1 . 93 µmol/L per allele , 95% CI = 1 . 55–2 . 31 µmol/L; P = 1 . 3×10−23 ) . Our findings provide an independent replication of the recently reported results by Köttgen et al [14] of an association of this 16p12 locus with CKD and eGFRcrea . The strongest SNP associations outside the UMOD region on chromosome 16p12 are shown in Table S1 ( for CKD ) and Table S2 ( for SCr ) , respectively . For further assessment , we tested rs4293393 in 1 , 819 Dutch subjects with SCr information . These were healthy population-based controls ( see Materials and Methods for sample set description ) with SCr values substantially different from the Icelandic measurements , generally showing lower values and much less variability ( Figure S3 ) . Interestingly , no association was observed in the 1 , 819 healthy Dutch subjects ( effect = 0 . 38 µmol/L , 95%CI = −0 . 48–1 . 25 µmol/L; P = 0 . 39 ) ( Table 2 ) . Significant heterogeneity was observed between the SCr association results for the Icelandic and Dutch populations ( I2 = 90 . 4% , P = 0 . 0013 ) . The SNP rs4293393 is located 550 basepairs upstream of UMOD , encoding uromodulin , also known as the Tamm-Horsfall protein ( Figure 1 ) . The protein is a glycosylphosphatidylinositol ( GPI ) -anchored glycoprotein , exclusively expressed in the thick ascending loop of Henle [16] and distal convoluted tubule [17] of the mammalian kidney . It is the most abundant protein in urine of healthy individuals , where it is present in a highly aggregated state [18] , [19] . While the exact physiological function of uromodulin remains to be elucidated , it has the capacity to bind to a variety of ligands . It has been reported to prevent bacteria from adhering to human kidney cells [20] and to inhibit calcium oxalate crystallization [21] . It may also have a role in ion transport and immunological processes [22] , [23] . UMOD knockout mice have been shown to have decreased creatinine clearance [24] and predilection for both urinary tract infections [25] and calcium oxalate stone formation [26] . The rs4293393 variant is in perfect LD in the HapMap CEU samples [15] with a synonymous SNP in UMOD , rs13335818 , coding for V264V ( D′ = 1 . 0 , r2 = 1 . 0 ) . The same perfect correlation between rs4293393 and rs13335818 was observed in a set of 3 , 364 Icelanders ( D′ = 1 . 0 , r2 = 1 . 0 ) . Both rs4293393 and rs13335818 are also in perfect correlation with rs12917707 ( D′ = 1 . 0 , r2 = 1 . 0 for both markers in the HapMap CEU samples [15] ) near the UMOD , the variant reported by Köttgen et al [14] to associate with both CKD and eGFRcrea with similar effect , indicating that these SNPs are tagging the same signal . As rs4293393 is on the Illumina 300/370K chips we used for direct genotyping , we refer to rs4293393 in the remainder of this article . Given that SCr varies substantially with both age and sex , we tested for an interaction between the effect of rs4293393-T and effects of age and sex on SCr . No interaction was found between the UMOD variant and sex ( P = 0 . 41 ) . In contrast , a strong interaction was observed between the UMOD variant and age in the Icelandic sample set ( P = 3 . 0×1017 ) . On average , SCr increased by an additional 0 . 09 µmol/L per year per allele of rs4293393-T ( 95% CI = 0 . 07–0 . 11 ) . In order to visualize this interaction , we stratified our Icelandic samples based on age and sex and tested for association within each stratum ( Figure 2A ) . Interestingly , rs4293393-T has little or no effect on SCr before the age of 50 years , but thereafter the effect increases rapidly with advancing age , especially around 70 years . Thus , the variant does not affect SCr in young individuals but rather how SCr increases with age . We note that due to the relatively short time span in which the SCr data were collected there is an inherent confounding between age and generation in our study , which will require further investigation to resolve . Similar interaction between the UMOD variant , age and CKD was also observed when the association analysis for CKD was stratified by year of birth used here as a proxy for age of onset ( Table 1 ) . Although it is well known that kidney function declines with age , GFR has been shown to decrease more slowly with senescence in healthy individuals than previously thought [3]–[5] , [11] . Comorbid conditions that increase in frequency with aging , including HTN , DM , atherosclerosis and heart failure are , however , increasingly recognized as important contributors of age-related decline in kidney function [3] , [4] , [6]–[8] . We thus proceeded to investigate whether the age effect observed in carriers of the UMOD variant is influenced by interaction with age-related risk factors for decline in kidney function . As HTN , type 2 DM and atherosclerosis are all well recognized age-dependent risk factors for CKD [3] , [4] , [6]–[8] , the association analysis was repeated after stratifying the SCr data based on these conditions . Incomplete information regarding history of HTN ( 5 , 705 cases ) , type 2 DM ( 1 , 422 cases ) and myocardial infarction ( MI , 2 , 551 cases ) was available for the Icelandic SCr sample set . In parallel with previous studies , the rate of increase in SCr with age was significantly higher in individuals with HTN than in individuals without this diagnosis ( effect = 0 . 23 µmol/L/year , 95% CI = 0 . 19–0 . 26 µmol/L/year; P = 2 . 9×10−35 ) . Similar results were obtained for type 2 DM ( effect = 0 . 26 µmol/L/year , 95% CI = 0 . 19–0 . 34 µmol/L/year; P = 1 . 1×10−11 ) and MI ( effect = 0 . 36 µmol/L/year , 95% CI = 0 . 30–0 . 42 µmol/L/year; P = 1 . 4×10−32 ) as well as the number of comorbid conditions ( Figure 2B ) . We also found that the effect of rs4293393-T on SCr increases with the number of comorbid conditions present ( Figure 2C ) . To further assess the relationship between genotype , age and risk factors for reduced kidney function , we investigated the effect of the rs4293393-T allele count on the increase in SCr with age stratifying on HTN and type 2 DM . A trend was observed for a higher rate of increase in SCr with age and rs4293393-T allele count in individuals with HTN compared to those without a diagnosis of HTN ( P = 0 . 077 ) as well as in those with type 2 DM compared to those without ( P = 0 . 063 ) . In other words , the age-related increase in SCr levels appears to be greater in rs4293393-T carriers that have either HTN or type 2 DM than in carriers who do not have these risk factors . However , an age effect was still observed after accounting for these age-related risk factors . Furthermore , we also observed a significantly higher rate of SCr increase with age and rs4293393-T allele count stratifying on the number of comorbid conditions present ( P = 0 . 0080 ) ( Figure 2D ) . To determine whether rs4293393-T influenced kidney function by directly affecting known risk factors , we tested the association of rs4293393-T in well powered Icelandic case-control sets of HTN , MI , stroke , and type 2 DM ( Table S3 ) . A weak nominally significant association of rs4293393-T with increased risk of HTN was observed ( OR = 1 . 07 , 95% CI = 1 . 01–1 . 12; P = 0 . 014 ) , but not with the other diseases tested . These data demonstrate that the effect of rs4293393-T on kidney function is not mediated through increased risk of these comorbid conditions , but rather suggest that the variant may affect the vulnerability of the kidney to these risk factors . These findings , demonstrating not only the effect of age on UMOD-associated increase in SCr but also the effect of age-related comorbid conditions , may explain why no association was observed between rs4293393-T and kidney function in the Dutch sample set of healthy population-based subjects with much lower SCr values and of much less variability than observed in the Icelandic samples ( Figure S3 ) . Urea is another small solute that is commonly used to assess renal function together with SCr . The correlation between SCr and serum urea in our data was 58% . We tested for association between rs4293393-T and serum urea in an Icelandic sample set that had urea measurements performed at the Laboratory in Mjodd ( N = 4 , 084 ) and found significant association with increased serum urea concentration ( effect = 0 . 36 mg/dL , 95% CI = 0 . 23–0 . 50 mg/dL; P = 1 . 0×10−6 ) . In humans , rare mutations in the UMOD gene that cause accumulation of abnormal uromodulin in the nephron and reduced urinary excretion of the normal protein [27] have been associated with two autosomal dominant kidney diseases with overlapping clinical features , medullary cystic kidney disease and familial juvenile hyperuricemic nephropathy [28] . These disorders are characterized by hyperuricemia , gout and progressive renal failure due to tubulointerstitial nephropathy . Given the link between UMOD and hyperuricemia , we tested rs4293393-T in Icelandic subjects with serum uric acid values from the Laboratory in Mjodd ( N = 6 , 583 ) and observed significant association ( effect = 6 . 1 , 95% CI = 1 . 7–10 . 4; P = 0 . 0064 ) . We then tested for association with gout in a set of 377 Icelandic cases and 39 , 916 controls ( see Materials and Methods for sample set description ) and found a suggestive association ( OR = 1 . 17 , 95% CI = 0 . 97–1 . 41; P = 0 . 097 ) . These data contrast the work of Köttgen et al [14] that neither detected association with serum uric acid nor gout . As uromodulin is thought to prevent the formation of calcium-containing kidney stones [21] , we tested rs4293393 in a sample set of 1 , 689 Icelandic patients with radiopaque kidney stones and 37 , 076 Icelandic population controls . We observed a significant association between rs4293393-T and reduced risk of kidney stones ( OR = 0 . 88 , 95% CI = 0 . 81–0 . 96; P = 0 . 0053 ) . In an attempt to replicate this observation , we genotyped rs4293393 in two additional sample sets of European ancestry , one from Iceland ( 1 , 271 cases and 3 , 177 controls ) and the other from the Netherlands ( 701 cases and 2 , 948 controls ) ( Table 3 ) . The effect size in both replication samples is consistent with the initial observation and the association is significant in the combined replication samples ( OR = 0 . 89 , 95% CI = 0 . 81–0 . 97; P = 0 . 0089 ) . There was no correlation between the effect size and year of birth of the kidney stone patients ( Table S4 ) . Köttgen et al [14] reported on variants at additional loci with GWS association to eGFRcrea ( SHROOM3 and GATM-SPATA5L1 ) or eGFRcys ( STC1 and CST3-CST9 ) . We tested these variants in our Icelandic datasets , including a small sample set with cystatin C measurements ( Table 4 ) . The association with SCr replicated for both the SHROOM3 and GATM-SPATA5L1 SNPs ( P = 0 . 00057 and P = 0 . 0067 , respectively ) and the association with cystatin C replicated for the CST3-CST9 SNP but not the STC1 SNP ( P = 0 . 00037 and P = 0 . 73 , respectively ) . It should be noted , however , that the Icelandic cystatin C sample set is very small ( N = 284 ) and possibly underpowered to replicate the reported association with the STC1 SNP . The STC1 SNP did show association with SCr in our dataset ( P = 1 . 6·10−6 ) as was observed in the analysis by Köttgen et al [14] . The SHROOM3 and GATM-SPATA5L1 SNPs showed suggestive association with CKD in the analysis by Köttgen et al [14] and our data support this association but do not constitute a conclusive replication . In contrast to the UMOD variant , we did not observe an interaction between the variants at these other loci and age . Furthermore , none of the SNPs did associate with kidney stones in Iceland ( data not shown ) . Finally , Köttgen et al [14] reported suggestive association between eGFRcrea and variants at JAG1 ( GeneID: 182 ) ; we did not replicate this finding in our SCr scan ( for rs6040055-T: effect = −0 . 27 , 95% CI = −0 . 60−0 . 05 ) , P = 0 . 17 ) . In summary , we describe sequence variants next to and in UMOD that associate with increased risk of CKD and elevated SCr but confer protection against kidney stones . We also demonstrate an interaction between these variants and both age and comorbid conditions that are related to decline in kidney function . Our observations indicate that UMOD is important for maintaining kidney function with advancing age and exposure to risk factors that are associated with aging , such as HTN , type 2 DM and cardiovascular disease .
Landspitali University Hospital ( LUH ) is a regional hospital for the greater Reykjavík area and a tertiary referral center for the entire Icelandic nation . The population of Iceland is comprised of 330 , 000 inhabitants of whom approximately 200 , 000 reside in the greater Reykjavik area . The nation's only nephrology clinic is located at LUH and all laboratory tests for the primary care clinics of the greater Reykjavik area are performed in the hospital's laboratories . We obtained results of all SCr measurements performed during the period 2003 to 2008 from the computerized database of the Clinical Laboratories at LUH and used the SCr values to identify those with chronic kidney disease ( CKD ) based on calculation of the estimated glomerular filtration rate ( eGFR ) by the original 4-variable Modification of Diet in Renal Disease ( MDRD ) study equation . We classified those with eGFR<60 ml/min/1 . 73 m2 as having CKD . All individuals with acute kidney injury and those who had eGFR<60 ml/min/1 . 73 m2 of less than 3 months duration were excluded from the CKD sample set . The study included CKD patients that had donated blood through various genetic programs at deCODE genetics . Biochemical measurements including SCr , serum urea , serum uric acid and serum cystatin C were available from two laboratories , the Laboratory in Mjodd , Reykjavik , Iceland , a private outpatient laboratory , and the Clinical Biochemistry Laboratory of LUH , serving both inpatients and outpatients . The main referral center for the Laboratory in Mjodd is a multispecialty medical clinic in Reykjavik ( Laeknasetrid ) . The laboratory tests were done in the years 1997–2008 at the Laboratory in Mjodd and in the years 2003–2008 at LUH . The Icelandic SCr measurements came from both laboratories , the Laboratory in Mjodd ( N = 10 , 260 ) and LUH ( N = 22 , 898 , of whom 8 , 523 also had measurements from the Laboratory in Mjodd ) . At the LUH , the same enzymatic method was used for measurement of SCr during the study period ( Vitros 950 Autoanalyzer , Ortho Clinical Diagnostics , Rochester , MN , USA ) , whereas at the Laboratory in Mjodd , SCr measurements were performed by modified kinetic Jaffe rection assays until May 2007 when an enzymatic method was introduced . The Icelandic kidney stone cases consisted of patients with confirmed radiopaque kidney stones from the Icelandic Kidney Stone Disease Registry at LUH . The study included kidney stone patients that had donated blood through various genetic programs at deCODE genetics . The coronary artery disease [29] , stroke [30] and type 2 DM [31] , [32] patient groups from Iceland have been described previously . The HTN sample set includes individuals who have been recruited into various genetic programs at deCODE and have: ( 1 ) self-reported HTN; ( 2 ) received the diagnosis of HTN at discharge from the LUH; or ( 3 ) have attended the Hypertension Clinic at LUH . The gout sample set includes subjects who were recruited into various genetic programs at deCODE and reported the use of either one of two specific anti-gout medications , allopurinol or colchicine . The Icelandic controls used in the case-control genome-wide association studies ( GWAS ) and replication studies were selected among individuals who had participated in the various genetic programs at deCODE genetics; tremor , preeclampsia , endometriosis , psoriasis , type 2 DM , Alzheimer's disease , osteoarthritis , schizophrenia , peripheral artery disease , abdominal aortic aneurysm , chronic obstructive pulmonary disease , stroke , osteoporosis , coronary artery disease , HTN , asthma , Parkinson's disease , sleep apnea , age-related macular degeneration , polycystic ovary syndrome , rheumatoid arthritis , lung cancer , longevity , benign prostatic hyperplasia , enuresis , migraine , glaucoma , attention deficit hyperactivity disorder , prostate cancer , infectious diseases , anxiety , expression studies , autism , dyslexia , melanoma , colorectal cancer , deep vein thrombosis , restless leg syndrome , studies on addiction and population controls . Individuals who reported a history of the trait being tested ( e . g . CKD ) were excluded from the control set . Some of the controls participated in more than one genetic program in which case their genotypes are only included once . The study was approved by the Icelandic Data Protection Authority and the National Bioethics Committee . All patients signed informed consent and donated blood samples . Personal identities of the patients and biological samples were encrypted by a third party system provided by the Icelandic Data Protection Authority . All samples with SCr measurements came from the Nijmegen Biomedical Study . The details of this study have been reported previously [33] . Briefly , this is a population-based survey conducted by the Department of Epidemiology and Biostatistics and the Department of Clinical Chemistry of the Radboud University Nijmegen Medical Center ( RUNMC ) , in which 9 , 371 individuals participated from a total of 22 , 500 age- and sex-stratified , randomly selected inhabitants of Nijmegen . The subjects were invited to participate in a study on gene-environment interactions in complex diseases . All participants filled out a questionnaire on lifestyle and medical history at baseline and 6500 of them donated blood samples for DNA isolation and biochemical studies . A fraction of the study participants were previously genotyped with the Illumina HumanHap300 or HumanHapCNV370 bead chips; these were selected to serve as controls in GWAS on prostate and breast cancer and were selected primarily based on age . A total of 1 , 819 individuals had both serum creatinine measurements and genome-wide SNP data available for analysis in this study . The Dutch patients with kidney stones were recruited from two sources: The outpatient clinics of the RUNMC and The Nijmegen Biomedical Study . All patients who present to the outpatient clinics of the RUNMC are invited to participate in a study on the effects of genes and lifestyle on the development of urological diseases . In case of consent , the patients fill out a questionnaire on lifestyle and donate a blood sample for DNA isolation . The controls for the analysis of kidney stone disease were also taken from the biobank of the Nijmegen Biomedical Study . All patients and controls were of self-reported European descent and were fully informed about the goals and the procedures of these studies . The study protocols for the recruitment of patients from outpatient clinics and the recruitment of participants to the Nijmegen Biomedical Study were approved by the RUNMC Institutional Review Board . All study subjects gave written informed consent . All Icelandic case and control samples were assayed with the Illumina HumanHap300 or HumanHapCNV370 bead chips ( Illumina , SanDiego , CA , USA ) , containing 317 , 503 and 370 , 404 haplotype tagging SNPs derived from phase I of the International HapMap project , respectively . Only SNPs present on both chips were included in the analysis and SNPs were excluded if they had: ( a ) yield lower than 95% in cases or controls; ( b ) minor allele frequency less than 1% in the population; or ( c ) showed significant deviation from Hardy-Weinberg equilibrium in the controls ( P<0 . 001 ) . Any samples with a call rate below 98% were excluded from the analysis . The final analysis included 302 , 379 SNPs . The genome-wide association scan was based on expected allele counts obtained with the IMPUTE software [34] , using the HapMap CEU samples as a training set [15] . The test for association was then performed using the expected allele counts as covariates . The imputation information was estimated by the ratio of the observed likelihood of allele counts and the likelihood of allele counts assuming perfect information under the assumption of Hardy-Weinberg equilibrium . The estimated information for the four SNPs imputed in Table 4 was high in all instances ( >0 . 96 ) . Single SNP genotyping of all samples was carried out at deCODE genetics in Reykjavik , Iceland , applying the same platform to all populations studied . All single SNP genotyping was carried out using the Centaurus ( Nanogen ) platform [35] . The quality of each Centaurus SNP assay was evaluated by genotyping each assay on the CEU samples and comparing the results with the HapMap data . All assays had a mismatch rate <0 . 5% . Additionally , all markers were re-genotyped on more than 10% of samples typed with the Illumina platform resulting in an observed mismatch in less than <0 . 5% of samples . For case-control association analysis , e . g . for CKD and kidney stones , we utilized a standard likelihood ratio statistic , implemented in the NEMO software [32] to calculate two-sided P values and odds ratios ( ORs ) for each individual allele , assuming a multiplicative model for risk , i . e . that the risk of the two alleles carried by a person multiplies [36] . Allelic frequencies , rather than carrier frequencies , are presented for the markers and P values are given after adjustment for the relatedness of the subjects . When estimating genotype specific OR , genotype frequencies in the population were estimated assuming Hardy-Weinberg equilibrium . Results from multiple case-control groups were combined using a Mantel-Haenszel model [37] in which the groups were allowed to have different population frequencies for alleles , haplotypes and genotypes but were assumed to have common relative risks . For the quantitative trait association analysis , e . g . for SCr and cystatin C , we utilized a robust linear regression based on an M estimator [38] as implemented in the rlm function of the R software package [39] . An additive model for SNP effects was assumed in all instances . All associations with quantitative traits were performed adjusting for age and sex . Interaction effects were tested by assuming all main effects and lower order interaction effects were present under the null model but not the interaction effect , resulting in a one degree of freedom model . For example , when testing for an interaction effect on SCr between age and the rs4293393-T allele count , the null model included as covariates age , sex and the rs4293393-T allele count . The alternative model included all these covariates as well as the product of the interaction of age and the rs4293393-T allele counts . Similarly , when testing for the interaction between age , the number of comorbid conditions and the rs4293393-T allele count , the null model included as covariates age , sex , rs4293393-T allele count , the product of interaction of age and rs4293393-T allele count , the product of interaction of the number of comorbid conditions and rs4293393-T allele count , and the product of interaction of age and the number of comorbid conditions and the alternative model added the product of interaction of age , the rs4293393-T allele count and the number of comorbid conditions . In the instances when an interaction effect was estimated , the main effect estimates and P values shown were obtained from fitting the appropriate model without an interaction effect . Some of the individuals in the Icelandic patient and control groups are related to each other , causing the chi-square test statistic to have a mean >1 and median >0 . 675 . We estimated the inflation factor for the genome-wide association by calculating the average of the 302 , 379 chi-square statistics , which was a method of genomic control [40] to adjust for both relatedness and potential population stratification . The inflation factor was estimated as 1 . 15 for CKD and 1 . 22 for SCr and all the results presented from association with these traits were adjusted based on these inflation factors .
|
Chronic kidney disease ( CKD ) is a common condition that is associated with substantial morbidity and mortality and has been recognized as a major public health problem worldwide . Common causes of CKD include hypertension , diabetes , and inflammatory disorders . Previous studies have shown a significant genetic contribution to kidney disease and a recent genome-wide association study yielded a variant in the UMOD gene that affects the risk of CKD . Here , we replicate the association between UMOD and CKD in an independent analysis . We also demonstrate for the first time an interaction between the UMOD variant and age that suggests that this variant may adversely affect the aging kidney and its adaptation to age-related risk factors of kidney disease , such as hypertension and diabetes . Furthermore , we show that the UMOD variant that affects risk of CKD also provides protection against kidney stone disease .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/population",
"genetics"
] |
2010
|
Association of Variants at UMOD with Chronic Kidney Disease and Kidney Stones—Role of Age and Comorbid Diseases
|
Host protection from fungal infection is thought to ensue in part from the activity of Syk-coupled C-type lectin receptors and MyD88-coupled toll-like receptors in myeloid cells , including neutrophils , macrophages and dendritic cells ( DCs ) . Given the multitude of cell types and receptors involved , elimination of a single pathway for fungal recognition in a cell type such as DCs , primarily known for their ability to prime T cell responses , would be expected to have little effect on innate resistance to fungal infection . Here we report that this is surprisingly not the case and that selective loss of Syk but not MyD88 in DCs abrogates innate resistance to acute systemic Candida albicans infection in mice . We show that Syk expression by DCs is necessary for IL-23p19 production in response to C . albicans , which is essential to transiently induce GM-CSF secretion by NK cells that are recruited to the site of fungal replication . NK cell-derived-GM-CSF in turn sustains the anti-microbial activity of neutrophils , the main fungicidal effectors . Thus , the activity of a single kinase in a single myeloid cell type orchestrates a complex series of molecular and cellular events that underlies innate resistance to fungal sepsis .
Candida albicans is the most prevalent fungal pathogen in humans causing local infections of skin , nails , oral cavity and genital tract [1] . In some instances , Candida can spread systemically via the bloodstream and lodge in the kidneys , which then act as the major site of fungal replication [2] . Despite the availability of several anti-fungal drugs , invasive candidiasis still has a high mortality rate ranging from 45 to 75% [3] , highlighting the need to further understand host-pathogen interactions and mechanisms of immune resistance to fungal spread . Despite its potential pathogenicity , C . albicans generally behaves as an innocuous commensal in immunocompetent individuals because it triggers host defense pathways that keep the organism in check . Host protection from infection ultimately depends on recognition of Candida by pattern recognition receptors ( PRRs ) and their associated signaling pathways that initiate immunity . Many PRRs recognizing Candida are expressed by myeloid cells and belong either to the Toll-like receptor ( TLR ) or the C-type lectin receptor ( CLR ) families . A role for TLRs in anti-fungal defense was first suggested by studies in mice deficient for the TLR adaptor MyD88 , which are highly susceptible to systemic candidiasis [4] , [5] . However , MyD88 additionally transduces signals from IL-1 and IL-18 receptors , which can impact innate anti-fungal immunity [4] , [6]–[10] , and human deficiency in MyD88 does not lead to loss of resistance to fungal organisms [11] . Therefore , the role of TLRs in protection from Candida infection remains unresolved [12]–[14] . In contrast , the role of CLRs in anti-fungal defense is becoming increasingly well-established . CLRs involved in fungal recognition include Dectin-1 , Dectin-2 , mannose receptor , MCL and Mincle , and mice or humans deficient in some of these receptors display enhanced susceptibility to candidiasis [15]–[19] . Dectin-1 , -2 and Mincle all signal via tyrosine-based motifs that recruit the spleen tyrosine kinase Syk [20]–[23] , leading to an NF-κB-dependent transcriptional program via CARD9 [24] . CLR/Syk signaling additionally promotes activation of NFAT , MAP kinase and PI3 kinase ( PI3K ) pathways [25] , [26] and can also lead to production of reactive oxygen species ( ROS ) and activation of inflammasomes [6] . Notably , Syk- or CARD9-deficient dendritic cells ( DCs ) fail to produce certain cytokines in response to Candida and fungal cell wall components [6] , [21] , [27] and CARD9-deficient mice are highly susceptible to systemic infection with C . albicans [24] . Likewise , human deficiency in CARD9 results in severe forms of superficial as well as invasive candidiasis [28] , [29] . Thus , Syk-dependent signaling by CLRs appears an important and non-redundant pathway for anti-fungal responses . It is presently unclear whether this reflects a dominant role for Syk in a given myeloid cell type or the additive effects of PRR signaling in multiple phagocytes . PRR signaling can trigger both innate and adaptive immune mechanisms . Adaptive immunity is initiated by DCs and important for defense against mucocutaneous candidiasis [30] but does not play a prominent role in combatting disseminated C . albicans infection [31] . Instead , innate immunity acts as the major barrier to systemic Candida spread . Indeed , the candidacidal activity of neutrophils is the key mediator of immunity to systemic candidiasis and neutropenia is a major risk factor for invasive Candida disease [31] , [32] . Macrophages and inflammatory monocytes also coordinate aspects of resistance to systemic Candida spread [33]–[36] while , recently , NK cells have been shown to be crucial for promoting neutrophil candidacidal activity during experimental systemic candidiasis in mice [37] . The collaborative impact of NK cells and neutrophils is also apparent in a model of invasive Aspergillus fumigatus where co-depletion greatly decreases survival compared to neutrophil depletion alone [38] . Thus , neutrophils , monocytes/macrophages and NK cells all mediate innate resistance to fungal hematogenous spread although whether all these cell types act individually or coordinately to provide host protection and which signals are involved in regulating their activity remains unknown . Experimental systemic candidiasis in mice mimics human candidemia in that fungal replication occurs primarily in the kidneys and resistance is mediated by neutrophils independently of T and B cells [39] . In this work , we report that the coordination of innate immunity to systemic C . albicans infection in mice is critically dependent on Syk and not MyD88 expression in CD11c+ cells . We identify the CD11c+ cells in question as DCs by ontogenetic criteria , thereby ascribing DCs a key role in innate immunity that is much less appreciated than their function in adaptive immunity . We show that this is because in the absence of Syk signaling , DCs do not produce IL-23p19 in response to C . albicans , which is necessary for fungus-driven production of GM-CSF by NK cells in the kidney . The loss of GM-CSF-producing NK cells leads to a failure to sustain the candidacidal ability of neutrophils and results in high kidney fungal burden and decreased survival to infection . Thus , effective immunity to systemic C . albicans infection involves a precise chain of sequential cellular activation events that is initiated by Syk-dependent signaling in DCs , depends on NK cells and culminates in neutrophil fungicidal activity .
To assess the relative contribution of Syk- and MyD88-dependent pathways in CD11c+ mononuclear phagocytes ( predominantly DCs ) to immunity during systemic C . albicans infection , we crossed Sykfl/fl [40] or MyD88fl/fl [41] strains to CD11c-Cre [42] mice to generate CD11cΔSyk and CD11cΔMyD88 lines , respectively . When CD11c+ MHC II+ cells ( henceforth called DCs – see below ) in the spleens and kidneys of CD11cΔSyk mice were compared to those in littermate controls ( CD11cCre− Sykfl/fl ) , DCs from CD11cΔSyk mice displayed a marked reduction in Syk mRNA , with a PCR signal barely above that obtained for T cells , which do not express the kinase ( Figure S1A ) . No reduction in Syk mRNA was seen in B cells ( Figure S1A ) and measurement of Syk protein levels by intracellular staining ( Figure S1B and S1C ) or Western blotting ( data not shown ) confirmed that the kinase was specifically deleted in CD11c+ cells . Importantly , levels of Syk were not reduced in neutrophils , indicating restriction of the deletion to the mononuclear phagocyte system ( Figure S1B and S1C ) . Likewise , in CD11cΔMyD88 mice , a reduction in MyD88 staining was observed specifically in CD11c+ MHC-II+ DCs and not in neutrophils or other leukocytes ( Figure S1D and data not shown ) , as reported [41] . Remarkably , CD11cΔSyk mice succumbed rapidly to systemic infection with 5×104 CFU of C . albicans when compared to controls , of which the majority survived for up to 3 weeks ( Figure 1A and data not shown ) . At higher inoculum doses , the mortality of control animals increased [43] , although , importantly , the difference in susceptibility between control and CD11cΔSyk mice was maintained ( data not shown ) . The kidneys of infected CD11cΔSyk mice showed a large number of fungal abscesses with prominent hyphae ( revealed by periodic acid Schiff ( PAS ) staining ) heavily surrounded by leukocytes ( shown by hematoxylin and eosin ( H&E ) staining ) ( Figure 1B ) . Consistent with these observations , fungal burden in the kidneys of CD11cΔSyk mice was around 100-fold higher than in control littermates ( Figure 1C ) . Reflecting the massive candidemia , fungus could additionally be recovered from spleen and liver of CD11cΔSyk mice , which additionally displayed liver lipolysis ( data not shown ) . In contrast , selective ablation of MyD88 in CD11c+ cells in CD11cΔMyD88 mice did not result in enhanced susceptibility to systemic Candida infection even though MyD88-deficient mice ( lacking MyD88 in all cell types ) were extremely susceptible ( Figure 1C ) . Thus , ablation of Syk but not MyD88 in CD11c-expressing cells greatly compromises innate resistance to systemic C . albicans infection in mice and leads to death from fulminant candidiasis . CD11c is not an exclusive marker of DCs . To narrow down the CD11c+ cell type required to express Syk in this model , we made use of recently-developed Clec9a-Cre mice in which Cre activity is restricted to cells derived from non-monocytic conventional DC precursors ( CDP ) [44] . Notably , despite the incomplete penetrance of Cre-mediated recombination in such precursors [44] , Clec9aΔSyk were nearly as susceptible as CD11cΔSyk mice to systemic Candida infection ( Figure 1D and data not shown ) . Because Clec9aΔSyk mice were used as homozygotes in these experiments and therefore lacked DNGR-1 expression [44] , we confirmed that DNGR-1 deficiency does not impact on susceptibility to candidiasis by assessing fungal burden in infected Clec9aegfp/egfp mice [45] ( Figure 1D ) . These data therefore suggest a key role for Syk signaling by conventional DCs . Consistent with that conclusion , all kidney DC sub-populations in CD11cΔSyk mice showed loss of Syk independently of infection ( Figure S1F ) . We conclude that Syk expression by DCs and , possibly , additional CD11c+ cells is a key determinant of innate immunity to systemic C . albicans infection . We assessed the composition of the leukocytic infiltrate in kidneys of infected CD11cΔSyk mice to determine if susceptibility to candidiasis correlated with loss of any particular CD11c+ phagocyte subset whose development or recruitment to the site of infection might depend on Syk . Interestingly , there was little change in the total size of the CD11c+ MHC-II+ DC compartment after infection ( Figure S2A ) , although its relative composition was altered: in kidneys from uninfected mice , CD11bINT F4/80+ DCs were prominent whilst in infected mice this population decreased in size and a CD11b+ F4/80INT population became more abundant ( Figure S2B and S2C ) . Importantly , despite infection-induced changes , there was no difference between control or CD11cΔSyk mice . For example , the total number of CD11c+ MHC-II+ cells was the same in the two strains and there was only a marginal difference in percentage ( Figure S2A ) . Similarly , the change in hierarchy of CD11c+ MHC-II+ populations following infection was largely equivalent between the strains ( Figure S2C ) . Small differences observed for the percentage but not total number of CD11c+ MHC-II+ CD11bINT F4/80+ cells ( Figure S2C ) might reflect changes in other leukocyte populations even though there was no obvious change in B , T or NK cells ( data not shown ) . In contrast to the CD11c+ mononuclear phagocyte pool , the numbers and percentages of CD11c− MHC-II− neutrophils increased greatly in the kidneys following infection in both strains ( Figure 2A ) , as expected [43] , [46] . However , CD11cΔSyk mice displayed higher levels of kidney neutrophilia , correlating with the greater fungal burden ( Figure 2A ) . Importantly , the phenotype of neutrophils in the kidneys but not the bone marrow of infected CD11cΔSyk mice was atypical , with a large fraction of the cells expressing only low levels of CD11b and CD11a ( Figure 2B , 2C and data not shown ) . The cells were also less granular but did not appear apoptotic or stain for active caspase 3 ( data not shown ) . We further assessed levels of myeloperoxidase ( MPO ) , a major constituent of azurophil granules necessary for generation of reactive oxygen species ( ROS ) , a key component of the neutrophil killing arsenal [47] . Kidney neutrophils from infected CD11cΔSyk mice had decreased levels of MPO when compared to controls ( Figure 2D ) . As these phenotypic differences might suggest impaired functionality [48] , we assessed the ability of neutrophils in CD11cΔSyk mice to kill C . albicans . We infected control and CD11cΔSyk mice with a strain of GFP-expressing C . albicans and measured GFP signal among kidney leukocyte populations . As expected , the majority of the GFP signal was found within neutrophils ( Figure 2E ) . However , a greater frequency of GFP+ neutrophils were present in CD11cΔSyk mice than in controls suggesting that kidney neutrophils from the former strain are impaired in their ability to destroy the fungus . To explicitly test this hypothesis , we sorted GFP+ and GFP− neutrophils from the kidney , lysed them and plated the lysates to determine C . albicans growth . This analysis confirmed that GFP+ neutrophils derived from CD11cΔSyk but not from control mice contained live C . albicans ( Figure 2F ) . We then evaluated if neutrophils could kill C . albicans ex vivo by sorting GFP− neutrophils and incubating them with live fungus . Consistent with their phenotypic differences , neutrophils from kidneys of infected CD11cΔSyk mice showed a decreased ability to kill C . albicans ex vivo when compared to their counterparts from control infected mice ( Figure 2G ) . In contrast , bone marrow neutrophils from either uninfected or infected CD11cΔSyk mice showed equivalent ex vivo candidacidal capacity ( Figure 2G and data not shown ) , which argues that neutrophil impairment occurs locally at the site of infection . We conclude that in CD11cΔSyk mice infected systemically with C . albicans there is undiminished recruitment of neutrophils to the kidney but the recruited cells display phenotypic alterations and are locally impaired in their candidacidal activity . We searched for local alterations in the inflammatory milieu of the kidney that might connect diminished neutrophil function to loss of Syk in DCs . Homogenates of kidneys from CD11cΔSyk mice showed higher levels of IL-6 , KC , MIP-1α , IL-1β , TNF , IL-1α and MCP-3 at days 1 or 2 post-infection when compared to control mice ( Figure S3A and S3B ) . This likely reflects the contribution of cell types other than DCs and macrophages as some of those cytokines are known to be produced in a Syk-dependent manner by mononuclear phagocytes in response to stimulation with Candida albicans [6] , [27] , [49] . Because the interpretation of the data was marred by the large differences in fungal burden between the two strains , a broader analysis was performed early after infection ( 16 h ) when fungal burdens are more equivalent . This analysis confirmed the discrepancy in IL-6 levels between infected mouse strains while revealing that many inflammatory mediators are in fact induced to similar levels in both control and CD11cΔSyk infected mice ( Figure S3C ) . A notable exception is GM-CSF , which was found to be selectively lost in the kidneys of infected CD11cΔSyk mice when compared to controls ( Figure 3A ) . GM-CSF has been reported to be important for enhancement of neutrophil maturation and neutrophil oxidative responses in both mice and man [50]–[52] . We therefore tested whether exogenous GM-CSF could decrease the susceptibility of CD11cΔSyk mice to infection . Recombinant GM-CSF administration resulted in a marked decrease in fungal burden in the kidneys of infected CD11cΔSyk mice ( Figure 3B ) . In contrast , the same GM-CSF treatment had only a modest effect in control mice and , importantly , did not impact the hyper-susceptibility of MyD88 KO mice , demonstrating selectivity ( Figure 3B ) . Altogether , these data suggest that the susceptibility of CD11cΔSyk mice to systemic Candida infection stems from a deficiency in GM-CSF production in the kidney , which results in failure to locally sustain neutrophil microbicidal activity . We have recently found that kidney-infiltrating NK cells serve as a non-redundant source of GM-CSF to promote the candidacidal activity of neutrophils during systemic Candida infection [37] . Therefore , we assessed recruitment of NK cells to the kidneys of infected control and CD11cΔSyk mice and measured their production of GM-CSF and IFN-γ . There was no difference between the two strains in the percentage or the total number of kidney NK cells either before or at different times after infection ( Figure 4A and data not shown ) . However , as early as 16 h after infection , a marked reduction was observed in CD11cΔSyk mice in both the percentage and number of GM-CSF-producing NK cells ( Figure 4B ) . In contrast , the percentage and number of NK cells positive for IFN-γ was equivalent between the two strains ( Figure 4B ) . The production of GM-CSF by NK cells was transient as levels of the cytokine diminished after 16 h in contrast to those of IFN-γ , which continued to greatly increase ( data not shown ) . A similar loss of GM-CSF+ but not of IFN-γ+ NK cells was seen in infected Clec9aΔSyk mice ( Figure 4C ) , strengthening the notion that the phenotype stems from loss of Syk in DCs . To determine if reduced GM-CSF production by NK cells and impaired neutrophil microbicidal activity are linked , we investigated if NK cells from control mice could restore the resistance of CD11cΔSyk mice to infection . Transfer of cell preparations enriched for NK cells from naïve control mice into CD11cΔSyk mice prior to infection had no impact on fungal burden ( Figure 5A ) . However , when the same preparations were isolated from infected control mice , fungal control was restored in subsequently infected CD11cΔSyk mice ( Figure 5A ) . The decrease in fungal burden conferred by adoptive NK cell transfer associated with an increased proportion of CD11bhi neutrophils ( Figure 5B ) and restoration of the ability of neutrophils to kill C . albicans ex vivo ( Figure 5C ) . Similar results were obtained upon transfer of pure NK cell populations that were isolated by cell sorting to exclude any confounding effect of contaminants ( Figure 5D , 5E ) . Protection was not observed when NK cells were isolated from mice infected 48 h earlier ( Figure 5D and data not shown ) , consistent with the notion that GM-CSF production is transient ( see above ) . Transfer of unsorted total spleen cells was also not protective even when the cells were taken at 16 h after infection ( Figure 5D , 5E and data not shown ) . We conclude that the susceptibility of CD11cΔSyk mice to C . albicans infection is due to a defect in GM-CSF-dependent NK cell “help” for neutrophils and can be prevented by transfer of appropriately-primed NK cells from recently-infected wild type mice but not by unprimed wild-type NK cells . Finally , we sought to identify the signal that links Syk signaling in DCs to the production of GM-CSF by kidney NK cells . IL-23p19 has been reported to be induced very rapidly yet transiently in the kidneys and lungs of Candida-infected mice [53] , [54] . In addition , IL-23p19 is important for early resistance to candidiasis [55] , [56] and can be synthesized by DCs in a Syk-dependent manner upon stimulation with CLR agonists [27] . Although IL-23R can be found on both NK cells and T cells [57] , we noted that NK cells but not T cells expressed IL-23R in the kidney ( Figure 6A ) . The expression of IL-23R on kidney NK cells was upregulated in control but not CD11cΔSyk mice following infection ( Figure 6B ) . In addition , we detected strong induction of IL-23p19 mRNA in kidney CD11c+ MHC-II+ DCs from control mice but , importantly , not from CD11cΔSyk mice early after infection with C . albicans ( Figure 6C ) . The increase in the proportion of IL-23R+ cells ( Figure 6B ) may therefore be a consequence of positive feedback signaling of IL-23R in response to ligand [57] , [58] . To test the significance of this observation , we infected IL-23p19 KO mice and measured NK cell production of GM-CSF . Notably , IL-23p19-deficient mice resembled CD11cΔSyk mice in that NK cells taken from the kidneys of either strain displayed markedly reduced levels of GM-CSF but not IFN-γ mRNA and did not secrete GM-CSF protein upon short-term ex vivo culture ( Figure 6D and 6E ) . Furthermore , purified NK cells produced GM-CSF in vitro when stimulated with recombinant IL-23 but not IL-17A/F , C . albicans , curdlan or zymosan ( Figure 6F and 6G ) . GM-CSF production by NKs in response to C . albicans occurred only in the presence of DCs derived from wild-type but not CD11cΔSyk or IL-23p19-deficient mice ( Figure 6H ) . Finally , IL-23p19 KO mice infected with C . albicans were undistinguishable from CD11cΔSyk mice in having massively increased kidney fungal burdens ( Figure 6I ) that could be reversed by GM-CSF therapy ( Figure 6J ) . Together , these data suggest that Syk-dependent IL-23p19 production by DCs in response to C . albicans acts directly on NK cells to promote GM-CSF production and subsequent resistance to systemic candidiasis .
Multiple receptors on macrophages , monocytes , neutrophils , NK cells and innate lymphocytes , as well as on non-immune cells , mediate the recognition of microbes and are thought to act co-ordinately and somewhat redundantly to provide innate resistance to infection . Here , we demonstrate that DCs , a cell type chiefly known for its ability to initiate adaptive immunity , coordinate the entire innate immune control to systemic infection with C . albicans and show that this orchestration depends on a single kinase , indicating a remarkable lack of redundancy in innate immune pathways . We further unravel a hitherto unappreciated series of cellular interactions whereby DCs provide IL-23p19 to NK cells that allows for production of GM-CSF , which in turn maintains the microbicidal activity of neutrophils , the main candidacidal effectors . Disruption of this cellular relay in CD11cΔSyk or IL-23p19 KO mice causes susceptibility to systemic candidiasis and restoration of resistance can be achieved with GM-CSF treatment . Thus , our analysis reveals Syk mediated IL-23p19 production by DCs as a central and non-redundant node of innate immunity to fungal infection and an unexpected indirect regulator of neutrophil microbicidal activity via NK cells . Although the central function of neutrophils in innate protection from disseminated candidiasis is undisputed , the role of mononuclear phagocyte populations is not well established . It is surprising that loss of Syk from CD11c+ cells in CD11cΔSyk mice causes such a dramatic phenotype . We show that this is not because Syk uniquely regulates the development of particular CD11c+ subsets that coordinate anti-fungal immunity or even their recruitment to the site of infection , as there were no gross alterations in the composition of CD11c+ populations in kidneys from infected CD11cΔSyk mice . As in spleen and many other organs , kidney CD11c+ cells are also MHC-II+ and would therefore traditionally be defined as DC . However , not all kidney CD11c+ MHC-II+ cells are derived from committed DC precursors , leading to debate as to whether they are best classified as DCs or macrophages [59] , [60] . Taking advantage of a new Clec9a-Cre line to selectively target those cells derived from pre-DC/CDP [44] , we show that deletion of Syk in the DC lineage ( as defined hematopoietically ) phenocopies deletion in total CD11c+ cells . This would suggest that the susceptibility of CD11cΔSyk mice to systemic candidiasis is primarily due to loss of Syk from DCs . This in turn adds to the emerging notion that DCs may act as central regulators of innate immunity to infection in some instances [61] . Loss of resistance to Candida was also seen in LysMΔSyk mice ( data not shown ) and we do not presently exclude a possible contribution of Syk on CD11c+ cells of monocytic origin ( although we note that such a result is ambiguous as LysM-Cre activity is also found on conventional non-monocytic DCs [62] ) . Whichever their origin , the central role of Syk in DCs suggests that ablation of CD11c+ cells should also have a dramatic phenotype on resistance to Candida infection . Surprisingly , this was not the case as CD11c-DTR ( diphtheria toxin receptor ) mice treated with diphtheria toxin were actually more resistant to infection with C . albicans ( data not shown ) . This apparent discrepancy can be explained by the recent observation that ablation of CD11c+ cells in CD11c-DTR mice is accompanied by marked neutrophilia , which provides a major barrier to bacterial or , in this case , fungal infection [63] . It is notable that deletion of MyD88 in CD11c+ cells had no impact on C . albicans infection even though it markedly impacts responses to TLR agonists in vivo [41] . This may suggest a primacy of Syk-coupled rather than MyD88-coupled receptors in fungal recognition by DCs [1] , [22] , [64] . Nevertheless , MyD88 remains an important component of anti-fungal resistance as we find , along with Villamon et . al . [65] , that MyD88-deficient animals are very susceptible to systemic C . albicans infection . Unlike that of CD11cΔSyk mice , this susceptibility is not preventable by exogenous GM-CSF therapy and presumably involves MyD88 signaling in CD11c− cells . Whether this happens downstream of TLRs or receptors for IL-1 family cytokines remains to be determined . Depletion of neutrophils dramatically increases susceptibility of mice to experimental systemic C . albicans infection [31] and neutropenia places patients at severe risk from systemic candidiasis [32] , [66] . Previous work showed that IL-6-deficient mice are highly susceptible to systemic C . albicans [67] , which was attributed to a lack of neutrophil recruitment without impairment of candidacidal capacity [68] . It was therefore surprising to observe the opposite phenotype , namely normal neutrophil recruitment but impaired activity in infected CD11cΔSyk or Clec9aΔSyk mice . The ample production of neutrophil recruiting proteins such as IL-6 , KC and MIP-2 ( CXCL2 ) in the kidneys of such mice might account for unabated neutrophil recruitment . In contrast , the lack of GM-CSF in the microenvironment appears to be responsible for the loss of neutrophil activity . Neutrophil activation triggers re-localization of intracellular pools of CD11b to the plasma membrane [69] allowing for adhesion , migration and phagocytosis [48] . Thus , we suggest that the presence of CD11blo neutrophils is indicative of poorly activated cells with decreased microbicidal potential , as highlighted by our killing assays . Interestingly , GM-CSF has been linked to neutrophil functionality and survival via physical coupling of Src family kinase Lyn to the GM-CSF receptor , resulting in down-regulation of pro-apoptotic factors and up-regulation of anti-apoptotic pathways such as PI3K/Ark [70]–[72] . While we have failed to observe obvious signs of neutrophil apoptosis in CD11cΔSyk mice , we cannot exclude that any apoptotic cells might be removed rapidly and that the CD11blo phenotype is indeed a prelude to cell death . Notably , intravenous GM-CSF infusion is curative in cases of severe drug-resistant chronic mucocutaneous candidiasis [73] and patients with oral pseudomembranous candidiasis resulting from radiotherapy for head and neck cancers have been successfully treated with a GM-CSF mouthwash [74] . In addition , human neutrophil activation and survival relies in part on NK cell-derived cytokines , including GM-CSF [75] , and activated human NK cells enhance neutrophil survival and promote an increase in neutrophil CD11b expression and ROS production in a GM-CSF dependent manner [76] . Thus , GM-CSF , in part derived from NK cells , may underlie resistance to Candida infection not only in mice but also in Man . This is seemingly at odds with the fact that NK cell deficiency is associated primarily with viral rather than fungal infections [77] . However , the very few NK cell-deficient individuals studied so far may not have been exposed to the conditions predisposing to systemic candidiasis such as catheter insertion or deep tissue surgery . Alternatively , NK-cell independent mechanisms may compensate in these individuals for GM-CSF-dependent fungal control . In this regard , the requirement for NK cells in antifungal immunity even in mice may vary depending on the Candida strain in question [34] , [78] . We have recently shown that the functional development of NK cells in mice requires cell-intrinsic IL-17RA-mediated signals [37] . NK cells that develop in the absence of such signals are impaired in their ability to produce IFN-γ , kill target cells , as well as produce GM-CSF to control Candida infection [37] . Here , we show that even in IL-17RA-sufficient mice , where NK cell functional development is unaffected , the response of NK cells to acute Candida challenge is under stringent environmental control and requires exogenous priming signals . Priming signals for GM-CSF but not IFN-γ production in turn require Syk signaling in DCs as demonstrated by the fact that transfer of resting NK cells does not restore resistance of CD11cΔSyk mice to Candida , yet resistance is achieved if the transfer involves activated NK cells that were primed in an environment in which DCs express Syk . Together with the fact that Clec9aΔSyk and CD11cΔSyk phenocopy each other , this argues against the possibility that the defect in CD11cΔSyk mice is due to deletion of Syk in the NK cells themselves ( even if a small population of NK cells can express CD11c and show evidence of Cre activity in CD11c-Cre mice [79] ) . Supporting this contention , purified NK cells do not respond directly to C . albicans ex vivo but will readily do so in the presence of DCs or conditioned medium from Candida-treated DC cultures . This is consistent with the notion that accessory cells , such as DCs , monocytes , and macrophages are necessary for activation of NK cells in response to most pathogens ( reviewed in [80] , [81] ) . Nevertheless , it is possible that the anti-fungal activity of primed NK cells additionally requires signaling via Syk-coupled NK cell receptors and it will be interesting to study the phenotype of mice in which Syk is selectively ablated in NK cells as opposed to DCs . It is well known that stimulation of DCs and macrophages by C . albicans yeast and hyphae induces the production of IL-2 , IL-6 , IL-12 , IL-23 and TNF-α in a Syk-dependent manner [21] , [27] . In searching for which one of these or other factors might be responsible for priming NK cells to produce GM-CSF we focused on IL-23p19 . We show that IL-23p19 is not induced in Syk-deficient DCs during systemic candidiasis and that Candida-stimulated control but not IL-23p19 KO DCs induce GM-CSF production by NK cells . We further show that IL-23p19 KO mice are very susceptible to systemic candidiasis , as previously suggested [55] , but can be protected by GM-CSF treatment . Together , these data suggest that IL-23 might be the key Syk-dependent cytokine driver of DC-mediated resistance to candidiasis , consistent with the fact that addition of recombinant IL-23 and not recombinant IL-17A/F to purified NK cells induces GM-CSF production . This sheds light on a novel regulatory mechanism of cytokine production by NK cells that selectively affects GM-CSF but not IFN-γ secretion . However , IL-23 is composed of both the IL-23p19 and the IL-12/IL-23p40 subunits and it has been reported that IL-12p40-deficient mice are resistant to candidiasis [82] , [83] . We have been able to reproduce this finding ( unpublished observations ) and therefore , at present , we are forced to conclude that the key mediator of resistance is either a novel IL-23p19-containing cytokine ( including , possibly , an IL-23p19 homodimer ) or that IL-23 deficiency impairs resistance to candidiasis in an IL-12-sufficient but not IL-12-deficient background . While work to assess these possibilities is ongoing , our existing data nevertheless argue for a model ( Figure 7 ) where Syk-mediated recognition of fungal particles by DC , possibly through Syk-coupled CLRs , leads to production of an IL-23p19-containing cytokine , which acts on NK cells in the kidney to induce GM-CSF production . In turn , GM-CSF acts on recruited neutrophils to sustain microbicidal function . This unusual cellular relay from DCs to NK cells and to neutrophils via IL-23p19 and GM-CSF , respectively , provides a key axis for protection from disseminated candidiasis in mice that may be worth exploring as a possible therapeutic target in the context of fungal sepsis in humans .
All animal protocols were carried out under the authority of a UK project license ( number PPL 80/2309 ) approved by the CRUK London Research Institute Animal Welfare and Ethical Review Body in strict accordance with UK governmental regulations ( Animal Scientific Procedures Act 1986 ) or under protocols approved by the Veterinary office of the Canton Zürich , Switzerland ( license number 184/2009 and 201/2012 ) in strict accordance with the guidelines of the Swiss Animal Protection Law . All efforts were made to minimize animal suffering and ensure the highest ethical and humane standards . Control mice ‘Control’ ( including C57Bl/6 and littermate CD11cCre− [42]×Sykflox/flox [40] ) , CD11cΔSyk ( CD11cCre+×Sykflox/flox ) , CD11cΔMyD88 ( CD11cCre+×MyD88flox/flox [41] ) , Control MyD88 ( CD11cCre−×MyD88flox/flox ) , MyD88 KO [84] , Clec9a ( egfp/egfp ) [45] ( DNGR1 deficient ) , Clec9aΔSyk ( Clec9aCre/Cre [44]×Sykflox/flox ) , Clec9a control ( Clec9a+/+×Sykflox/flox ) and IL-23p19 KO [85] , [86] were crossed and bred at Cancer Research UK and at the Institute of Laboratory Animal Sciences , University of Zürich , Switzerland , in specific pathogen-free conditions . Candida albicans strains SC5314 and CAI4-pACT1 GFP ( described in [87] ) were grown by agitation overnight at 30°C in yeast peptone dextrose ( YPD ) or synthetic complete medium ( SC ) containing 2% glucose and Drop-out mix without Uridine . The cells were then washed twice with PBS before use as live yeasts . Heat-killed C . albicans ( HKCA ) yeast or hyphae were generated by boiling samples for 1 h . Mice aged 8–20 weeks were infected intravenously with 2×105 live C . albicans yeast unless stated . The mice were killed 2 days post-infection except where indicated and perfused with cold PBS . Mice that received GM-CSF treatment had two intraperitoneal doses of murine GM-CSF ( Peprotech ) 5 µg/mouse at time of infection and 24 h later . The kidneys were removed and homogenized in 1 ml PBS using an IKA T25 digital Ultra-Turrax homogenizer or a Qiagen Tissue Lyser . Serial dilutions were plated on YPD agar plates and the total number of colony forming units was calculated . Samples were fixed in 10% Neutral Buffered Formalin and processed by the histopathology laboratory at Cancer Research UK . Samples were dehydrated with ethanol and embedded in paraffin . Periodic Acid Schiff ( PAS ) and hematoxylin and eosin ( H&E ) were used to assess fungal invasion and leukocyte infiltration respectively . Single-cell kidney , spleen and bone marrow ( BM ) suspensions were prepared from PBS perfused mice . Kidneys and spleens were chopped into small pieces and digested in RPMI 1640 medium supplemented with glutamine , penicillin , streptomycin , ( all from Gibco ) , collagenase type IV ( 200 u/ml , Worthington ) , and DNase 1 ( 0 . 2 mg/ml , Roche ) for 1 h or 30 min respectively at 37°C . Cells were then passed through a 70 µm cell strainer ( BD bioscience ) , washed with RPMI 1640 supplemented with 10% fetal calf serum , glutamine , penicillin and streptomycin ( RPMI complete medium ) . Single cell kidney samples were then placed onto a non-continuous isotonic Percoll ( GE Healthcare ) gradient of 78% and 37% and centrifuged for 30 min at 550 g . The interface was collected from these samples and washed in RPMI complete medium . For isolation of BM cells , the femur and tibia were collected from both hindquarters . Bones were flushed with RPMI complete medium and passed through a 70 µm cell strainer to obtain single cell suspensions . Splenic and BM erythrocytes were lysed with Red Blood Cell Lysis Buffer ( Sigma ) for 3 min at room temperature ( RT ) . Single cell populations were subsequently used for either FACS staining , in vitro candidacidal activity or cell sorting . Data were collected on LSR Fortessa , FACSAria or LSRII ( all BD Biosciences ) and analyzed using FlowJo software ( Tree Star ) . The staining protocols used combinations of antibodies listed below . Antibodies purchased from BD bioscience included: anti-CD3e ( 145-2C11 ) , anti-CD4 ( RM4-5 ) , anti-CD8 ( 53-6 . 7 ) , anti-CD11b ( M1/70 ) , anti-CD11c ( HL3 ) , anti-CD16/CD32 ( 2 . 4G2 , Fc block ) , anti-CD19 ( 1D3 ) , anti-CD45R/B220 ( RA3-6B2 ) , anti-CD49b ( DX5 ) , anti-CD64 ( X54-5/7 . 1 ) , anti-IFN-γ ( XMG1 . 2 ) , anti-Ly6G ( 1A8 ) and Streptavidin-APC . The following antibodies were purchased from eBioscience: anti-CD3e ( 145-2C11 ) , anti-CD11b ( M1/70 ) , anti-CD103 ( 2E7 ) , anti-GM-CSF ( MP1-22E9 ) , anti-MHC-II ( M5/114 . 15 . 2 ) and anti-NK1 . 1 ( PK136 ) . The following antibodies were purchased from Biolegend: anti-CD11a ( M17/4 ) , anti-CD11c ( N418 ) , anti-CD18 ( M18/2 ) , anti-CD45 . 2 ( 104 ) , anti-F4/80 ( BM8 ) , anti-Syk ( 5F3 , purified and conjugated to AF647 with AF647 Antibody labelling kit ( Molecular Probes ) ) . Additional antibodies used were Polyclonal Goat anti-Mouse/Rat MyD88 ( R&D ) , anti-Mouse IL-23R ( 753317 , R&D ) , Rabbit anti-Goat AF488 ( Molecular Probes ) and anti-MPO ( 8F4 , biotinylated , Hycult Biotech ) . Single cell suspensions were surface stained directly ex vivo or following 7 h incubation with brefeldin A ( Sigma ) . Most cell staining involved dead cell exclusion by live/dead fixable violet dye ( Invitrogen ) , followed by fixation with 2% paraformaldehyde ( Electron Microscopy Sciences ) for 20 min at RT . Cells were washed twice with FACS buffer ( PBS , 1% FCS , 2 mM EDTA ) . For intracellular staining , cells were subsequently permeabilised and stained with saponin containing Reagent B ( ADG Bio Research GMBH ) . Kidneys were removed 1 or 2 days post-infection following PBS perfusion and homogenized on ice in 0 . 5 or 1 ml of PBS respectively . Chemokines and cytokines from homogenates and cell culture supernatants were analyzed according to manufacturer's instructions . Briefly , clarified samples were incubated with either BD cytometric bead array kits ( IL-6 , KC , MIP-1α , TNF , IL-1α ) , FlowCytomix Kits ( IL-15/IL-15R , MCP-3 and IL-10 ) , R&D Quantikine ELISA kit ( IL-1β ) or eBioscience Ready-Set-Go ELISA kit ( GM-CSF ) . Bead based assays were assessed using a LSR Fortessa whilst ELISA samples were read at 450 nm with all concentrations determined relative to a standard curve . For proteome profiling , kidneys were removed from naïve or 16 h post-infection mice following PBS perfusion and homogenized in 1 ml PBS with protease inhibitor cocktail ( cOmplete Roche ) with Triton ×100 added at a final concentration of 1% prior to a freeze thaw step . Samples were clarified prior to addition to the R&D Proteome profiler ( Mouse cytokine array panel A ) and developed as per manufacturer's instructions . Relative pixel density of each duplicate blot was measured using Image J software . The data are presented as fold change in signal from infected samples compared to naïve control samples . Single cell suspensions of kidney and bone marrow were prepared as above prior to staining and sorting for neutrophils , identified by CD11b+ Ly-6G+ F4/80− and GFP+ or GFP− . Sorted neutrophils were incubated with C . albicans ( 10∶1 ) in serum free medium on ultra low attachment plates ( Costar ) for 1 h at 37°C . Wells were collected and cells lysed with water prior to plating on YPD agar . C . albicans colony formation from neutrophil-containing wells was compared to that from control neutrophil-free wells to calculate the percentage of C . albicans killed . Data are combined from three independent experiments with each data point representing an individual well . Neutrophils were also isolated from mouse blood using a density gradient of Histopaque 1119 and Histopaque 1077 ( both Sigma ) . Blood neutrophil killing activity was assessed using 104 C . albicans yeast co-cultured with 104 neutrophils ( usually >80% Ly6G+ ) in protein low binding tubes ( Sarstedt ) for 2 h . The percentage of C . albicans killed was assessed as above with data combined from two independent experiments . Single cell suspensions of spleens from control mice were obtained as described above and NK cells were either enriched with anti-DX5 microbeads ( Miltenyi Biotech ) or purified by FACS based on DX5 and CD3 expression . 8×106 enriched NK cells or 4×106 FACS purified NK cells ( >95% pure and viable ) were adoptively transferred into recipient mice 1 h prior to infection . DCs were differentiated from BM precursors in presence of GM-CSF for 7 days . 105 FACS-purified NK cells from naïve spleens were cultured alone or co-cultured with 5×104 DCs in presence of heat-killed C . albicans ( 10 M . O . I . ) , 100 µg/ml Curdlan , 50 µg/ml Zymosan , recombinant IL-23 ( BD bioscience; 100 ng/ml ) or recombinant IL-17A/F heterodimer ( BD bioscience; 1 µg/ml ) . The culture supernatant was collected after overnight incubation and GM-CSF was quantified by ELISA ( eBioscience ) according to manufacturer's instructions . RNA was extracted from whole organs disrupted using the Tissue Ruptor ( Qiagen ) using TRIzol ( Invitrogen ) according to manufacturer's instructions . RNA from FACS sorted cell samples was isolated using either TRIzol or the QIAcube ( Qiagen ) . Isolated RNA was reverse transcribed into complementary DNA using random primers ( Invitrogen ) . Quantitative PCR was performed using Taqman primer/probe sets ( Invitrogen ) , Sykb ( Mm01333035_m1 ( exon boundary 1–2 ) ) , csf2 ( Mm00438328_m1 ) , ifng ( Mm01168134_m1 ) and house keeping Gapdh ( Mm99999915_g1 ) or SYBR Green ( Qiagen ) with primer pairs il23a ( F-GCCAAGAAGACCATTCCCGA R-TCAGTGCTACAATCTTCTTCAGAGGACA ) and Gapdh ( F-CAGTATTCCACTCTGAAGAAC R-ATACGGCCAAATCTGAAAGAC ) using either the Viia7 or 7500 Fast Real-Time PCR System ( Applied biosystems ) . Prism version 6a ( GraphPad ) was used for plotting data and for statistical analysis . Survival data are presented as a Kaplan-Meier plot with a log rank test used to compare significance between groups . Data was subjected to D'Agostino & Pearson omnibus normality test , Shapiro-Wilk normality test and Kolmogorov-Smirnov test to determine the subsequent statistical tests applied . Statistical significance of differences between two groups or groups with fewer than three samples was determined by 2-tailed t test . For experiments with more than 2 groups , significance of any differences was determined using a 1-way ANOVA with Tukey multiple comparison of all pairs for post-test analysis . If the data was assessed to be non-gaussian then a Kruskal-Wallis with Dunn's multiple comparison test was undertaken . The level of significance was defined as *p<0 . 05 , ** p<0 . 01 , *** p<0 . 001 , **** p<0 . 0001 . The test used for statistical analysis is indicated in each figure legend .
|
Multiple cell types bearing a vast array of immune receptors with different modes of signaling ensure that the host response to infection is both robust and reliable . For this reason , loss of a single signaling pathway in a given cell type is often not enough to impact host resistance . Here , we find , surprisingly , that this is not the case in a mouse model of systemic fungal infection with Candida albicans . We show that a single kinase ( Syk ) in a single cell type ( dendritic cells , DCs ) coordinates the entire host resistance network . We highlight Syk-dependent production of IL-23p19 by DCs as the key to protection and show that IL-23p19 acts on another white blood cell type , NK cells , to specifically induce production of another mediator , GM-CSF . The latter is key for yet another cell , the neutrophil , to be mobilized into action and kill Candida organisms . This study places DCs , best known for their role in priming T cells , at the center of a cellular relay of innate immunity to fungal infection . It highlights key nodes of antifungal immunity that could be targeted in combination with antifungal drugs to provide new ways to treat patients with fungal sepsis , who generally have poor outcomes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"developmental",
"biology",
"innate",
"immune",
"system",
"white",
"blood",
"cells",
"immune",
"cells",
"cytokines",
"cell",
"biology",
"animal",
"cells",
"nk",
"cells",
"immunity",
"molecular",
"development",
"antigen-presenting",
"cells",
"immunity",
"to",
"infections",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immunology",
"immune",
"response",
"immune",
"system"
] |
2014
|
Syk Signaling in Dendritic Cells Orchestrates Innate Resistance to Systemic Fungal Infection
|
Apolipoprotein E ( apoE ) is a forefront actor in the transport of lipids and the maintenance of cholesterol homeostasis , and is also strongly implicated in Alzheimer’s disease . Upon lipid-binding apoE adopts a conformational state that mediates the receptor-induced internalization of lipoproteins . Due to its inherent structural dynamics and the presence of lipids , the structure of the biologically active apoE remains so far poorly described . To address this issue , we developed an innovative hybrid method combining experimental data with molecular modeling and dynamics to generate comprehensive models of the lipidated apoE4 isoform . Chemical cross-linking combined with mass spectrometry provided distance restraints , characterizing the three-dimensional organization of apoE4 molecules at the surface of lipidic nanoparticles . The ensemble of spatial restraints was then rationalized in an original molecular modeling approach to generate monomeric models of apoE4 that advocated the existence of two alternative conformations . These two models point towards an activation mechanism of apoE4 relying on a regulation of the accessibility of its receptor binding region . Further , molecular dynamics simulations of the dimerized and lipidated apoE4 monomeric conformations revealed an elongation of the apoE N-terminal domain , whereby helix 4 is rearranged , together with Arg172 , into a proper orientation essential for lipoprotein receptor association . Overall , our results show how apoE4 adapts its conformation for the recognition of the low density lipoprotein receptor and we propose a novel mechanism of activation for apoE4 that is based on accessibility and remodeling of the receptor binding region .
Apolipoprotein E ( apoE ) is a member of the superfamily of exchangeable apolipoproteins . It mediates cellular uptake of cholesterol-rich lipoproteins by acting as a high affinity ligand for cell surface receptors belonging to the low-density lipoprotein ( LDL ) receptor family [1] . An imbalance in cholesterol homeostasis increases the risk for cardiovascular diseases and is also linked to neurodegenerative disorders [2 , 3] . Therefore , the receptor binding property of apoE stresses its importance in the transport of lipids and metabolism of cholesterol both within the plasma and the central nervous system [4 , 5] . In blood plasma , the receptor mediated uptake and endocytosis of apoE-containing lipoproteins lowers the overall levels of circulating lipoproteins , explaining the anti-atherogenic effect of apoE [6] . In the brain , although apoE is involved in lipid redistribution and neuronal growth and repair , the presence of the ε4 allelic form of the apoE gene also represents the most significant genetic risk factor of developing Alzheimer’s disease [7] . An abnormal trafficking of lipids and cholesterol by apoE4 is among the pathogenic mechanisms that are proposed to contribute to the susceptibility of ε4 carriers for Alzheimer’s disease [8 , 9] . ApoE is a ~34 kDa protein composed of 299 amino acids . Single point variations at positions 112 and 158 distinguish the three main isoforms of apoE: apoE2 ( Cys112 , Cys158 ) , apoE3 ( Cys112 , Arg158 ) and apoE4 ( Arg112 , Arg158 ) [10] . These sole amino acid substitutions result in structural differences between these isoforms [11] and marked effects on their lipid binding abilities [12] , providing grounds to explain their different physiological role ( s ) in cardiovascular and Alzheimer’s diseases [13] . In the lipid-free state , all three apoE isoforms possess two independently folded structural domains linked by a protease sensitive loop [14] . The N-terminal ( NT ) domain ( res . 1 to 191 ) comprises an elongated four-helix bundle that contains the binding region to the members of the LDL receptor family on the fourth helix [15] . The C-terminal ( CT ) domain ( res . 210 to 299 ) presents the major lipid binding region [16] and is particularly challenging to study , as it is involved in the oligomerization of apoE in the absence of lipids [17] . Several mutations had to be introduced in the CT domain to generate a stable monomeric protein leading to the so far only available full-length high resolution three-dimensional ( 3D ) structure of a lipid-free apoE protein . In this structure , the CT domain variant contains three α-helices folded upon the NT domain conferring a globular shape to apoE [18] . Upon binding to lipid particles , apoE undergoes a large conformational conversion to accommodate and stabilize the lipids through its amphipathic α-helices , allowing thereby their trafficking in the circulation [19] . Additionally , lipid binding induces apoE to adopt a biologically active conformation that is a prerequisite for the binding of lipoproteins to cell surface LDL receptors and their internalization [1] . Analysis of reconstituted discoidal phospholipid-apoE particles ( rHDL , more recently termed nanodisk ) presented a major step forward towards a structure of lipid-bound apoE [19] . These particles mimic in vivo nascent high density lipoproteins ( HDL ) in shape , size , density and functional properties [20] . It was demonstrated that in these systems , the α-helices of apoE are oriented perpendicularly to the acyl chains of the lipids and the apolipoprotein molecules circumscribe the edge of the discoidal particles [21–23] . Lipid-binding also triggers the elongation of NT domain helix 4 which was proposed to represent a key lipid-induced conformational change allowing for the recognition of apoE by LDL receptors [24 , 25] . However , the conformation adopted by apoE molecules at the surface of these discoidal particles remains an open question . While it is accepted that the CT domain adopts an extended α-helical structure [19 , 22 , 23] , the conformation of the NT domain has not converged towards a single model . Based on calorimetry measurements , it was proposed that the four-helix bundle opens to expose the hydrophobic faces of the amphipathic helices towards the lipids and that further reorganization of helices occurs , triggered by lipid binding [26] . Although this bundle opening was suggested to ultimately lead to a fully extended conformation of apoE that wraps around the entire circumference of the lipid bilayer of the disc [22] , several studies have indicated that apoE adopts a hairpin structure for which distinct hinge localizations were proposed [27–29] . Supported by low resolution X-ray density and electron paramagnetic resonance ( EPR ) measurements , an alternative model was developed . In this case , even though apoE also folds in a hairpin structure , the hydrophobic faces of apoE helices are suggested to interact with each other , while the polar faces contact the phospholipids leading to ellipsoidal lipoparticles [30 , 31] . Despite two decades of intensive structural studies , a consensus on the conformation of lipidated apoE has not yet been reached . With the aim of deciphering the molecular structure adopted by apoE at the surface of rHDL in solution , we designed an approach where complementary low resolution structural data were combined with 3D structural modeling ( Fig 1 ) . We decided to focus our present work on rHDLs containing only apoE4 , considering the prevalent role of this isoform in Alzheimer’s disease [8 , 9] . Experimental data were primarily generated from chemical cross-linking ( XL ) coupled to mass spectrometry ( -MS ) which produces covalently connected pairs of peptides that provide a set of distance restraints between cross-linked residues on the native protein , enabling low resolution models to be elaborated . XL-MS has seen significant progress recently [32–37] and has been successfully applied to a large number of protein complexes [38–41] . The distance restraints from our intramolecular XLs , together with additional experimental data obtained in this work and information from the literature were then used in our hybrid molecular modeling approach . Two alternative models of lipidated apoE4 were validated by our XL-MS results and assessed by molecular dynamics simulations . Our resulting models represent the most detailed structures obtained so far on full-length apoE4 associated to rHDL and they provide unprecedented insight into the active structure of apoE4 . Taken together the data allowed us to propose a novel molecular mechanism that explains how apoE is recognized by the members of the LDL receptor family .
ApoE can bind to lipoproteins of variable sizes and shapes due to its conformational flexibility [42] . To obtain detailed information on lipidated apoE4 conformation , it was desirable to obtain highly homogeneous lipoproteins , in order to stabilize a uniform apoE4 conformation . For the preparation of such rHDL , we used the cholate dialysis method [43] and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine ( POPC ) as a source of phospholipids . The initial lipid:protein molar ratio was optimized to enhance homogeneity of apoE4/POPC particles by following the resulting rHDL migration on native PAGE ( S1 Fig ) . At an initial apoE4/POPC molar ratio of 1:110 , a single population was apparent , displaying a Stokes diameter of ~105 Å ( Fig 2A ) . A single population was also detected by gel filtration of apoE4/POPC rHDL ( Fig 2B ) . Finally , quantification of the protein content indicated that the apoE4/POPC reconstitution allowed recovering up to ~50% of the protein initially engaged in rHDL . These reconstitution parameters allowed us to prepare highly homogeneous apoE4/POPC rHDL particles in a reproducible manner . To provide detailed input data for the structural modeling of lipid-bound apoE4 , we extensively characterized their composition and shape . Quantification of the concentration of lipid by phosphorus assay revealed that each particle contained about 200 POPC molecules . Chemical XL with large excess of bis ( sulfosuccinimidyl ) suberate ( BS3; molar ratio BS3:apoE4 200:1 ) followed by SDS-PAGE analysis resulted in a single apoE4 dimer band at approximately 70 kDa , indicating that two molecules of apoE4 were present on each apoE4/POPC rHDL . Infrared measurement revealed a sharp peak for the amide I band ( 1700–1600 cm-1 ) centered at 1652 cm-1 characteristic of α-helical structures [44] ( Fig 2C ) . An α–helical content of approximately 60% was estimated by curve-fitting of the amide I band . ApoE4/POPC rHDL were next characterized by both negative staining ( NS ) and cryo-transmission electron microscopy ( TEM ) . Most representative class-averages of NS-TEM images revealed that apoE4/POPC rHDL mainly appeared to be of circular shape with a diameter of 115 ± 10 Å ( S2A Fig ) . To determine the overall shape of the rHDL particles in hydrated state and to avoid possible artifacts due to sample drying and heavy metals on the observed shape , we visualized the apoE4/POPC rHDL by cryo-TEM ( S2B Fig ) . The homogeneity of the apoE4/POPC reconstitution enabled us to perform single particle analysis . Two-dimensional averages identified both top and side views of the rHDL particles ( Fig 2D ) . These projections displayed a diameter similar to the one previously measured in NS-TEM and a thickness of 50 ± 10 Å , in good agreement with the expected thickness of a POPC bilayer [45] . NS- and cryo-TEM images therefore strongly support a discoidal shape for the apoE4/POPC rHDL . These particles will further be designated as apoE4 nanodiscs in this work . The apoE4 protein conformation at the surface of nanodiscs was investigated by XL-MS using the homobifunctional disuccinimidyl suberate ( DSS ) cross-linker that reacts with primary amino groups ( Lys residues and protein N-termini ) . The extended Cα-Cα distance for lysine pairs that can be cross-linked by DSS is usually considered to have an upper limit of about 30 Å [46 , 47] . An equimolar mixture of light DSS ( DSS-H12 ) and heavy DSS ( DSS-D12 ) was used , providing a unique isotopic signature to cross-linked peptides and facilitating their detection and identification by MS [48 , 49] . The apoE4 molecules at the surface of the nanodiscs were cross-linked with 8 moles of DSS for 1 mole of apoE4 . The low DSS/apoE4 molar ratio was chosen to minimize the risk of disturbing the structure adopted by apoE4 in the nanodiscs . The resulting species were isolated by SDS-PAGE revealing two bands of comparable intensity at approximatively 35 and 70 kDa , which were assigned to cross-linked monomeric apoE4 and cross-linked dimeric apoE4 , respectively ( Fig 3A ) . To generate exclusively unambiguous intramolecular XL products , the monomeric apoE4 band at 35 kDa was processed by in-gel digestion with trypsin and analyzed by liquid chromatography MS/MS . The resulting fragment ion spectra were analyzed using the dedicated software pipeline xQuest/xProphet [46 , 48] . 27 cross-linked peptides were identified for monomeric apoE4 ( S1 Table ) , which corresponded to 22 unique Lys-Lys distance restraints ( Fig 3B and Table 1 ) . Evaluation of the intramolecular XL data set obtained for the apoE4 nanodiscs revealed that both CT and NT regions of the protein are covered by the ensemble of XLs , with 11 out of the 12 apoE Lys residues involved in at least one XL . The XLs can be classified into two main categories . The first , and largest , group comprises XLs that were formed between Lys residues located in the NT domain and the CT domain ( Fig 3B and 3C , dotted lines ) . The second group contains pairs of Lys residues belonging to the NT domain only , the vast majority connecting different helices forming the four-helix bundle adopted by apoE in its soluble form ( Fig 3B and 3D , dashed lines ) . Topological information on the conformation of lipidated apoE4 could be deduced from the distance restraints derived from the XL data . The distribution and number of intramolecular XLs between Lys residues of the NT and CT domains were inconsistent with a completely extended conformation of apoE4 at the surface of the nanodiscs . They rather suggested a hairpin conformation ( Fig 3C ) . Besides , the scattering and number of intra-NT domain XLs were indicative of a relatively compact state of the NT helix bundle ( Fig 3D ) . Once the in-depth experimental characterization of the nanodiscs was achieved , we set out to generate a model of apoE4 bound to rHDL . To do so a two-step procedure was set up ( Fig 1 ) : first , monomeric conformations of apoE4 were constructed by molecular modeling using experimental data to guide the modeling process . Then , dimer assemblies of these monomer structures were wrapped around an explicit lipid disc and the evolution over time of these systems was investigated by molecular dynamics simulations . In a first modeling approach , we directly used all the intramolecular XLs as long and medium-range distance restraints so as to generate a structural model of monomeric lipidated apoE4 . However , this attempt was unsuccessful as the ensemble of XLs restraints could not generate any concluding structures that would fit the experimental characterization of the nanodiscs ( shape and size ) . From this first approach , it appeared evident that the ensemble of XL data would not be satisfied by a single ultimate model , hinting at the presence of at least one alternative conformation . We therefore devised a second approach in which here-acquired structural data were rationalized in the light of current knowledge on lipidated apoE to narrow the range of conformational states apoE4 could adopt at the surface of nanodiscs ( Fig 1 ) . They were implicitly included in sets of constraints for the structure generation ( S1 Text and S2 Table ) . First , to fulfill the hairpin conformation suggested by the NT-CT spatial proximity , evidenced by our XL-MS data ( Fig 3C ) , we inserted a hinge , allowing the CT domain to fold back along the NT domain . To determine the apex of the hairpin , we tested three different hinge positions in the non-structured portions of apoE4 connecting the NT and CT domains ( res . 164 to 168 , 186 to 193 , or 201 to 208 ) . Although our XL data pointed toward a relatively compact conformation of the NT domain ( Fig 3D ) , we conjectured that an apoE4 NT domain conformation completely folded as in solution would be hardly compatible with a receptor active conformation , as it is commonly accepted that opening of the NT bundle upon lipid interaction is a prerequisite for exposure of NT helix 4 containing the region involved in recognition of LDL receptors [19] . Therefore , based on literature [18 , 22 , 28 , 29] , we decided to partially open the NT four-helix bundle by unfurling the turn in between NT helices 3 and 4 and aligning these two helices with the CT domain in a hairpin conformation by using a zipping procedure . Second , we maintained NT helices 1 to 3 bundled together by applying a zipping procedure between NT helices 2 and 3 , thus promoting their spatial proximity to comply with the XL-MS data and to place NT helix 2 outside of the implicit lipid disc . On the other hand , due to the lack of XL data for NT helix 1 , which does not contain any Lys residue , preventing us to rule on its position , we chose to keep this helix in contact with NT helix 2 by using the distances and angles from the NMR study of full length mutated apoE3 [18] . A partially opened state comprising a NT three-helix bundle with NT helix 4 detached was hence generated . Finally , we imposed a curvature to adapt the conformation of apoE4 molecules to the experimental discoidal shape of the nanodiscs and applied a distance constraint to move the flexible CT end outside of the nanodisc . Validation of our models by the XL data revealed that , from the structures generated with the three different hinge positions , the model featuring a hairpin structure containing the hinge formed by res . 186 to 193 best matched the XL pattern , satisfying 12 out of 22 XLs ( Table 1 ) . This model was named “opened hairpin” model ( Fig 4A ) and the selected hairpin apex placed NT helices 3 and 4 in juxtaposition to the CT domain , in good agreement with 6 XLs ( out of 11 ) formed between these two domains ( Table 1 ) . Nevertheless , the opened hairpin model left out 10 XLs that failed to comply with its structure . These non-satisfied XL were either intra NT domain ( helices 2/3 connected to helix 4 ) or NT-CT domains XL ( helices 2/3 connected to a different region of the CT domain ) ( Table 1 ) . Careful inspection of the opened hairpin model suggested that these XLs were likely to be satisfied if the NT domain adopted a four-helix bundle . We thus constructed a second monomeric apoE4 model , using the same hinge region ( res . 186 to 193 ) but adjusted the constraint list ( S2 Table ) to retain a compact state of the NT domain bundle . Remarkably , in this second model , named “compact hairpin” model in the following ( Fig 4B ) , 19 out of the 22 identified XLs were validated ( Table 1 ) . The three non-satisfied XLs involved a subset of the NT-CT links ( helices 2/3 with res . 262 and 282 of the CT domain ) that otherwise supported the opened hairpin model . The two conformations proposed here may therefore represent distinct states of lipidated apoE4 that dynamically co-exist in solution . Both the opened and compact hairpin monomeric models were dimerized in either a head-to-head or head-to-tail orientation . They were wrapped around a solvated POPC disc producing four different molecular systems ( S1 Text and S3 Fig ) . In all 4 setups , the final number of lipids contained in the nanodisc is in good agreement with the experimental values we measured , providing a first validation of our models before we further studied their dynamic behavior using molecular dynamics simulation . In the first nanoseconds of the trajectories , the amphipathic α-helices were observed to rearrange so as to more efficiently protect the hydrophobic acyl chains of the lipids located at the edge of the nanodiscs from the solvent . By adjusting their α-helical segments contacting the lipids , two apoE molecules are able to accommodate the number of lipids contained in each lipoprotein particle and match the average diameters of the nanodiscs as they were observed in this study by native PAGE ( Fig 2A ) , NS- and cryo-TEM ( Fig 2D and S2 Fig ) . Further , our 75-ns long trajectories highlighted that the lipid structures kept their disc shape in all cases ( Fig 5A and S4 Fig ) and the majority of the XLs remained satisfied at the end of our simulations ( S3 Table ) . The α–helical content at the end of the simulations calculated with DSSP [50] ranged between 51% and 66% in good agreement with the 60% estimated from our infrared measurements . No significant differences could be evidenced between the systems featuring either a head-to-head or head-to-tail apoE dimer and we therefore could not discriminate between both orientations . However , during the trajectories local changes in the secondary structure were observed in some regions of the protein . Remarkably , a short stretch ( res . 164 to 168 ) at the end of NT helix 4 switched from a random coil to an α-helical conformation and remained α-helical for the rest of the simulation in one of the monomers in all models ( Fig 5B ) . This structural change , close to the binding region to LDL receptors , promoted an extension of helix 4 resulting in a long amphipathic helix spanning res . 131 to 180 ( Fig 5B ) . Furthermore , upon this change Arg172 , known to be involved in the recognition of LDL receptors [51] and other upstream basic residues , also known to interact with the receptor [52] , underwent a reorientation leading to their respective alignment ( Fig 5B ) . Comparison of the solvent accessibility of these residues in our two models ( S5 Fig ) indicated that , while most residues binding to the LDL receptors featured a low accessibility in the compact hairpin model , they really pointed into the solvent in the opened hairpin model regardless of the dimer arrangement . Therefore , although both conformations may co-exist in solution , they may exert variable binding activities towards receptor recognition with the opened hairpin model representing the active conformation of lipidated apoE4 .
The XL-MS distance restraints obtained here from the cross-linked monomeric apoE4 molecules argued against a model where apoE could adopt a completely extended structure surrounding the nanodisc , with two molecules of apoE running along each other in a ‘double-belt’ organization as was proposed previously [22] . A large subset of our intramolecular XL data rather inferred a hairpin fold of lipidated apoE as previously proposed in other studies [29 , 31] . Alike previous models of full-length apoE , our XLs implied the hinge of the hairpin to be situated in the unstructured region connecting the NT and CT domains but with subtle differences resulting in significant structural and mechanistic implications . Specifically , in the so far most detailed Xray/EPR model of lipidated apoE4 [30 , 31] , the hinge is situated at res . 162 to 169 ( vs res . 186 to 193 here ) and suggested to bring in close proximity regions that are known to be important for the interaction with LDL receptors , the region spanning res . 134 to 150 and Arg172 . However , due to the hinge location in this model , the α-helical extension of NT helix 4 , suggested to be essential for receptor binding activity [24 , 25] , is no longer possible . This hinge location was also not supported here , as the model we built with the hinge on res . 164 to 168 only satisfied 9 out of the 22 identified XLs . In spite of the difference in hinge localization , spatial proximity of significant pairs of residues could be reconciled between our and previous hairpin models . For instance , for apoE4 , the spatial proximity of two residues , Arg61 and Glu255 , that are proposed to form a salt bridge promotion the interaction between NT and CT domains in the lipid-free form [60] , was confirmed to be maintained in the lipid-bound state in discoidal particles [29] . The proximity of these residues was also preserved here , thanks to the partially closed conformation of the NT domain . Further , a significant number of EPR constraints [31] were also validated in our models , including the intramolecular spatial proximity of residues 76/77 with residues 239/241 that were established in our study to be intramolecular by the selection of the monomeric band for in-gel digestion ( Fig 3A ) . The here-produced hairpin models thus allow at the same time both spatial proximity of recognized pairs of important residues in CT and NT regions and the opportunity for the extension of helix 4 needed for the recognition of LDL receptors ( Fig 5B ) . However , a limitation of our study is that , in the current setting , we did not specifically discriminate between intra- and intermolecular cross-links within homodimeric apoE proteins and the respective organization of the two apoE molecules on the lipid particle could therefore not be deduced . The head-to-head and head-to-tail dimerizations , as presented in S3 Fig , therefore remain to a certain degree speculative . Two alternative models , featuring three or four bundled amphipathic helices from the NT domain , were constructed that together satisfied the ensemble of XL derived spatial restraints ( Fig 4 and Table 1 ) . The compact hairpin model features a NT four helix bundle laid along the CT domain and interacting with the lipids only via helix 4 . In this conformation , helix 4 , that contains essential residues for recognition of the members of the LDL receptor family [52] , was shielded from the solvent by helix 3 ( Fig 4B and S5 Fig ) . In the opened hairpin model , the turn between helices 3 and 4 was unfurled , in agreement with previous studies [18 , 28] , and allowed an opening of the bundle with NT helix 3 now interacting with the lipids . This partial opening of the bundle was sufficient to expose helix 4 to the solvent ( Fig 4A and S5 Fig ) . The opening movement from the compact to the opened hairpin model therefore provides us with a possible regulatory mechanism of apoE4 lipoproteins ( Fig 6A and 6B ) . In contrast to earlier studies that indicated that the interaction of the NT domain with the lipids would engage an open and active conformation of the receptor binding region [18 , 26] , our models strongly suggest that , in both the open and compact state , the NT domain of apoE is associated with lipids at the surface of the nanodisc . The domains outside the lipid disc in the compact hairpin model ( S3C and S3D Fig ) are not clearly resolved by NS-TEM ( S2 Fig ) . Heterogeneity in the disc size , dynamic structure of NT domain switching between compact and open conformation , and small size of the folded domain preclude its visualization by single particle technique based on averaging of projections of individual aligned particles . Moreover NT helix 1 that was considered in our modeling as part of the NT helix bundle despite the absence of structural data could instead adopt a more extended conformation . Each of these two cases would then contribute to decrease the compactness of the NT portion that may be observed by NS-TEM . We speculate that both the open and compact hairpin model co-exist in a dynamic equilibrium where the different forms could concurrently be captured by our XL experiments . Further , we propose that in presence of the receptor this equilibrium is shifted to the opened hairpin model , the model that represents the state accessible to LDL receptors , and therefore allows us to draw a mechanism of accessibility of the LDL receptor binding region ( Fig 6A and 6B ) . A relatively small structural change , observed in all models during the molecular dynamics trajectories , elongated helix 4 and connected it with a subsequent small helix spanning res . 169 to 180 , leading to the formation of a 50 residue-long amphipathic helix ( res . 131 to 180 ) ( Fig 5B ) . This helix extension upon lipidation has already been proposed experimentally by NMR and EPR [24 , 25] and it was suggested to act as a molecular switch that stably anchors the receptor binding region on the lipid surface or/and correctly positions residue 172 with other basic residues ( in the region 136 to 150 ) known to be required for an optimal interaction with the LDL receptors [51 , 52] . These receptors share highly conserved structural domains , including ligand-binding domains containing cysteine-rich ligand binding type-A ( LA ) repeats . For the LDLr , the most prevalent member of this family of receptors , it is now well established that among the 7 LA repeats , LA5 is essential for binding of apoE lipoproteins [61] and that the pair LA4-LA5 is sufficient to bind apoE in rHDL [62] . The residues known to interact with the LDLr on apoE [51 , 52] span a too large region to be recognized by a single LA repeat of the LDLr and would thus allow for the binding of the two LA repeats to the same apoE molecule as we proposed earlier based on the lipid-bound structure of an apoE-derived peptide [25] . In addition , our study showed that the elongation of NT helix 4 upon lipidation led to a reorganization of the LDLr binding residues that could promote their binding to the LDLr LA4-LA5 repeats ( Fig 5B ) . To support this hypothesis , we performed docking assays in which LA4 and LA5 were docked individually to such an elongated NT helix 4 ( S4 Table ) . The results confirmed that the distance between the two docked modules was in agreement with the long loop between LA4 and LA5 repeats ( S6 Fig ) . This distance is unique between this pair of LA repeats [63] , highlighting its importance in lipidated apoE recognition . Contrary to the soluble form of apoE4 , the elongation of NT helix 4 conferred upon lipidation would therefore represent an additional prerequisite for binding to LDL receptors ( Fig 6D ) . In summary , our data advocate that several requirements need to be met to provide a fully receptor-active apoE ( Fig 6D ) : lipid binding , exposure of the receptor binding region and elongation of NT helix 4 . We speculate that the here proposed compact hairpin model is a stable conformation co-existing with the active , receptor-competent open structure , explaining why these two alternative conformations could be trapped in the XL-MS experiments . Both conformations may therefore be part of a regulation mechanism of apoE function at the surface of lipids . Our work represents a building stone towards a better understanding of the strong anti-atherogenic effect of apoE and the models we are proposing could prove useful in the study of lipidated apoE in totally different contexts , such as understanding its role in Alzheimer’s disease . Overall our hybrid approach , compatible with the presence of lipids , results in 3D structures of lipidated apoE4 that represents the most comprehensive model of the active form of apoE4 to date and might be applied to the study of other ( membrane ) proteins where such complementary low resolution structural data are available .
Unless otherwise stated , all chemicals were obtained from Sigma-Aldrich at the highest purity available . Water was double-distilled and deionized using a Milli-Q system ( Millipore ) . The human full-length apoE4 gene fused to a self-cleavable intein tag and a chitin binding domain cloned into a pTYB2 vector was a kind gift of Dr . Vasanthy Narayanaswami ( University of Long Beach , California , U . S . A . ) . ApoE4 was expressed in a T7 expression strain of Escherichia coli ( ER2566 ) in 2xYT medium by the addition of isopropyl β-D-thiogalactopyranoside . Pelleted cells were resuspended in buffer A ( 20 mM HEPES , 500 mM NaCl , 1 mM EDTA , pH 8 ) , supplemented with 0 . 5% ( v/v ) triton-X-100 ( TX100 ) and anti-proteases ( cOmplete EDTA-free protease inhibitor cocktail , Roche ) , and apoE4 was released by high-pressure homogenizer . The protein was purified following standardized protocols previously described for intein-labelled proteins [64] . Briefly , the clarified cell lysate was loaded onto chitin beads ( Impact system , New England Biolabs ) equilibrated with 5 column volumes ( CV ) of buffer A containing 0 . 5% ( v/v ) TX100 and incubated at 4°C during 1 h on a rotating wheel . The flow through was discarded and the beads were washed with 10 CV of buffer A containing 0 . 3% ( v/v ) TX100 . ApoE4 was released by incubation of the chitin beads with buffer A containing 30 mM dithiothreitol ( DTT ) at 4°C during 40 h and finally eluted with 3 CV of buffer A containing 5 mM DTT . ApoE4 was then buffer exchanged against buffer B ( 20 mM ammonium bicarbonate , pH 8 ) with PD-10 desalting column ( GE healthcare ) and lyophilized overnight . Prior to utilization , lyophilized apoE4 was solubilized in buffer C ( 20 mM HEPES , 150 mM NaCl , pH 7 . 4 ) containing 6 M guanidine-HCl and further purified by size exclusion chromatography on a Superose 6 matrix ( GE Healthcare ) eluted with buffer C containing 4 M guanidine-HCl . Fractions containing apoE4 were pooled together and dialyzed against buffer B during 48 h at 4°C . ApoE4 concentration and purity were assessed by absorbance at 280 nm and SDS-PAGE . ApoE4 rHDL were formed using 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine ( POPC , Avanti polar lipids ) following a modified version of the protocol initially developed for apoA-I by Matz and Jonas [43] . POPC solubilized in chloroform was dried under nitrogen and resuspended to a concentration of 20 mg/ml in buffer C . Sodium cholate was added at a POPC:sodium cholate molar ratio of 1:2 and the mixture was sonicated for 1 h with vortexing every 15 min . ApoE4 was added to the mixture at different molar ratio and incubated overnight at 4°C on a rotating wheel . Sodium cholate was eliminated by dialysis during 24 h against buffer B ( 3 buffer exchange ) . Samples were purified on a Superose 6 matrix eluted with buffer C . Fractions containing apoE4 rHDL were pooled together and concentrated by filtration up to an apoE4 concentration of 0 . 5 mg/ml ( Vivaspin 6 , 50 K MWCO , Sartorius ) . The homogeneity and the size distribution of the rHDL were both assessed by blue native PAGE ( S1 Fig ) while their apoE4 and POPC content was evaluated by measuring the concentration of proteins and lipids by absorbance at 280 nm and phosphorus assay [65] , respectively . Blue native PAGE ( 3 . 5–13% ) electrophoresis was realized following a procedure described in [66] using HMW Native Marker Kit ( GE Healthcare ) for protein standards . SDS-PAGE ( 8% ) was realized according to [67] using prestained protein ladder ( Fermentas ) as molecular weight size marker . Both blue native PAGE and SDS-PAGE were revealed using Coomassie blue staining . Infrared spectroscopy was performed in attenuated total reflection mode and infrared spectrum was recorded on an Equinox 55 spectrophotometer ( Bruker Optics ) . Measurement was made at room temperature by spreading 2 μL of apoE4/POPC rHDL solution ( 0 . 5 mg/ml ) on the surface of the internal reflection element made of a diamond crystal . Excess water was removed under nitrogen flow . The spectrum represents the mean of 256 spectra recorded at a 2 cm-1 resolution . Data were analyzed using Kinetics software ( SFMB , Brussels , Belgium ) and processed for baseline correction and subtraction of the water vapor contribution . Curve-fitting on the non-deconvoluted spectrum was performed to determine the global secondary structure content of a protein . The proportion of a particular structure is computed to be the sum of the area of all the fitted bands having their maximum in the frequency region where that structure occurs divided by the total area of the amide I band between 1700 and 1600 cm-1 . They were chosen by the program on the basis of the shape of the most deconvoluted spectrum ( α-helices and random coil absorb at 1637–1662 cm-1 , turns at 1662–1682 cm-1 and β-sheets at 1613–1637 cm-1 and 1682–1689 cm-1 ) . TEM data were collected at a nominal magnification of 60 , 000 and pixel size 1 . 9 Å/pix on a JEM-1400 ( JEOL ) operating at 120 kV with a LaB6 filament and equipped with a CMOS TemCam-416 4016x4016 camera ( TVIPS ) . For NS-TEM , grids were glow discharged system and 2 μl of sample at concentration of 0 . 01 mg/ml was applied and stained with uranyl formate . Images were collected at defocus between 1 and 2 μm . A total of 6 , 766 particles were picked manually . The set of particles was first classified with multiple cycles of k-means classification and multi-reference alignment using SPARX with the exclusion of particles less representative . In complement , a second set of classifications was performed using EMAN2 [68] resulting in 25 class-averages corresponding to 1188 symmetric top-view particles . For cryo-TEM , a frozen-hydrated grid was prepared by blotting 2 μl of sample ( 1 mg/ml ) on a Quantifoil holey carbon-film-coated 400-mesh copper grids and plunge-frozen . From a total of 660 images , with an average dose of 41 electrons/Å2 , 29 , 134 particles were carefully manually selected . The particles were classified using EMAN2 . The final representative class-averages were calculated from 10 , 249 particles . XL-MS analysis was carried out essentially as described elsewhere [49] . Briefly , an 8-fold molar excess of DSS ( Creative Molecule Inc . ) over apoE4 concentration was added to the apoE4 nanodiscs . The mixture was incubated for 30 min at 37°C and the XL reaction was quenched by the addition of ammonium bicarbonate to a final concentration of 50 mM for 10 min at room temperature . The products resulting from the XL reaction were separated by SDS-PAGE and visualized with Coomassie blue staining . Bands containing the cross-linked species of interest were sliced from the gel into cubes of 1 mm3 , transferred into protein low binding tubes ( Eppendorf ) and submitted to in-gel digestion [69] using trypsin ( Promega ) . MS analysis was carried out on a Thermo Orbitrap Elite mass spectrometer ( Thermo Scientific ) and data analysis was performed using xQuest [48] . False discovery rates ( FDR ) and delta score ( deltaS ) of cross-linked peptides were assigned using xProphet [46] . Cross-linked peptides that were identified with an assigned FDR below 5% and a deltaS below 95% were selected for this study ( S1 Table ) . In all cases the FDR , which denotes the false-discovery rate as calculated by xProphet [46] , was equal to zero . All selected XLs were further analyzed by visual inspection in order to ensure good matches of ion series on both cross-linked peptide chains for the most abundant peaks . Lys146 was not detected as a XL site . This most likely resulted from trypsin digestion producing a dipeptide which is too short to be considered by our applied XL-MS method in order to ensure good matches of cross-linked peptides . Based on intramolecular XLs and information on the shape/organization of the apoE4 nanodiscs , the structure of an apoE4 monomer surrounding an implicit POPC disc was modeled with CNSsolve [70 , 71] . The modeling procedure is described in Supporting Information S1 Text . All molecular dynamics calculations were performed in the isothermal-isobaric ensembles at 300 K with the program NAMD2 . 9 [72] . The CHARMM 27 force-field [73 , 74] with CMAP corrections [75] was used for protein , water and ions and a united atom force field [76] described the lipid molecules . The protocols of the molecular dynamics simulations and of the molecular dynamics trajectory analysis are described in the Supporting Information S1 Text . LDLr LA repeats LA4 ( res . 179–214 ) and LA5 ( res . 121–167 ) , as extracted from the first and representative conformation in the NMR structure ( PDB code 2LGP ) [77] , were docked individually to a part of apoE4 ( res . 125–185 ) containing the elongated helix 4 extracted from one molecule of apoE4 in the head-to-head opened hairpin model using the Haddock web server [78] . LA4/LA5 repeats were docked individually ( and not together ) as the loop between the two is highly flexible . In accordance to an earlier docking study [79] the LA5 module was docked to apoE4 using four unambiguous constraints ( S4 Table ) . For the docking of the LA4 module to apoE4 ambiguous constraints were used in accordance to mutational data [51] and structural data of LA4 repeat bound to other proteins [80–82] ( S4 Table ) . For the individual dockings of LA5/apoE4 and LA4/apoE4 , 4 and 40 poses were obtained respectively . Each possible combination of LA4/LA5 repeats pairing was screened for the distance between the backbone carbon atom of residues Y167 ( LA4 ) and F179 ( LA5 ) . The shortest distance between these residues is about 35 Å ( S6 Fig ) , a value in good agreement with the loop length between LA4 and LA5 repeats ( 12 residues ) .
|
Among the proteins involved in the transport of lipids and their distribution to the cells , apolipoprotein E ( apoE ) mediates the internalization of cholesterol rich lipoproteins by acting as a ligand for cell-surface receptors . In the central nervous system , while apoE is the major cholesterol transport protein , a dysfunction of apoE in the transport and metabolism of lipids is associated with Alzheimer’s disease . A molecular understanding of the mechanisms underlying the receptor binding abilities of apoE is crucial to address its biological functions , but is so far hindered by the dynamic and complex nature of these assemblies . We have designed an original hybrid approach combining experimental data and bioinformatics tools to generate high resolution models of lipidated apoE . Based on these models , we can propose how apoE adapts its conformation at the surface of lipid nanoparticles . Further , we propose a novel mechanism of regulation of the activation and receptor recognition of apoE that could prove valuable to interpret its role in Alzheimer and apoE-related cardiovascular diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"pattern",
"recognition",
"receptors",
"medicine",
"and",
"health",
"sciences",
"molecular",
"dynamics",
"immunology",
"lipoprotein",
"receptors",
"lipid",
"structure",
"immune",
"system",
"proteins",
"lipids",
"proteins",
"lipoproteins",
"chemistry",
"physics",
"biochemistry",
"biochemical",
"simulations",
"signal",
"transduction",
"molecular",
"structure",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"immune",
"receptors",
"physical",
"sciences",
"computational",
"chemistry",
"chemical",
"physics",
"computational",
"biology"
] |
2018
|
Lipidated apolipoprotein E4 structure and its receptor binding mechanism determined by a combined cross-linking coupled to mass spectrometry and molecular dynamics approach
|
Bacterial whole genome sequencing holds promise as a disruptive technology in clinical microbiology , but it has not yet been applied systematically or comprehensively within a clinical context . Here , over the course of one year , we performed prospective collection and whole genome sequencing of nearly all bacterial isolates obtained from a tertiary care hospital’s intensive care units ( ICUs ) . This unbiased collection of 1 , 229 bacterial genomes from 391 patients enables detailed exploration of several features of clinical pathogens . A sizable fraction of isolates identified as clinically relevant corresponded to previously undescribed species: 12% of isolates assigned a species-level classification by conventional methods actually qualified as distinct , novel genomospecies on the basis of genomic similarity . Pan-genome analysis of the most frequently encountered pathogens in the collection revealed substantial variation in pan-genome size ( 1 , 420 to 20 , 432 genes ) and the rate of gene discovery ( 1 to 152 genes per isolate sequenced ) . Surprisingly , although potential nosocomial transmission of actively surveilled pathogens was rare , 8 . 7% of isolates belonged to genomically related clonal lineages that were present among multiple patients , usually with overlapping hospital admissions , and were associated with clinically significant infection in 62% of patients from which they were recovered . Multi-patient clonal lineages were particularly evident in the neonatal care unit , where seven separate Staphylococcus epidermidis clonal lineages were identified , including one lineage associated with bacteremia in 5/9 neonates . Our study highlights key differences in the information made available by conventional microbiological practices versus whole genome sequencing , and motivates the further integration of microbial genome sequencing into routine clinical care .
Primary and nosocomial bacterial infections are a major source of morbidity and mortality worldwide [1] . In the United States alone , community-acquired bacterial infections represent a leading cause of death , particularly among the elderly [2 , 3] , while healthcare-associated bacterial disease affects approximately 2 million patients annually and results in nearly 100 , 000 fatalities [4 , 5] . Bacterial infectious disease is especially pervasive among patients undergoing intensive medical care , who are predisposed to disease because they are frequently immuno-compromised [6] , have undergone invasive procedures , and/or have long-standing placement of foreign bodies ( such as indwelling catheters and mechanical ventilators ) [7] . Furthermore , effective treatment of bacterial infections has been complicated by the emergence and dissemination of multi-drug resistant strains [8–10] , which can colonize healthcare environments and persist in them over time [11 , 12] . Although predominantly restricted to research studies at present , bacterial genome sequencing of clinical isolates is becoming increasingly tractable as the cost of DNA sequencing continues to decline [13 , 14] . Bacterial genome sequencing has shed light on diverse aspects of a number of medically-relevant pathogens , including global and intra-hospital strain transmission [9 , 15–19] , the evolution of antibiotic resistance [15 , 20 , 21] , horizontal exchange of antibiotic resistance plasmids within hospitals [11] , the distribution of genomic material across a bacterial species [22 , 23] , and the identification of novel bacterial pathogens [24] . Although such studies have provided important insights and an improved understanding of medically important bacterial pathogens , in general , nearly all have been limited in one or more respects: they have been retrospective in nature , limited in scope to particular bacterial species , and/or have extended to a subset of bacterial isolates ( e . g . a recognized outbreak ) , rather than globally across an entire patient population . These limitations impose an artificially narrow and biased perspective on the repertoire of infectious organisms affecting patient populations [25] . As a different approach , we sought to perform prospective collection and whole genome sequencing of all bacterial isolates obtained from the intensive care units of a single tertiary care hospital over a period of one year . This collection of 1 , 229 microbial genomes enables quantification of several features of infectious bacteria within the patient population of a single hospital , including genetic diversity , the prevalence of previously unreported bacterial species , and large-scale , longitudinal reconstruction of both intra-patient and inter-patient molecular epidemiology .
This study was conducted from August 2012 to August 2013 at the University of Washington Medical Center , a 450-bed hospital in Seattle , Washington , that serves as a tertiary care center for a geographic region comprising 5 states . Bacterial isolates were obtained during the course of routine medical care of patients in the hospital’s three intensive care units ( neonatal , medical/surgical , and cardiac ) . All individually characterized bacterial isolates ( i . e , isolates receiving a species-level , genus-level , or descriptive classification ) that were reported in patients’ laboratory results were collected for sequencing . A total of 1 , 316 viable isolates obtained from 421 individual patients were subjected to shotgun whole genome sequencing . Libraries from 87 isolates failed quality control , leaving in a final collection of 1 , 229 isolates obtained from 391 different patients ( 222 males and 169 females ) ( S1 Data File ) . The average age of patients in the neonatal intensive care unit was 34 days at the time of sampling ( range 0–109 days ) , while the average age of patients occupying other units was 56 . 6 years ( range 17–94 ) . The sequenced isolates originated from a variety of anatomic locations , although cultures of broncheoalveolar lavage fluid , sputum , blood , urine , and wounds collectively accounted for 80% of sampling sites ( S1 Fig ) . Because sampling locations included normally non-sterile sites , in some cases our collection represented both known bacterial pathogens and normal microbiota sampled incidentally . In accordance with the laboratory’s testing practices , we considered an isolate to be clinically relevant if there was an indication necessitating antibiotic susceptibility profiling , either as part of the clinical laboratory’s reflex testing algorithms or by special physician request . Conventional clinical microbiology workup resulted in a species-level classification for nearly half of the bacterial isolates ( 610/1 , 229 ) , the majority of which corresponded to well-characterized bacteria such as Escherichia coli and Staphylococcus aureus . The remaining isolates were classified at the genus level ( ie , Lactobacillus sp . ) , or on the basis of morphological and/or metabolic characteristics ( for example , “lactose-fermenting gram negative rod” ) , according to recommended practice guidelines [26] . Altogether , the harvested organisms encompassed 83 distinct taxonomical or functional classifications ( S2 Data File ) as determined by routine clinical laboratory testing . To enable more precise taxonomic classification of isolates sampled from the clinical wards , we performed de novo genome assembly for each isolate and identified the closest matching ( partial or complete ) genome present in the NCBI nucleotide database . Using this strategy , we identified 126 different species from 60 different genera ( S2 Data File ) , providing a highly resolved catalog of the organisms collected from the ICU system ( Fig 1 ) . For many clinical classifications not extending to the species-level , it is perhaps unsurprising that a genomic approach readily uncovers heterogeneity not represented through existing bacterial typing algorithms , and more definitively establishes organism identity . Because our isolates were derived from otherwise unselected primary clinical material , we hypothesized that our collection could contain organisms , potentially including medically significant organisms , which are not well represented in existing curated repositories . In order to explore the prevalence that novel species were represented within the primary clinical isolates , we calculated pairwise average nucleotide identity by BLAST ( ANIb ) [27] between each isolate and its closest identified match from NCBI . Pairwise ANIb values of less than 95% are generally accepted as the cutoff for circumscribing separate species [27 , 28] . Species demarcations based on this definition of genomic similarity are known by the technical definition of ‘genomospecies’ , indicating that they have not been established on the basis of identifying phenotypic or morphological characteristics . Pairwise ANIb values for our isolate collection ranged from 71–100% , with 35% ( 428/1 , 229 ) of all isolates falling below a 95% identity threshold to its closest matched genome , and thereby qualifying as novel genomospecies ( S2 Fig ) . Such organisms were isolated from all culture sites ( S1 Fig ) , suggesting that they were not restricted to heavily colonized anatomic locations such as the gut and skin . Although none of the classifications by the clinical laboratory were overtly different from the closest-matched genome identified for any given isolate , 12% ( 76/612 ) of isolates receiving a species-level classification displayed <95% ANIb to the specified taxon , and thus represented a genomospecies distinct from the given clinical classification . Novel genomospecies were unevenly distributed across bacterial genera; at the extremes , all isolates from the genera Escherichia , Lactobacillus , and Proteus were grouped with a previously reported genome , whereas all isolates from the genera Neisseria and Corynebacterium fell below a 95% ANIb identity to previously reported genomes ( S3 Fig ) , likely reflecting the high degree of genomic diversity inherent to those lineages [29 , 30] . We next evaluated groups of novel genomospecies with respect to their genomic similarity to known organisms . We performed an all-by-all computation of pairwise ANIb values among sequenced isolates , built a graph from these data , and clustered like isolates in accordance with their pairwise ANIb scores ( Fig 2A ) . The 1 , 229 isolates fell into a most likely configuration of 78 distinct groups ( indicated by the maximal intra-cluster enrichment metric [31] , S4 Fig and S3 Data File ) , defined by chains or clusters of 2 or more isolates linked with values of ≥95% pairwise ANIb , and an additional 29 isolates were left unclustered as singletons . This number of categories is somewhat less than the number of species obtained by conventional or genomic classification and suggests that our “de novo” clustering approach provides a more conservative estimate of the number of genomospecies ( S2 Data File ) . Pairwise comparisons of isolates in the same group showed significantly ( p< 0 . 0001 , Mann-Whitney test ) higher ANIb values than those in different groups ( 97% ANIb versus 43% ANIb , respectively ) , and a clear separation of these categories around the 95% ANIb threshold , supporting the validity of clustering results ( S6 Fig ) . Fourteen ( 18% ) non-singleton clusters were comprised entirely of isolates classified to a previously described species , that is , they contained exclusively organisms with ≥95% ANIb to representatives of that species; 22 clusters ( 29% ) contained ≥50% isolates matching a particular known species ( with a minority of isolates qualifying as a novel but related genomospecies ) ; in another 12 clusters ( 15% ) the majority of isolates were novel genomospecies , with a minority assigned to known taxa . Finally , 30 clusters ( 39% ) were made up entirely of isolates qualifying as novel genomospecies , although many of these ( 19/30 ) were composed of only two or three isolates each . 76% ( 22/29 ) of unclustered isolates were novel genomospecies . In order to best quantify the number of novel genomospecies represented in the collection , we repeated clustering using only those isolates that did not match previously sequenced genomes with ≥95% ANIb ( Fig 2B ) . These distributed among 53 separate clusters ( S5 Fig and S4 Data File ) and 34 singletons , conservatively suggesting that our survey has identified 87 novel taxonomic groups . Of those categorizations , 32% ( 28/87 ) contained at least one isolate with an indication for antibiotic susceptibility testing , and so harbored a fraction of isolates which satisfied our definition for clinical significance . Of note , 20 of these 28 groups contained two or more representatives obtained independently from more than one patient , corroborating their significance in multiple contexts and individuals . Existing pan-genome analyses have not included large numbers of primary clinical isolates . To measure genomic diversity observed within established clinical pathogens , we assessed the pan-genome content of the 20 most prevalent species cultured from our patient population ( range: 6 isolates for Moraxella catarrhalis to 162 isolates for Staphylococcus epidermidis ) . For each species , we considered the number of unique genes represented in existing reference genomes and subsequently factored in predicted genes from the sequenced clinical isolates . We restricted analysis to isolates with >95% ANIb to established reference genomes in order to avoid inflating these estimates . For all bacterial species examined we added additional genes to the pan-genome with each clinical isolate sequenced ( Fig 3A and S5 Data File ) . This analysis indicates that common clinical pathogens all display “open” pan-genomes with versatility in gene content [32] , albeit to varying degrees . Relatedly , across taxa we consistently identified additional unique sequences not represented in reference genomes , indicating that even in well-studied pathogens , there remains a reservoir of uncatalogued genomic variation harbored by clinical isolates ( Fig 3B ) . Considering the collection as a whole , there was substantial variation between species with respect to the overall size of the pan-genome , the number of unique sequences that were discovered outside of existing reference genomes , and the current rate of gene discovery ( Fig 3C ) . There was a weak relationship between the number of isolates sequenced and the rate of gene discovery ( R2 = 0 . 129 ) . Perhaps unsurprisingly [33] , Escherichia coli had the largest pan-genome size of the species analyzed with 20 , 432 non-homologous genes , while Moraxella catarrhalis contained the fewest , with 1 , 420 . At the conclusion of this study , new genes were discovered in Rothia mucilaginosa at a rate of 152 per additional isolate sequenced ( 11 clinical isolates sequenced ) , making its pan-genome the least explored of the organisms included in our sample set . At the other extreme , only 1 new gene was discovered with each additional S . aureus isolate sequenced ( 108 clinical isolates sequenced ) , suggesting that there is relatively little unexplored distributed genetic content in our collection . We explored the molecular epidemiology of bacterial infections within the hospital , investigating both the possibility of bacterial transmission events and the dynamics of infection within individual patients . To ensure sufficient genomic similarity for accurate mapping and variant calling , we again restricted analysis to isolates with >95% ANIb to established reference genomes . Initially , we sought to quantify the level of artifactual genomic variation arising from our protocols . We randomly selected sixteen isolates distributed across 15 different species , and for each , paired technical replicates were separately taken through library construction , sequencing , and analysis . No discordant single nucleotide variants ( SNVs ) were identified between any replicate pair . Because strains may undergo genomic diversification during the course of an epidemic [34–36] , or even during infection in a single patient [37–39] , establishing a quantitative definition for the number of genomic polymorphisms which identify isolates as members of the same outbreak remains an unresolved challenge in molecular epidemiology [34] . Indeed , accurately reconstructing transmission chains in light of this so-called “cloud of diversity” will require defining new practices in molecular epidemiology , including examination of multiple isolates from the same individual [35 , 39] . Given these issues , here we defined strain clonality ( relationship by direct descent ) using two different thresholds: A] low stringency: isolates distinguished by up to 40 SNVs , a threshold based on empiric measurements of within-host bacterial genomic variability [40] , and B] high stringency: isolates distinguished by no more than 3 SNVs [41] , based on the variability of genomic sequencing in other studies . It is expected that the low stringency threshold encompasses potentially indirect transmission events , such as those occurring through intermediate reservoirs such as the environment or asymptomatic colonized hosts , where lineages have had time to accumulate some number of differentiating genomic differences [39 , 40] , whereas the high stringency threshold is more likely to represent direct transmission of bacterial clones [41] . A key point is that both of these cutoffs fall vastly below the average pairwise distance that is expected between any two randomly selected isolates of the same species from the larger community ( S1 Table ) . We considered isolates to be members of the same clonal lineage if they were linked either directly or through intermediate connections of equal to or less than these thresholds .
It has been posited that large-scale , whole genome sequencing of clinical bacterial isolates holds promise as a transformative technology in the practice of clinical microbiology [9 , 13 , 44 , 45] but the technology has not yet been systematically applied in a clinical context to explore what information can be readily obtained from using this approach . In one noteworthy study , all 130 bacterial isolates recovered from a clinical microbiology laboratory over a single day were subjected to whole genome sequencing [46] , although the analytic scope of that effort was limited to taxonomical classification of the organisms . Other , recent work has explored the prospective [19] and “real-time” [47] use of bacterial whole genome sequencing to define the molecular epidemiology of suspected disease outbreaks . However , to the best of our knowledge , our study is unique in that we prospectively performed large-scale , unbiased collection of all bacterial isolates from a defined set of hospital units over an extended time period and performed genomic sequencing and analysis . Consequently , this project reveals a number of telling differences between the information obtained by existing microbiological practices and what can be learned from genomic analysis of clinical bacterial isolates . First of these is the straightforward task of classifying an isolate within the bacterial taxonomy . In standard clinical practice , taxonomic classification of bacterial species has long relied on phenotypic characteristics such as Gram stain , colony morphology , and biochemical characterization . The degree to which particular isolates from a specimen are assigned a species-level classification is designed , in part , to avoid reporting members of the normal microbiota or other organisms that are non-contributory to a disease process , and thus discouraging antibiotic “overtreatment” by physicians [26 , 48] . The reporting of species-level classifications therefore depends on the availability of microbiological methods for classifying an organism , the clinical indication for culture , the anatomic sampling site , and special requests of the ordering provider . Consequently , bacterial isolates from a specimen are neither typed comprehensively nor all typed to the same level of taxonomic resolution , and some species-level identifications made in the clinical laboratory are not reported to providers . Although the current utility to clinical practice is debatable , with whole genome sequencing it is possible to readily and unilaterally classify isolates on the basis of sequence similarity to other reported draft or reference genomes , providing improved granularity and eliminating ambiguities presented by genera or group-level assignments ( i . e . , “Coagulase negative Staphylococcus” ) . Although no overt inconsistencies were observed between isolates’ genomic classifications and conventional taxonomic assignment , whole genome sequencing provided substantially greater taxonomic resolution than standard practices ( Fig 1 ) . Additionally , a sizable fraction of all isolates ( 35% ) corresponded to novel genomospecies upon genomic analysis , including 12% of the isolates that were assigned a species-level taxonomic classification by conventional methods . Although somewhat surprising , this finding is compatible with earlier , 16S rRNA-based typing studies wherein a significant fraction of infections were caused by previously uncharacterized organisms [49] that conventional microbiology practices are incapable of resolving . Further characterizing these organisms and exploring their prevalence and potential pathogenicity will be important work for future studies . The flexible genome content of bacteria is a major contributor to their pathogenicity [14 , 22 , 50] , and investigation of the pan-genomic content of a species can therefore provide insights into aspects of biology relevant to pathogenesis . All pan-genomes examined in this study ( Fig 3 ) increased predictably with additional isolates sequenced , suggesting that additional genomic content will be identified with continued sequencing of primary clinical isolates . Further , these results suggest that the availability of high quality reference genomes is not necessary to explore genomic content and to identify potentially important genes relevant to human disease . Properties of pan-genomes varied considerably across species , particularly with respect to the rate of gene discovery per additional strain sequenced . Staphylococcus aureus clinical isolates in our collection had the lowest rate of novel gene discovery , suggesting that clinical lineages are unlikely to experience major phenotypic shifts owing to the acquisition of new functions coded by external genes . The finding of an essentially “closed” pan-genome for S . aureus strains is surprising in light of previous reports of extensive genomic variation for that species [51] , and likely reflects the relative clonality of the predominant clinical strain [52] . In contrast , organisms such as Klebsiella pneumoniae and E . coli display genomic content that is broadly distributed across its members , signifying an “open” pan-genome [11 , 23] . The observation that pangenome size for those species does not dramatically plateau even with large numbers of additional strains sequenced supports the notion that their pangenomes are actively evolving [23 , 53][14] , and that discoveries in additional genomic diversity may be sustained indefinitely . With respect to intra-patient infection dynamics , we identified a measurable frequency of polyclonal bacterial infections ( 30 . 5% of isolates from a given species that were recovered from a patient within 20 days ) . However , since it is not clinical practice to isolate phenotypically indistinguishable isolates of the same species , this is almost certainly an underestimate secondary to incomplete representation in our isolate collection of all lineages involved in particular infections . Relatedly , although it is current microbiological practice to characterize the antibiotic profile of phenotypically distinct representatives of the same species ( typically , those with different colony morphologies ) , our data indicate that in some cases typing multiple isolates from a particular species could yield significantly different results . In contrast , the majority ( 17/32 ) of persistent bacterial infections , recovered from a patient over 20 days or more , were marked by longitudinal recovery of the initial infecting clone . These findings suggest that many cases of long-lasting or chronic infection reflect continuous colonization of the original infectious strain , or autoinoculation from untreated bodily sites . The most unanticipated results of our study are from molecular epidemiological analysis of bacterial clonal lineages , where we find data consistent with direct ( 0–3 SNV differences among isolates ) and less direct ( up to 40 SNV differences among isolates ) bacterial transmissions involving multiple patients in these ICUs . The dynamics underlying bacterial colonization and transmission are surprisingly complex [34 , 39 , 40] , and circumscribing transmission events according to specific thresholds of strain relatedness is certainly an oversimplification . For example , this strategy could exclude transmission of hypermutator strains separated by an unusually high mutational burden [54 , 55] , or those spread by individuals over the course of long-term carriage [39] . In this respect , our analysis may be considered somewhat conservative , but is appropriately restrictive given the unbiased sampling employed in our study . Though the importance of nosocomial transmission of bacteria has long been a major public health concern [5 , 56] , whole genome sequencing studies by several groups have recently shown that the true incidence of nosocomial transmission is quite rare for several common pathogens of major concern [14 , 16 , 18 , 57 , 58] . Our findings support the conclusions of those studies when overlapping the specific pathogens that have been previously surveyed; however , evidence is provided in our study for the inter-patient sharing of several opportunistic pathogens: S . epidermidis , E . faecium , and P . aeruginosa , and to a lesser extent , E . faecalis , S . aureus and S . maltophilia ( Fig 4 ) . The most notable of these clonal lineages affected the NICU , and involved at least seven different S . epidermidis clonal lineages ( 3 involving pairs of patients and 4 spanning 3 or more patients ) . Because neonates are initially colonized with microbiota both from the maternal genital tract [42] and microbes from the external environment [59] , they may be especially susceptible to colonization with transmitted agents compared to patients with more established microbiomes . A substantial fraction of the observed clonal lineages were associated with bacterial disease that required clinical treatment ( 44/71 patients ) . However , the sharing of closely related strains among patients was not recognized nor further investigated in the absence of genomic data . There are several explanations for this . First , except in rare cases , bacterial characterization of coagulase negative Staphylococcus ( other than Staphylococcus lugdunensis ) was not performed to the species level , so that identification of recurrent S . epidermidis infections would not have been recognizable . Secondly , many of the clonal lineages were transmitted over extended periods of time ( up to nearly a year ) and in some cases without any temporal overlap of affected patients , making it difficult to link the cases on the basis of patient biogeographical information . All organisms involved are known to be transmitted by person-to-person interaction as well as via intermediate colonization of the hospital environment itself [10 , 17 , 58 , 60–63] . Thus , in addition to potential environmental reservoirs that remain unexplored in our study , it is possible that some of these events were mediated by assimilation of clonal lineages into the normal microbiota of individuals of the hospital staff without causing disease ( “cryptic” or “silent” transmissions [9 , 17] ) , and inadvertently transmitted to patients over time through the circulation of medical staff among the different units . It is possible that key epidemiological links between individuals were not captured by , or represented in , our study . Third , the organisms involved all carry the distinction of being both commensal organisms and opportunistic pathogens , and consequently , autoinfection of a patient with their resident microbiota would be the most likely explanation for any isolated case . We emphasize that , although our analysis reveals isolates that are likely to be related by descent , it is unclear how and when transmission of clonal lineages occurs . For example , it is possible that some transmissions reflect indirect infection or colonization events occurring outside the hospital from endemic , clonal bacterial pools present in the community [18 , 64] . However , both the recovery of several genomically indistinguishable clones from multiple patients , and the association between shared clonal lineages and overlapping patient admission dates is strongly suggestive , and merits future investigation . The modes of transmission and the potential reservoirs of these transmitted organisms must also be investigated through future work that more comprehensively surveys occupants of a hospital system , possibly extending to medical staff and patient family members . Regardless , it is clear that genomic analysis will be necessary to identify cryptic sharing of clonally related microbes , and to more fully illuminate the population biology of bacterial infections observed in the hospital setting . Routine , unbiased , and large-scale sequencing of bacterial clinical isolates has the power to reveal unknown and unsuspected properties about bacterial infectious disease , and is becoming increasingly feasible with sustained advancements in massively parallel sequencing technologies . Comprehensive sequence information , derived from all bacterial isolates , has the capability to transform hospital practices not only within clinical microbiology laboratories but also by directly informing patient care in the form of infection control and treatment practices . Even when performed on a large scale , whole genome sequencing of bacterial isolates may prove cost effective in healthcare practice , considering the financial savings that would accompany potentially reduced patient morbidity and mortality . Although it will take some time to fully explore the power of this approach in the clinical identification and management of bacterial infectious disease , we have demonstrated that there are practical applications to be realized through the immediate application of large-scale , unbiased sequencing of clinical bacterial isolates .
Primary isolation and characterization of bacterial isolates was performed by the University of Washington Medical Center Clinical Microbiology Laboratory according to routine laboratory practices [26] . Antibiotic susceptibility testing was performed using a combination of the TREK Sensititre system ( TREK Diagnostic Systems ) and E-test ( bioMerieux SA ) , depending on the organism isolated , when clinically indicated . Two mycobacterial isolates ( Mycobacterium tuberculosis and Mycobacterium avium ) were excluded from the study . Primary bacterial isolates were collected from primary diagnostic culture or subculture plates , and grown overnight in liquid cultures of Luria-Bertani medium or streaked for confluence and grown as a lawn on appropriate solid media , depending on nutritional requirements . Organisms reported as part of aggregated populations ( for example “mixed Gram positive flora” ) were not analyzed . Use of microbiological specimens and patient chart review was approved by the University of Washington Human Subjects Review Board ( approval number 42541 ) and was conducted in accordance with the Declaration of Helsinki . As a minimal risk study utilizing surplus microbiological isolates , a waiver of consent was approved , as it was not possible to contact all subjects associated with the isolates nor feasible to obtain consent from the study population , and the waiver of consent was not deemed to adversely affect the rights and welfare of the subjects . Results of this study research study were not directly communicated to clinical providers or used as actionable information . After the completion of isolate collection , sequencing libraries were constructed in serial batches of ~192 organisms each , using the same lots of reagents in order to limit batch effects . Sequencing libraries were prepared as described elsewhere [30] , with the addition of a size selection step to enrich for library fragments 400–900 bp in size . Sequencing was performed using Illumina HiSeq 2000 and Illuimina MiSeq platforms with 101 bp paired end reads . Adaptors were trimmed and PCR duplicates removed using the program Fastq-Mcf ( http://code . google . com/p/ea-utils/ ) with skew filtering disabled and other parameters at default . Draft genomes were assembled using AbySS v1 . 3 . 5 [65] , with k-mer values empirically optimized to maximize the N50 statistic ( the length for which all contigs of equal or larger size contain half the sum of the entire assembly ) of assemblies ( S1 Data File ) . To perform species-level classification of isolates , contigs from each isolate were BLAST [66] searched against bacterial genomes in the NCBI non-redundant nucleotide database ( NT , accessed 4-15-14 ) . The best BLAST hit ( as measured by e-value ) for each contig matching a partial or complete genome was recorded , and the cumulative length of best-matched contigs was recorded for each organism: the organism with the greatest cumulative length of best matched designated as the closest match . Estimated read depth per strain was calculated by the Lander-Waterman method [67] , assuming the length of the closest-matched genome . Genomes with less than 7X estimated average read depth were discarded as a quality control measure . Average Nucleotide Identity by BLAST ( ANIb ) values were calculated for draft genomes using the jSpecies algorithm [27] as implemented through a standalone script ( https://github . com/widdowquinn/scripts/blob/master/bioinformatics/calculate_ani . py ) . Isolates were considered to be from a novel genomospecies if they exhibited <95% ANIb when compared against their closest-matched genome , as identified above . Clustering of like organisms was performed on the basis of ANIb values using an agglomerative hierarchical clustering approach , both for the entire collection of isolates and for isolates from novel genomospecies . ANIb values were first calculated for all bacterial isolates to generate an all-by-all comparison matrix . We then created a network in which each node corresponded to a bacterial isolate , and the edges between nodes were given weights equal to the isolates’ pairwise ANIb values , up to a maximum of 0 . 95 . We next applied agglomerative hierarchical clustering as described elsewhere [31] . To determine the optimal number of clusters , which corresponds to an estimate of the number of distinct species , we calculated the intra-cluster enrichment metric across all cluster numbers from 0 to 150 ( S4 and S5 Figs ) . Individual isolates that were not initially assigned to a cluster were added to existing clusters if they matched an isolate within it with ANIb ≥ 0 . 95 . Clustering images were generated with igraph ( http://www . igraph . org/ ) . Some isolates were left unclustered because they did not connect closely to any other isolates . Analysis was confined to isolates matching an available reference genome with an ANIb value of ≥95% . Single nucleotide variants were called by aligning short read data from sequenced strains against appropriate reference genomes using BWA version 0 . 6 . 1-r104 [68] and SAMtools version 0 . 1 . 18 [69] , discarding reads with mapping quality of less than 10 . To avoid artifacts relating to the inclusion of genomically disparate species , we excluded isolates qualifying as novel genomospecies by ANIb analysis . Single nucleotide variant calling was performed using SAMtools with a haploid genome model and minimum variant frequency of 0 . 5 . Variants supported by fewer than 5 reads or a likelihood score of less than 200 were masked as “unknown” data . All-by-all pairwise distance matrices were constructed using custom perl scripts by comparing sites of variation among isolates , masking sites at which one or both isolates displayed “unknown” data or fewer than 15× read coverage , and counting only those variants sites at which both isolates could be confidently genotyped . Pairwise genomic distances were expressed as the absolute number of passing variant sites which distinguished such pairs . For focused mutational analyses , variants were annotated using snpEFF version 4 . 1G [70] . Draft genome sequences and available complete genomes from NCBI were combined for pan-genome analysis as described elsewhere [14] , with minor modifications . Briefly , gene predictions were made using RAST version 4 . 0 [71] . A “meta-reference” was constructed to represent all unique coding sequences ( CDS ) in all strains , and were clustered using CD-HIT v4 . 6 [72] to de-duplicate proteins ≥80% identical . Putative phage and insertion sequences were identified by BLAST search against a prophage database as described [73] and eliminated . BLASTX and BLASTP were used to search de novo assemblies and complete genomes , respectively , against the meta-reference , and a CDS was considered present in an isolate if ≥80% of the CDS was covered by an alignment and protein-level identity was ≥80% . Sequences of ≤75 amino acids in length and assemblies with an N50 statistic of <8x103 bp were excluded from pan-genome analysis . 1 , 000 different random input orders of genomes were performed using the specaccum function in the vegan package implemented in R 3 . 1 . 1 , and standard power functions with offsets were fit to resultant data using pyeq2 ( http://code . google . com/p/pyeq2/ ) to enable imputing gene diversity for different numbers of genomes sequenced .
|
Bacterial whole genome sequencing is becoming increasingly common to microbiological research , but despite its great potential , has not yet been meaningfully integrated into clinical care . Here , we generated whole genome sequencing data from nearly all of the bacterial isolates prospectively collected from a hospital’s intensive care units over an entire year . Our analysis identifies novel microbiota in hospitalized patients , a high incidence of patient infection with multiple unrelated lineages of a bacterial species , and the possibility of cryptic transmission of bacteria among patients . Our study is unprecedented in providing a broad and unbiased view of bacterial infections that affect the hospital’s sickest patients , and demonstrates the extent of information that can be learned from comprehensive genomic surveillance of clinical bacterial isolates over an extended period of time .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
A Year of Infection in the Intensive Care Unit: Prospective Whole Genome Sequencing of Bacterial Clinical Isolates Reveals Cryptic Transmissions and Novel Microbiota
|
Lymphoid oncogenesis is a life threatening complication associated with a number of persistent viral infections ( e . g . EBV and HTLV-1 in humans ) . With many of these infections it is difficult to study their natural history and the dynamics of tumor formation . Marek's Disease Virus ( MDV ) is a prevalent α-herpesvirus of poultry , inducing CD4+ TCRαβ+ T cell tumors in susceptible hosts . The high penetrance and temporal predictability of tumor induction raises issues related to the clonal structure of these lymphomas . Similarly , the clonality of responding CD8 T cells that infiltrate the tumor sites is unknown . Using TCRβ repertoire analysis tools , we demonstrated that MDV driven CD4+ T cell tumors were dominated by one to three large clones within an oligoclonal framework of smaller clones of CD4+ T cells . Individual birds had multiple tumor sites , some the result of metastasis ( i . e . shared dominant clones ) and others derived from distinct clones of transformed cells . The smaller oligoclonal CD4+ cells may represent an anti-tumor response , although on one occasion a low frequency clone was transformed and expanded after culture . Metastatic tumor clones were detected in the blood early during infection and dominated the circulating T cell repertoire , leading to MDV associated immune suppression . We also demonstrated that the tumor-infiltrating CD8+ T cell response was dominated by large oligoclonal expansions containing both “public” and “private” CDR3 sequences . The frequency of CD8+ T cell CDR3 sequences suggests initial stimulation during the early phases of infection . Collectively , our results indicate that MDV driven tumors are dominated by a highly restricted number of CD4+ clones . Moreover , the responding CD8+ T cell infiltrate is oligoclonal indicating recognition of a limited number of MDV antigens . These studies improve our understanding of the biology of MDV , an important poultry pathogen and a natural infection model of virus-induced tumor formation .
Virus driven lymphoid oncogenesis is a serious consequence of infection with a wide range of herpes and retroviral pathogens in a variety of hosts . Major lymphoma-associated infections of humans include Epstein Barr virus ( EBV ) and Human T cell lymphotropic virus ( HTLV ) [1] , [2] . With both EBV and HTLV tumor progression is a relatively rare event considering the prevalence of infection and the persistent nature of the virus [2] , [3] . In contrast , Marek's Disease Virus ( MDV ) is a widespread , oncogenic α-herpesvirus infection of chickens which readily causes lymphoid tumors and has immense impact on the poultry industry [4] . The oncogenicity of MDV , combined with the ability to vaccinate against tumor formation make the MDV-chicken system an excellent natural infection model for understanding the biology and treatment of viral induced lymphomas [1] , [5]–[7] . The spread of MDV occurs through the inhalation of infectious particles in dust . After a brief lytic phase in B lymphocytes ( ∼2 to 7 days post infection [dpi] ) , MDV establishes a life-long latent infection in CD4+ T lymphocytes [8] . The life-cycle is completed by transfer of the MDV to the feather follicle epithelium [8] . In susceptible birds , MDV infection leads to a high incidence of CD4+ T cell tumors ( up to 100% ) in a wide range of organs including heart , liver , ovary , testes , lungs and skin [9]–[14] . These CD4+ tumors express high levels of CD30 , a tumor necrosis factor receptor II family member , also over-expressed on human lymphomas with diverse etiologies [5] . MDV latency and tumor formation is dependent upon viral encoded genes such as EcoRI-Q ( meq ) , a c-Jun related molecule [15]–[17] . The penetrance ( up to 100% ) and temporal reproducibility of tumor appearance after infection ( within 3 to 4 weeks ) in susceptible lines of bird raises important questions regarding tumor clonality . These include the clonality of transformed cells in individual sites and between sites where multiple discrete solid tumors are evident in a single individual . The MDV genome readily integrates into the host cell genome particularly at telomeric or sub-telomeric locations [18] , [19] . The profile of MDV integration within the tumor host cell suggested restricted clonality of most Marek's Disease-derived cell lines and cells taken from tumor sites [18] , [19] . Between two and twelve independent integration sites were detected in each sample and the pattern of integration was stable over time in culture . In contrast , analysis of T cell receptor ( TCR ) Vβ family usage in CD30hi cells from primary lymphomas led to the conclusion that MD tumors were polyclonal [10] . During the analysis of MDV integration patterns , samples obtained from a single chicken contained at least two major distinct patterns [19] suggesting at least two independent transformation events . These data coupled with the possibility of favoured sites for MDV integration [e . g . telomeric or sub-telomeric preference , [19]] suggest that a non-viral integration site dependent analysis of clonality would be appropriate . Since the tumors are derived from CD4+ T cells , the clonally expressed T cell receptor ( TCR ) would be an appropriate target for the molecular definition of tumor clonality . The development of successful anti-tumor vaccines against MDV has been critical in poultry production and led to the proposal of utility for MDV as a model for developing vaccines against other lymphoma-inducing viral infections reviewed in [1] , [6] . The vaccines are highly effective at preventing tumor formation but fail to eliminate infection or block transmission over prolonged periods [20] . Periodically , circulating strains of MDV develop enhanced pathogenicity and vaccine break has necessitated the development of different generations of vaccines over the past 50 years [Reviewed in [21] , [22]] . The success of vaccination indicates acquisition of protective adaptive immunity and both antibody and T cell responses are readily detected [23] , [24] . Other evidence for immune protection includes the association of genetic resistance with the MHC ( B locus ) haplotype [25]–[29] . Similarly , natural infection induces measurable natural killer cell , antibody , T cell and cytokine and interferon responses [30]–[34] . The highly cell associated nature of MDV supports the notion that cell mediated responses may predominate in protective immunity ( reviewed by [23] , [35] , [36] with the CD8+ T cell mediated cytotoxic killing demonstrated in several studies [37]–[39] . The cytotoxic activity in MHC B19 and B21 homozygous chickens was focussed on the MDV-encoded pp38 , meq and gB antigens [38] . Importantly , transient depletion of CD8+ T cells rendered chickens more susceptible to infection with MDV [40] . The response to persistent viral infections in humans is often characterised by cytotoxic T cells specific to latency-associated antigens . Indeed , large clones of T cells are readily detected during infection with CMV [41] , [42] and EBV [2] , [43] , [44] . This type of clonal structure within CD8+ T cells is indicative of a response focussed on very few antigens . The issue of tumor clonality and the nature of the CD8+ T cell response during MDV infection prompted application of the T cell receptor repertoire analysis tools we have recently developed for the chicken [45] . The chicken TCRβ locus in chickens is much simpler than in mammals containing 13 Variable ( V ) , 1 Diversity ( D ) , 4 Joining ( J ) segments and 1 C segment [45]–[49] . The Vβ segments group into two families , which simplifies global analysis of the chicken TCR repertoire . We applied a combination of CDR3 length analysis ( spectratyping ) and sequencing of the VDJ-junction ( also known as the complementary determining region 3 [CDR3] ) to define the clonality of MDV cell lines and different populations of cells from tumors or other sites within MDV infected birds . These approaches revealed clonal structure within MDV tumors ( but not always monoclonal ) and a pattern of shared and distinct clonal origin in different sites within a single individual . Analysis of the tumor infiltrating and splenic CD8+ T cells allowed identification of large T cell clones within an oligoclonal framework of responding CD8+ T cells .
Inbred line P ( MHC , B19/19 ) white leghorn chickens were reared pathogen free at the Poultry Production Unit of the Institute for Animal Health . One-day-old birds were infected with of MDV strain RB-1B [50] by intra-abdominal injection of ∼1000 pfu cell associated virus and observed for the development of MD using methods described previously [51] , [52] . Two of the birds ( 15 and 16 ) were sentinel birds and infected by exposure to experimentally infected birds . Birds were reared with ad libitum access to water and vegetable-based diet ( Special Diet services , Witham , UK ) and wing-banded to allow identification of individuals . This study was carried out according to the guidance and regulations of the UK Home Office with appropriate personal and project licences ( licence number 30/2621 ) . As part of this process the work has undergone scrutiny and approval by the ethics committee at the Institute for Animal Health . Single-cell suspensions of lymphocytes were prepared from spleen , blood and tumor tissues by Histopaque-1083 ( Sigma-Aldrich , Steinheim , Germany ) density-gradient centrifugation . CD4+ and CD8+ T cell populations were isolated by positive magnetic cell sorting ( AutoMACS Pro Separator , Miltenyi Biotec , Bergisch Gladbach , Germany ) according to manufacturer's instructions using FITC conjugated mouse anti-chicken CD4 , clone CT-4 and anti-chicken CD8β antibodies , clone EP42 [[53]; SouthernBiotech , Birmingham , Alabama , USA ) ] and goat anti-mouse IgG microbeads ( Miltenyi Biotec ) . After each antibody treatment , cells were washed three times with PBS containing 0 . 5% bovine serum albumin with centrifugation at 450 xg for 10 min . The purity of sorted cells was >99% by flow cytometry . Established lymphoma cell lines derived from MDV-1-induced tumors included MSB1[54] , HP8 [55] and HP18 [56] , RPL-1 [57] . Four additional MDV cell lines were established from four line P birds infected with pRB-1B5 [51] , from testes ( T ) , ovary ( O ) and spleen ( S ) tumors according to standard methods [56] . These have been given the following identifiers 4523 ( T ) , 4525 ( O ) , 4590 ( S ) and 760 ( O ) . The Reticuloendotheliosis virus T ( REV-T strain ) -transformed CD4+ T-cell line AVOL-1 [58] , [59] was included as a MDV-negative transformed cell line . Cell lines were grown at 38 . 5°C in 5% CO2 in RPMI 1640 medium containing 10% fetal calf serum , 10% tryptose phosphate broth and 1% sodium pyruvate . Tissue samples were stored in RNAlater ( QIAGEN Ltd . Crawley , United Kingdom ) at −20°C before disruption by homogenization ( Mini-bead beater; Biospec Products , Bartlesville , Okla . ) . Isolated cell subsets or cultured cells were disrupted by resuspension in RLT buffer ( QIAGEN ) and stored at −20°C . RNA was extracted with the RNeasy Mini kit ( QIAGEN ) according to the manufacturer's instructions . Contaminating DNA was digested on column with RNase-free DNase 1 ( QIAGEN ) for 15 min at room temperature . The RNA was eluted with 50 µl RNase-free water ( QIAGEN ) and stored at −80°C . Reverse transcription reactions were performed using the iScript Reverse Transcription system ( iScript Select cDNA synthesis Kit , Bio-Rad , USA ) according to manufacturer's instructions , using 2 µg of isolated RNA from each sample and oligo ( dT ) primers . Twenty µl of cDNA was obtained for each sample and stored at -20°C . PCR were performed according to standard protocols . Briefly , cDNA ( 2 µl ) was incubated with 200 µm dNTP , 1 . 5 mM MgCl2 , 1x reaction buffer [50 mM KCl , 20 mM Tris–HCl ( pH 8 . 4 ) ] , 2 units Platinum Taq DNA polymerase ( Invitrogen ) , 1 µl of each primer at 10 µM working concentration , in a 50 µl final reaction volume . The forward primer used for Vβ1 and Vβ2 was 5′ACAGGTCGACCTGGGAGACTCTCTGA CTCTGAACTG-3′ and 5′-CACGGTCGACGATGAGAACGCTACCCTGAGATGC-3′ respectively with a common Cβ reverse primer 5′ACAGGTCGACGTACCAAA GCATCATCCCCATCACAA-3′ [60] . The TCRβ locus lies on chromosome 1 with Vβ and Cβ primer design based upon genomic sequence ( version 82; http://www . ensembl . org/Gallus_gallus ) as described previously ( 45 ) . The use of primers that lie in conserved regions of the TCR segments minimises any bias associated with PCR amplification . Sequence analysis of samples derived from uninfected birds reveals a polyclonal population of amplified TCR CDR3 with no evidence of PCR bias ( 45 and our unpublished data ) . PCR conditions were as follows , one cycle , 94°C for 2 min , followed by 35 cycles of 94°C for 30 s , 50°C for 40 s and 72°C for 1 min , followed by one cycle at 72°C for 10 min using a G-storm thermocycler ( Gene Technologies , Essex , UK ) or Eppendorf mastercycler ( Eppendorf , Hamburg , Germany ) . The amplified products were analysed by electrophoresis through 1% agarose ( Sigma-Aldrich Ltd , Poole , UK ) gels in 1x Tris-borate-EDTA buffer at 50 mA for 1 hr , and products visualized by staining with ethidium bromide ( Bio-Rad , Ltd ) or GelRed nucleic acid stain ( Biotium ) . PCR products were purified using QIAquick PCR purification kit ( Qiagen Ltd ) according to manufacturer's instructions . DNA was eluted in 50 µl nuclease free water and stored at −20°C . To determine the sequence of the expressed Vβ-chain , PCR products were cloned directly into the pCR4-TOPO vector ( Invitrogen ) and used to transform competent E . coli , TOP10 ( Invitrogen ) according to the manufacturer's instructions . After incubation on selective LB agar plates containing 100 µg/ml Ampicillin ( Sigma ) , single bacterial colonies were picked and screened for insert of correct size by PCR followed by agarose gel electrophoresis . Positive colonies were processed using the Qiagen Miniprep kit ( Qiagen Ltd ) and subsequently sequenced with plasmid-specific ( M13 Forward; 5′-GTAAAACGACGGCCAG-3′or M13 reverse; 5′-CAGGAAACAGCTATGAC-3′ ) or Cβ specific reverse primer ( 5′-TGTGGCCTTCTTCTTCTCTTG-3′ ) . Alternatively , the plasmid insert amplified by PCR was purified using QIAquick PCR purification kit ( Qiagen Ltd ) according to manufacturer's instructions and sequenced directly using a nested Cβ specific reverse primer ( above ) . Sequencing was carried out by capillary electrophoresis on the CEQ 8000 sequencer according to the manufacturer's instructions ( Beckman Coulter , Fullerton , CA ) . Up to 22 ( usually ∼15 ) independent sequences were obtained with each sample . The sample size ( n ) was chosen with reference to the coefficient of variation of the binomial distribution , which is proportional to 1/√n . This means that the increased precision obtained by raising sample size above ∼n = 15 rapidly reaches a point of diminishing return . Appropriate confidence limits for the repeated sequence frequencies were calculated using the Adjusted-Wald method for binomial proportinos [61] . All sequence data was considered with reference to data generated by spectratype analysis of the CDR3 length profile generated from the total population of cells examined . To determine the CDR3 lengths of the amplified PCR products by spectratype analysis , a run-off reaction was performed as follows . Five µl of purified PCR product was incubated with 200 µm dNTP , 1 mM MgCl2 , 1x reaction buffer [50 mM KCl , 20 mM Tris–HCl ( pH 8 . 4 ) ] , 0 . 5 units Taq DNA polymerase ( Invitrogen ) , 1 µl of a WellRED dye D4 ( Sigma ) labelled nested Cβ specific reverse primer ( 5′-TCA TCT GTC CCC ACT CCT TC-3′ ) at 4 µM working concentration in a 20 µl final reaction volume . The reaction conditions were as follows , one cycle 95°C for 2 min , followed by 4 cycles of 57°C for 2 min and 72°C for 20 min using a G-storm thermocycler ( Gene Technologies , Essex , UK ) or Eppendorf mastercycler ( Eppendorf , Hamburg , Germany ) . The run-off reaction products were diluted 5x with nuclease free water and 1 µl of the diluted product was mixed with 40 µl sample loading dye ( Beckman Coulter , Fullerton , CA ) containing 0 . 25 µl DNA size standard kit-600 ( Beckman Coulter , Fullerton , CA ) . Samples were transferred into a 96 well plate , overlaid with mineral oil and immediately loaded into a capillary sequencer ( CEQ8000 Genetic Analysis System , Beckman Coulter ) for fragment analysis . For optimal results , samples were analysed using a modified fragment analysis program ( Frag-4 ) by increasing separation time to 75 min . The data was compiled in CEQ8000 analysis module and for each sample the range of base pair lengths of products was identified and displayed as spectratype profiles . Peak size data was extracted from the fragment analysis software and transferred into Microsoft Excel . Chi-squared tests were used to test whether each CDR3 length distribution differed significantly from that obtained with uninfected birds ( TCRVβ1 and TCRVβ2 from unsorted cells or those positively sorted for expression of CD4 or CD8β ) . The spectratype profiles derived from uninfected birds ( n = 3 for each population ) exhibited consistently broad CDR3 length distributions that were not statistically different to each other . Reference CDR3 length distributions were constructed for each population by calculating the mean proportion of signal obtained at each CDR3 length from uninfected samples .
In the first instance we selected eight MDV-transformed cell lines [[54] , [56] , [57] , [62] and our unpublished data] and subjected these to TCR repertoire analysis . The REV-T-transformed CD4+ T-cell line AVOL-1 [58] , [59] was included for comparison . All of the MD tumor cell lines expressed either Vβ1 or Vβ2 exclusively , whereas the REV-T transformed AVOL-1 cell line expressed both TCR Vβ1 and Vβ2 ( Figure 1A ) . The majority of the randomly selected cell lines ( 6/7 ) expressed Vβ1 suggesting a bias in tumor formation between the two avian TCRβ families . The spectratype-derived CDR3 length profiles for each MD cell line comprised a single spectral peak , whereas AVOL-1 contained multiple spectral peaks ( Figure 1B ) . PCR products were cloned into the pCR4-TOPO vector and the inserts sequenced from single colonies of transformed E . coli . For each MD cell line , all inserts contained identical TCRβ CDR3 sequences whereas three sequences were obtained for Vβ1 in AVOL-1 ( Figure 1C and S1 ) . Taken together , these data indicate the clonal nature of MD cell lines compared with an oligoclonal structure in the REV-T transformed AVOL-1 cell line . A fresh ovarian tumor was obtained from one pRB-1B5 MDV-infected bird ( designated Bird1 ) at post mortem ( 90 DPI ) . Spectratype analysis revealed a restricted TCRβ repertoire ( Figure 2A ) with a single spectral peak for Vβ1 . The Vβ2 spectratype profile of the ovarian tumor had two main peaks and 3 or 4 minor peaks . With Vβ1 all CDR3 sequences were identical ( Figure 2B ) corresponding in size to the CDR3 length observed by spectratyping , a profile similar to the tumor-derived cell lines . In contrast , with Vβ2 two repeated CDR3 sequences were detected , one which coded for the amino acid ( aa ) sequence ‘GIDSD’ at a frequency of 9/21sequences which translates to an estimate of 43% ( 95%CI 24-63% ) of the population and the second , ‘DRG’ at 7/21 ( 33% , 95%CI of 17–54% of the population ) . The remaining 5 sequences were singlets . The expanded Vβ2 clones may indicate presence of additional tumor clones , latently-infected T cells or a focussed T cell response infiltrating the tumor . These data demonstrate that MD tumor may consist of a monoclonal Vβ1 and an oligoclonal Vβ2 population . Application of spectratype and CDR3 sequence analysis to T cell populations from uninfected Line P birds revealed polyclonal repertoire profiles with no duplicated CDR3 sequence identified in any sample ( data not shown ) . Since MDV transforms CD4+ cells [9]–[12] , [14] we compared the CDR3 length distribution within unsorted and CD4+ populations of cells derived from tumors . Spectratype analysis of the liver and kidney tumors ( 32 DPI ) from two additional individuals ( designated Bird 2 and 3 ) revealed dramatic restriction in Vβ1 CDR3 length in unsorted cells ( Figure 3 , left column ) . These profiles were mirrored by the spectratypes of the CD4+ cell populations in all four tumor samples ( Figure 3 , middle column ) . Flow cytometry analysis showed that CD4 + cells represented between 88–98% of the cells derived from whole tumor ( data not shown ) . Cell lines were established from three tumors , two of which had spectratype profiles identical to those detected within isolated CD4+ cells ( Figure 3 , right column ) . With the kidney tumor of Bird 3 , the CDR3 spectra of cultured cells included a dominant peak of identical length to that in CD4+ T cells but also included a second slightly shorter peak . Sequence analysis revealed dominant sequences that were enriched by sorting for CD4+ cells and by ex vivo culture with the majority being derived from monoclonal expansions ( Figure 4 ) . The second spectral peak in the cultured cells of Bird 3 represented a second sequence detected once in the sorted CD4+ cells . Moreover , as a result of analysing two tumors from different organs from each individual , this data set also demonstrated that different tumor clones were present in different sites , with each site dominated by a single Vβ1 clone ( e . g . CDR3 aa sequences EWDRGTY and VGGDRLS for Bird 2 ) . In contrast to Vβ1 , the Vβ2 spectratype profiles of the 4 tumors ( Figure S2 ) and corresponding sequences ( Figure S3 ) indicate a wider repertoire although relatively large CD4 + T cell clones were detected in Bird 2 liver and kidney . However , none of these clones could be generated into transformed T-cell lines and may represent non-culturable tumors or a focussed T cell response . To identify the frequency of profiles consistent with metastatic tumor clones ( shared clones in multiple sites ) and those with independent origin ( different clones ) , we carried out the spectratype analysis of multiple tumor sites from further seven birds ( Bird 4 to 10 ) . The profiles obtained for both Vβ1 and Vβ2 are shown in Figure S4 ( A for Vβ1 and B for Vβ2 ) . Dominant spectral peaks shared between multiple sites were found in 6 of 7 birds but there were also site-specific over-represented spectral peaks in most individuals , for example with the kidney Vβ1 of bird 7 . Overall , the data indicate large bias in the profile of CDR3 length in all tumor sites ( p<0 . 001 ) and the shared peaks between sites will often be due to a common CDR3 sequence . However , as seen with Bird 2 sometimes the sequence will be distinct despite shared CDR3 length ( Figure 4 ) . Interestingly , the dominant spectral peak seen in multiple tumor sites was often evident in spleen and/or blood samples supporting an interpretation of metastatic spread for some tumor clones . Further spectratype and sequencing analyses were performed to identify the nature of the CD8+ response ( see below ) , where cells from multiple tumor sites were sorted into CD4 and CD8 fractions . The spectratype profiles for whole tumor or sorted CD4+ cells from tumor sites in Birds 11 to 14 were similar to those seen with Bird 1 to 10 , with dominant spectral peaks in tumor sites ( Figure S5 ) . Some of the dominant peaks were shared between tumor sites within a single bird whilst others were specific for particular sites . The Vβ1 and Vβ2 products were sequenced for all tumor sites in Birds 11 and 12 ( Figures S6 to S9 ) . In the absence of culturable T-cell lines generated from these tumors , we tentatively defined tumor-like clones as CD4-enriched and representing greater than 30% of the sequences in any one site ( most were much higher frequency than 30% ) . Specifically , the sequence data for Vβ1 in CD4+ cells from Bird 11 ( Figure S6 ) identifies three large tumor-like clones , “LDGTGGY” ( liver only ) , “RRLTGD” ( kidney and as a singlet in ovary ) and “LDTGGS” ( liver , kidney and ovary ) . The sequence for Vβ2 in CD4+ cells of Bird 11 ( Figure S7 ) revealed one highly over-represented sequence in all sites ( ILRDRGW ) that may represent a metastatic tumor and a second in the spleen ( IRLGTGGY ) . For Bird 12 ( Figure S8 ) no Vβ1 CDR3 were represented at over 30% of CD4+ T cell derived sequences but one Vβ2 sequence ( Figure S9 ) with the CDR3 motif “QG” was dominant in the kidney ( 18/19 CD4+ sequences ) and detected in ovary and spleen . A second CD4+ , Vβ2 CDR3 sequence “FVMRGID” was dominant in the ovary but not detected elsewhere . In most individuals the sequencing approach revealed smaller clones of CD4+ cells ( repeated but <30% of sequences in any site ) including Vβ1 with Birds 2 , 11 and 12 and in Vβ2 with Birds 3 , 11 and 12 ( Figures 4 , S3 , S6 to S ) . These sequences may also represent small tumor clones or responding cells but the expansion of one of these sequences in cultured cells from Bird 3 kidney indicates that the “small tumor clone” explanation is valid . Global attribution of the smaller clones of CD4 T cells to a response or tumor phenotype is not possible with the current data sets . Nonetheless , our data clearly demonstrated that culturable tumors were usually dominated by a single T cell clone but that different sites within the same individual can contain independent tumor clones . The detection of tumor clones in the blood , at post-mortem raised the possibility of identifying tumor clones prior to the occurrence of overt disease . Initial analysis with samples of blood collected ∼2 weeks before the birds exhibited clinical signs supported the notion that the TCR spectratype would be useful to detect tumor clones circulating in the blood . The results of Vβ1 analysis of peripheral blood leukocyte ( PBL ) samples for two birds ( Bird 15 and 16 ) are given in Figure 5 . The samples from liver , kidney , muscle , heart and spleen taken at 49 DPI from Bird 15 revealed a dominant spectral peak that could also be detected in the blood at 42 and 35 DPI ( leading to a significant bias in the spectral profile; p<0 . 001 ) . Similarly , Bird 16 shared the same spectral peak in liver , kidney and ovary with an overrepresented peak and a biased CDR3 profile in the blood at 35 DPI ( p<0 . 001 ) , one week prior to the onset of clinical disease . In Bird 16 , there was also a second spectral peak in the ovary and a non-shared spectral peak in the muscle that were not detected in the blood . A further two birds ( 17 and 18 ) were blood sampled serially ( twice a week ) throughout infection for more precise detection of the tumor clones in the blood , and the results for Vβ1 and Vβ2 spectratypes are depicted in Figure 6 . The tumor profile for Bird 17 at post-mortem ( 33 DPI ) indicated a shared spectral profile for Vβ1 in kidney , testes and spleen ( and in CD4+ cells isolated from kidney and spleen ) and a second site-restricted tumor in the kidney comprising CD4+ Vβ2+ cells . The multi-site tumor CDR3 spectral length was readily detected in the PBL from 16 DPI ( p<0 . 01 and at later time points p<0 . 001 ) whereas earlier PBL samples exhibited a “normal” distribution of CDR3 lengths that were not significantly different to the spectral profiles obtained from uninfected birds . In contrast , the site specific Vβ2 tumor was not detected as a spectratype bias in the PBL at any time . The tumor profiles of Bird 18 revealed one shared site ( ovary and spleen ) Vβ1 tumor , one single site Vβ1 tumor ( liver ) and one shared site Vβ2 tumor in all three sites ( although the more complex ovarian tumor spectratype suggest it may be less highly represented ) . The multi-site Vβ1 tumor was detected as spectral bias in the PBL between 16 and 19 DPI ( p<0 . 001 ) although the overall bias was less dramatic than seen with Bird 17 . The spectral profiles of PBL from MDV infected birds indicate that multi site tumor clones can be readily detected in the blood over two weeks prior to clinical symptoms . Unlike the multi site tumors , those restricted to a single site were not detected in the blood . The appearance of tumor clones in the blood affected the repertoire of the overall PBL population especially within the TCRVβ family that comprise the tumor ( e . g . for Bird 17 , the blood Vβ1 profile was completely dominated by the tumor ) . Moreover , the disturbance caused by a large CD4+ T cell tumor clone in Vβ1 also affected the repertoire profile of Vβ2 ( compare pre- and post- 12 DPI spectratype profiles ) with significantly altered CDR3-length profiles in the PBL of Bird 17 at 16 DPI ( p<0 . 005 ) , 29 DPI and 33 DPI ( both p<0 . 001 ) . Although the nature of the tumor complicates identification of CD4+ T cell responses the CD8+ TCRαβ+ T cells clearly represent a responding T cell population capable of specific recognition , cytokine production and anti-MDV capability [38] , [39] , [63] . Moreover , in humans infected with persistent viruses ( e . g . EBV , CMV and HTLV ) the responding CD8+ T cells develop a highly focussed repertoire [2] , [41]–[43] , [64] , [65] . Hence , to define the repertoire of the CD8+ response in MDV infected birds , we isolated CD8+ T cell populations from a range of tumor sites and subjected them to spectratype and sequence based repertoire analysis ( simultaneous analysis of CD4+ populations was used to determine the nature of the tumor profiles in these individuals , Figures S5 , S6 , S7 , S8 , and S9 ) . Spectratype profiles obtained for Vβ repertoire analysis of CD8+ cells isolated from multiple tumor sites in four birds ( 11 to 14 ) are presented in Figure 7 . CD8+ cells represented a minority cell population within the tumor , ranging between 0 . 4 and 5% by flow cytometric analysis ( data not shown ) . Highly purified CD8+ cells ( >99% ) exhibited a restricted Vβ1 CDR3 length spectral profile ( p<0 . 001; Figure 7 ) . Within birds , the spectral profiles taken from different sites often included shared peaks detected in multiple samples . The Vβ2 spectral profiles were more variable but were also characteristic of biased populations ( p<0 . 01 to p<0 . 001 ) with large over-represented peaks in some samples . The Vβ1 products were sub-cloned and sequenced from all sites in two birds ( 11 and 12 ) ( Figure 8 ) allowing identification of clonal expansion by the presence of repeated sequences . These sequences included the CDR3 aa motif “GGS” present in both Bird 11 and 12 as a large , multi-site , overrepresented “public” CDR3 sequence . Considering this clone was the only sequence at this length in either Bird 11 or 12 it is intriguing that this spectral peak was also over-represented in the CD8+ T cells from Bird 13 and 14 . Other repeated CDR3 sequences in CD8+ T cells included “RDRGIY” ( in liver kidney and spleen ) , “SRTGGS” ( ovary and spleen ) and “IFGIY” ( spleen ) of Bird 11 and “GGSI” in the spleen of Bird 12 . Further candidate CD8+ CDR3 sequences were identified as present in unsorted populations and not present in CD4+ sorted populations . These included those revealed by the Vβ2 sequencing efforts; two from Bird 2 ( ETGGVY and FAFIDRGI ) , one from Bird 3 ( TIERVD ) , two from Bird 11 ( EVGEILY and TTPQGDRSQ ) and one from Bird 12 ( RGGYQPA ) . Collectively , these results indicate a highly focussed CD8+ T cell response with some clones present at high frequencies in multiple tumor sites and the spleen . The tumor profile of Bird 11 ( 5 tumor-like clones , with two metastatic ) and Bird 12 ( 2 tumor-like clones with one metastatic and one ovary-restricted ) may relate to the identity of the CD8+ T cell expansions seen in different sites . For example the public GSS CDR3 sequence was detected at most tumor sites , whereas some other CD8+ clones were more restricted in their distribution to certain locations . Based upon an assumption of similar TCR mRNA levels in all cells and the known numbers of Vβ1+ and Vβ2+ CD8+ cells in the tumor and spleen we can estimate the size of the CD8+ clones in the tumor site and spleen . For example , within Bird 11 , the splenic population of the three CD8+ Vβ1+ clones “GGS” “RDRGIY” and “IFGIY” each represented 12 . 5% of the CDR3 which translate into populations of ∼25 million cells ( 95%CI 4–74×106 ) . . In Bird 12 , the public CDR3 aa sequence “GGS” was present in 6/22 ( 27% ) CDR3 sequences from CD8+ T cells representing ∼54 million cells ( 95%CI 24–96×106 ) and the private “GGSI” represented in ∼27 . 2 million cells ( 95%CI 8–72×106 ) . For comparative purposes we have displayed the aa identity of all over-represented CDR3 sequences identified in this study and grouped these according to frequency in different T cell subsets ( Figure 9 ) . All cell lines contained monoclonal CDR3 sequences except for one short-term cultured cell line , which was biclonal . Within CD4+ T cells derived from tumor sites , fourteen high frequency CDR3 ( >50% ) were identified with ten represented at greater than 70% of the sequences obtained . Of the 21 “high frequency” CDR3 ( established cell lines , ex vivo cultured cells and tumor sites ) , these were distributed in Vβ1 and Vβ2 based CDR3 ( 13 and 8 respectively ) . All four Jβ segments were represented . Other CD4+ CDR3 were present at 10 to 30% with a small number of low frequency ( <10% ) repeated sequences . Within positively sorted CD8 or non-CD4 ( presumably CD8+ ) populations some large clones were detected , most of which represented private CDR3 but one represented a public CDR3 sequence detected in multiple birds . Samples of T cells from uninfected birds were polyclonal ( no repeated CDR3 ) and none of the CDR3 seen in MDV infected birds was detected ( data not shown ) .
Virus driven transformation of lymphoid cells is a major clinical consequence of infection with persistent infections such as EBV and HTLV in humans . Progress in understanding these human diseases is hindered by the lack of suitable model systems . MDV represents a natural α-herpesvirus of galliform birds capable of inducing rapid onset of tumors in susceptible birds . Losses caused by this group of viruses also represent a substantial problem in their own right; without MDV vaccination the poultry industry would be unsustainable . Indeed the ability to vaccinate against MDV tumor formation has implications for control of medically relevant tumors [1] , [6] . Within this framework , we addressed the issue of T cell clonality during infection and tumor formation , dissecting the tumor , spleen and blood to identify repertoire changes in the transformed CD4+ cells and the responding CD8+ cells . With MD almost all cell lines and in vivo tumors have been characterised as CD4+ , [9]–[11] , [13] , [14] . In one study using the intraperitoneal infection route one of twelve cell lines was CD4- CD8α+ but this lacked expression of CD8β [12] . All of the CD8+ samples in this study were prepared using anti-CD8β to avoid isolation of non-classical CD8αα T cells . We chose to examine the Vβ profiles as a measure of clonality since this receptor is clonally expressed with a single in-frame sequence present in each clone of T cells due to the process of allelic exclusion that takes place during T cell development in the thymus [66]–[68] . Tumor clonality is a fundamental issue in MD pathogenesis . The infectious cycle involves transfer of the virus from the lungs to initiate a cytolytic infection in B cells . This is followed by spread and lytic cycling infection largely within CD4 TCRαβ T cell population , before development of latent infection and transfer of MDV into the feather follicle epithelium from where the infectious virus is shed [8] . All infected birds experience a persistent , latent infection , and susceptible birds develop tumors usually within 4 to 5 weeks . Herein resides the problem; if the transformation event is rare , how to explain the high penetrance and temporal reproducibility of the tumor phenotype , unless the “tumors” are induced as a result of polyclonal transformation . Previous studies have addressed this issue in relation to the pattern of MDV genomic integration within the host cell genome [18] , [19] or by cell surface staining with CD30 as a tumor associated marker [10] . These two studies reached opposing conclusions , with the restricted MDV integration profiles used to propose clonal tumors ( with metastasis ) , contrasted with the high expression of CD30 in both TCRαβ families within a single tumor being used to propose polyclonality . Our studies using TCRVβ repertoire analysis techniques [45] as a viral integration independent clonal “bar-code” to identify the repertoire of CD4+ TCRαβ+ T cells in tumor-derived cell lines and with in vivo derived tumor samples revealed a characteristic of clonal dominance within an oligoclonal framework of tumor-capable CD4+ T cells . All of the established tumor-derived cell lines were monoclonal ( each expressing a single TCRβ CDR3 sequence ) , although one short-term line developed during the course of these studies was biclonal at second passage . In contrast the REV-T-transformed AVOL-1 cell line was oligoclonal after over 37 passages expressing at least three TCRVβ1 and one Vβ2 TCR CDR3 sequences . The spectratype profiles obtained with all cell lines were diagnostic in terms of the clonality of the CDR3 as defined by sequence analysis . The clonal structure of the cell lines was not influenced by the length of time in culture which suggests that monoclonality is not an artefact of in vitro selection as a result of multiple passages . It is therefore likely that selection for dominant transformed clones had already occurred in vivo and is retained in MDV cell lines as suggested previously [19] . Furthermore , cell lines generated in this study from fresh tumors expressed a TCR identity shared with the source tumor in vivo . Where cell lines were established most ( ∼90% ) expressed the Vβ1 family of T cell receptors with only one expressing Vβ2 , a ratio consistent with the 84% bias previously reported [13] . The profile of most primary tumors was dominated by a single clone of transformed T cells , although biclonal dominance in individual tumor sites was not uncommon . However , sequence analysis revealed smaller secondary clones of expanded CD4+ T cells in most tumors ( ∼10% of the CDR3 sequences ) and the outgrowth of one of these during ex vivo culture indicates the tumor potential of sub-dominant CD4+ clones . Some of the very large clonal populations also failed to establish as tumor cells lines ex vivo , perhaps indicating a phenotypic variability in transformation state . Indeed , considering the very large TCR clones ( 40 to 100% of CD4+ CDR3 in one site ) these were evenly distributed between TCRVβ1 ( 9 sequences ) and TCRVβ2 ( 8 sequences ) ( Figure 9 ) . The bias in TCRVβ usage within cell lines may represent a cultivation artefact or reflect the biology of cells expressing different TCR family members . Nonetheless , the multi-step analysis of dominant CDR3 in the primary tumor , in sorted CD4+ cells and after establishment of lymphoblastoid cell lines ex vivo are important in confirming the capacity of the identified large clones to express a tumor capability . All four Jβ segments were present in the CDR3 of both TCRVβ1 and TCRVβ2 expressing large tumor-like clones or in cultured tumor cell lines . Our data resolves many of the issues surrounding MD tumor clonality . Essentially , we demonstrate clonal dominance within MD tumors ( broadly similar to that reported by Delacluse et al . , [19] ) although our integration-site independent analysis using the T cell receptor CDR3 region revealed a more complex clonal framework within , and between , tumor sites in vivo . Different tumor sites within a single individual may be dominated by shared or distinct clones , hence a single individual may experience multiple transformation events giving rise to tumors that have very different characteristics . On most occasions the dominant in vivo clone present at a particular site was the only clone represented in ex vivo cultured cell lines grown under tumor culture conditions . However , on one occasion one of the lower frequency clones exhibited tumor-like growth patterns ex vivo ( alongside the dominant clone in the original site ) indicating that some of the smaller clones exist in a transformation capable state . The fact that many individuals harbour both metastatic and single-site tumor clones indicates a complex interplay between transformation and clonal competition . Indeed , with most individuals the overall tumor burden was the result of a small number of independent transformation events ( i . e . more than one but fewer than 3 or 4 ) . In contrast , with some individuals the multi-site tumors were the result of metastasis from a single tumor clone . The relationship between these “successful” tumor clones and the infected cell population deserves attention . In a broader context , the monoclonal origin of adult T-cell leukaemia/lymphoma ( ATLL ) induced by the human T-lymphotropic virus type -1 ( HTLV-1 ) associated malignancy is well documented [2] , [69] , [70] . This profile is probably related to the rarity of ATLL even among HTLV-1 seropositive individuals [3] reflecting the acquisition of secondary genomic lesions in persistently infected T cells . Nonetheless , the rapid onset MD tumors with clonal dominance in the context of a more complex framework of oligoclonal expansion may also reflect a circumstance common to other tumor associated persistent viruses of lymphocytes including HTLV-1 . Perhaps the main differences may lie in the vigor of MDV-induced T cell replication leading to a compressed time-frame compared with other lymphotropic , tumor associated viruses . Biological differences were also detected amongst the very large clonal CD4+ “tumors” , with some clones found in multiple sites including the blood and spleen whereas others were located in a single site , indicating phenotypic diversity based upon metastatic capability . The identification of metastatic tumor clones in the blood allowed serial analysis of blood samples from infected birds to determine the dynamics of the appearance of the tumor clone , in relation to the time of infection and onset of clinical signs . The spectratype analysis of blood samples prior to infection and in the first 10–14 days revealed a profile consistent with a polyclonal population of circulating cells . However in some cases , the ‘tumor-specific’ spectratype signature could be detected in blood 12 to 16 dpi , more than two weeks before appearance of clinical signs . The appearance of the tumor clone at detectable levels in the blood supports the proposal of an early transformation event . The level of tumor clone expansion in the blood compartment at the onset of clinical disease was extreme , and in some individuals , these were the only T cell clones detectable ( e . g . within TCRVβ1 for Bird15 and 17 ) represented the tumor ( Figure 5 and 6 ) . There was also evidence for disturbance within the polyclonal repertoire in TCRVβ2 expressing cells ( Figure 6 ) suggesting that the blood niche for T cells was being filled by the tumor . Hence , with a circulating TCR profile dominated by a single clone , it is of little surprise that MDV-infected birds develop immune deficiency [reviewed in [71]] . These dramatic repertoire changes would have greater impact than the reported changes in cytokine production [72] and would be immunologically catastrophic . Infiltration of the skin with CD4+ T cells is a consequence of MDV infection [73] , [74] [75] and the high frequency tumor clones in the blood are likely to represent the relocation of MDV to the site of transmission . In mammals , many persistent viral infections including EBV , CMV and HTLV stimulate highly focussed repertoire expansion in responding CD8+ T cells [2] , [76] , [77] . Although the MDV tumors were populated by relatively small numbers of CD8+ T cells , their repertoire was highly structured and oligoclonal in nature . The CD8+ T cell clone sizes of around 25 to 50 million cells are similar to those reported during persistent viral infections in humans [78] . However in the case of MD , these are developed over a much shorter period of time than considered with mammalian infections . For example , taking a conservative estimate of prolonged T cell division of 12 hours/division [79] and assuming no cell death ( unlikely ) , the latest time point for initial stimulation of the CD8+ T cell would be ∼15 days prior to sampling . This calculation would place the initiation of these clones of specific CD8+ T cells at ∼15 DPI , probably earlier , around the time at which latent infection was initiated . The rapid focussing and clonal expansion of the MDV-specific repertoire suggests restriction to a small selection of MDV antigens . Indeed , Omar and Schat [38] examined the cytolytic response of infected birds against a panel of cell lines expressing individual genes from MDV found that in MHC B19 homozygote Line P2a birds , the cytolytic activity was restricted to meq , gB and pp38 antigens , while the genetically-resistant line N2a ( B21 ) birds also detected the ICP4 antigens . In our studies , tumor-infiltrating CD8+ T cells produce greater levels of IFNγ mRNA than CD8+ T cells derived from the spleen of uninfected birds ( unpublished data , Mwangi , Peroval et al . , ) . The CD8+ T cell response of susceptible birds is insufficient to prevent tumor progression; our data provides a framework for comparisons with resistant or vaccinated birds which do not develop tumors . Our sequence analysis clearly detected large CD8+ T cell clones and allowed an approximation of the clone size , the application of higher throughput sequencing technologies may be useful in the future to identify smaller clonal expansions and provide more accurate estimations of clone sizes . Understanding the nature of the TCR repertoire to specific antigens after infection and vaccination can be used to improve vaccine approaches in the future . The rapid nature of focussing within the CD8+ population may reflect a combination of the minimal MHC configuration where each haplotype is dominated by presentation through a single MHC class I gene [80] and the minimal TCRVβ locus with 13 Vβ segments in just two families [45] . The high frequency CD8+ T cell clones were found in both tumor sites and in the spleen of infected individuals , either restricted to one tumor site or present in multiple tumor sites . One of the largest CD8+ clones has a CDR3 sequence ( “GSS” ) of note , in that identical sequences were detected in different individuals . This type of CDR3 is known as a “public” TCR rearrangement and , although previously reported with mammals , is relatively rare [81] . Upon closer examination , it was clear that the public GSS amino acid sequence for the CDR3 also represented shared nucleotide sequence in different individuals . Interestingly , the GSS sequence represents retention of a fragment of the D segment , after deletion of six nucleotides in the D and three nucleotides in the Vβ1 segment . Although not noted previously , it is clear that a CDR3 constructed by deletion ( with no retained nucleotide addition ) is much more likely to occur in multiple individuals than one generated by addition of nucleotides . We propose that public CDR3 sequences in other contexts ( e . g . in humans ) may also conform to this arrangement , representing a deletion-based junctional modification . This feature might be useful and exploitable in diverse scenarios to improve “public” responses to vaccines . The remaining CDR3 sequences positively identified as clonal expansions in CD8+ cells ( or as not in CD4+ cells ) all represented “private” CDR3 identities ( Figure 9 ) . In this report , we have documented the TCR Vβ repertoire changes associated with infection , tumor development and anti-tumor response that characterise MDV pathogenesis . Upon consideration of our data in the context of previous reports , we propose that the MD tumors are dominated by clonal expansion in an oligoclonal framework of minor clones of pre-cancerous cells . We propose that this type of population structure explains the penetrance and narrow temporal window that characterise MD in susceptible birds . The CDR3 analysis identified that all established MDV-transformed cell lines tested were clonal ( with one bi-clonal short term culture ) , and that these clones represent dominant clones detected in vivo . Within birds harbouring multiple tumors there was a mixture of metastatic and site-specific tumor clones . Overall , we examined 50 tumors derived from 21 individuals , and all tumors were dominated by one or two clones with some birds harbouring a single metastatic tumor clone and others with different clones in different sites . The TCR repertoire analysis system has allowed examination of diverse areas of MD lymphoma biology and the CD8+ response against the infection . We consider that this type of approach can be used to further define MD pathogenesis and the response generated against infection and/or tumors . These types of study also have the potential to impact much more broadly , identifying strategies to vaccinate against or otherwise control viral driven lymphomas in medical and veterinary fields .
|
Many viral infections target the immune system , making use of the long lived , highly proliferative lymphocytes to propagate and survive within the host . This characteristic has led to an association between some viruses such as Epstein Barr Virus ( EBV ) , Human T cell Lymphotrophic Virus-1 ( HTLV-1 ) and Mareks Disease Virus ( MDV ) and lymphoid tumors . We employed methods for identifying the T cell receptor repertoire as a molecular bar-code to study the biology of MDV-induced tumors and the anti-tumor response . Each individual contained a small number of large ( high frequency ) tumor clones alongside some smaller ( lower frequency ) clones in the CD4+ T cell population . The tumor infiltrating CD8+ T cell response was highly focused with a small number of large clones , with one representing a public CDR3 sequence . This data is consistent with the recognition of a small number of dominant antigens and understanding the relationship between these and protective immunity is important to improve development of new vaccination strategies . Collectively , our results provide insights into the clonal structure of MDV driven tumors and in the responding CD8+ T cell compartment . These studies advance our understanding of MDV biology , an important poultry disease and a natural infection model of virus-induced tumor formation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"physiology/immunity",
"to",
"infections",
"virology/animal",
"models",
"of",
"infection",
"infectious",
"diseases/viral",
"infections",
"virology/viruses",
"and",
"cancer",
"pathology/immunology",
"oncology/myelomas",
"and",
"lymphoproliferative",
"diseases",
"immunology/immunity",
"to",
"infections",
"virology/host",
"antiviral",
"responses"
] |
2011
|
Clonal Structure of Rapid-Onset MDV-Driven CD4+ Lymphomas and Responding CD8+ T Cells
|
A complete understanding of the mechanisms underlying the acquisition of protective immunity is crucial to improve vaccine strategies to eradicate malaria . However , it is still unclear whether recognition of damage signals influences the immune response to Plasmodium infection . Adenosine triphosphate ( ATP ) accumulates in infected erythrocytes and is released into the extracellular milieu through ion channels in the erythrocyte membrane or upon erythrocyte rupture . The P2X7 receptor senses extracellular ATP and induces CD4 T cell activation and death . Here we show that P2X7 receptor promotes T helper 1 ( Th1 ) cell differentiation to the detriment of follicular T helper ( Tfh ) cells during blood-stage Plasmodium chabaudi malaria . The P2X7 receptor was activated in CD4 T cells following the rupture of infected erythrocytes and these cells became highly responsive to ATP during acute infection . Moreover , mice lacking the P2X7 receptor had increased susceptibility to infection , which correlated with impaired Th1 cell differentiation . Accordingly , IL-2 and IFNγ secretion , as well as T-bet expression , critically depended on P2X7 signaling in CD4 T cells . Additionally , P2X7 receptor controlled the splenic Tfh cell population in infected mice by promoting apoptotic-like cell death . Finally , the P2X7 receptor was required to generate a balanced Th1/Tfh cell population with an improved ability to transfer parasite protection to CD4-deficient mice . This study provides a new insight into malaria immunology by showing the importance of P2X7 receptor in controlling the fine-tuning between Th1 and Tfh cell differentiation during P . chabaudi infection and thus in disease outcome .
Despite efforts to develop vaccines and antimalarial drugs , Plasmodium infection still causes the death of about half a million people yearly [1] . The most prevalent Plasmodium species , Plasmodium falciparum and Plasmodium vivax , persist for very long time periods in the bloodstream of infected individuals . Parasite persistence is also ensured by repeated re-infections in hyperendemic areas . Although the major clinical manifestations of the disease attenuate after a few malaria episodes , repeated exposure to the parasite over several years is required to control parasite population growth [2] . Furthermore , protective immunity is usually lost in the absence of continued exposure to the parasite [3] . The ability to survive the effector mechanisms of innate immunity and to evade the acquired immune response for long periods shows how difficult it is to combat malaria and how appropriate the immune response must be to eliminate the parasite . Therefore , a complete understanding of the mechanisms underlying the acquisition of protective immunity is crucial to improve vaccine strategies to eradicate malaria . Particularly concerning the stimulatory signaling required for optimal activation of CD4 T cells , which have a central role in protection against malaria by producing IFNγ and helping B cells to secrete antibodies [4] , [5] . The intraerythrocytic cycle of P . falciparum and P . vivax has a synchronic periodicity and , consequently , the delivery of immune stimulatory molecules and subsequent fever episodes occur periodically after the rupture of infected red blood cells ( iRBCs ) . It has been shown that parasite components , such as glycosylphosphatidylinositol ( GPIs ) -anchored molecules and DNA from P . falciparum , activate macrophages through the toll-like receptor ( TLR ) 1/TLR2 and TLR9 signaling , respectively [6] . Nevertheless , it is still unclear whether recognition of damage signals contributes to activating the immune system in individuals suffering from malaria . During pathogenic infection and tissue injury , nucleic acids and their metabolites are released from dead cells and induce inflammatory and reparatory responses [7] , [8] . Adenosine triphosphate ( ATP ) is released passively from necrotic cells and through pannexin-1 hemichannels from apoptotic cells [9] . ATP accumulates in iRBCs and is released into the extracellular milieu through ion channels in the erythrocyte membrane or upon iRBC rupture [10] , [11] . ATP is also released through pannexin-1 hemichannels in the immune synapsis formed between T cells and antigen presenting cells ( APCs ) , triggering ATP-gated ionotropic P2X receptors that promote IL-2 secretion and T cell proliferation [12] . Unlike P2X1 and P2X4 receptors , which translocate into the immune synapsis , the P2X7 receptor remains uniformly distributed across the cell surface , allowing T cells to sense environmental ATP . The P2X7 receptor is activated only at high ATP concentrations [13] , hence it may be particularly important to help T cells distinguish tissue-damaging infections from quiescent infections or reminiscent antigens from a previous infection . Transient P2X7 activation promotes T cell response due to the formation of a non-selective cation channel that allows calcium influx [14] . However , sustained signaling induces the formation of large transmembrane pores and , consequently , leads to loss of membrane integrity and T cell death [15] . Although P2X7 signaling has important consequences for T cell biology , few studies have addressed the direct effects of this signaling pathway on T cell fate in vivo . Interestingly , the P2rx7 gene is highly expressed in follicular helper T cells ( Tfh ) located in Peyer's patches and the P2X7 receptor critically controls their numbers and , consequently , the production of IgA against gut commensals [16] . Increased P2rx7 expression is also a feature of regulatory T cells ( Treg cells ) ; ATP stimulation inhibits Treg cell generation and suppressive activity through the P2X7 receptor [17] . Moreover , it has been shown that P2X7 activation by extracellular ( eATP ) can be abrogated by CD39 ( nucleoside triphosphate diphosphohydrolase-1 ) , an ectoATPase that degrades ATP or adenosine diphosphate ( ADP ) to adenosine monophosphate ( AMP ) . CD39 is constitutively expressed on Treg cell surface [18] , providing protection against ATP-induced cell death [19] . CD39 and CD73 ( ecto-5’-nucleotidase ) also contribute to the suppressive activity of Treg cells [18] . CD73 hydrolyses extracellular AMP to adenosine , which is an important physiological regulator of the immune response [20] . In this study , we investigated using the blood-stage Plasmodium chabaudi ( Pc ) murine infection model whether P2X7 signaling contributes to CD4 T cell subset differentiation in malaria . The infection with synchronic Pc parasites develops from an acute phase to a long-lasting chronic phase , which accurately reproduces several aspects of human malaria [21] . IFNγ production is associated with the development of protective immunity [22] , [23] . A major source of IFNγ during acute Pc infection is class II MHC ( major histocompatibility complex ) -restricted CD4 T cells , which also help B cells to secrete antibodies [24] . The complete elimination of chronic parasitemia and protection against reinfection require Th1 cells [25] , [26] , which are particularly important in ensuring long-term strain-transcending immunity [27] . A Th1 cell population co-expressing IFNγ and IL-10 also plays a key role in protecting against severe malaria pathology [28] . Furthermore , Tfh cells provide critical help to B cells to produce high affinity antibodies [29] , [30] , and have been the focus of recent studies in murine and human malaria [31] . Tfh cells are implicated in protection against both Pc and Plasmodium yoelii 17XNL parasites [32]-[34] . Our results suggest that P2X7 receptor is required for Th1 cell differentiation during Pc infection but it also controls the Tfh cell population . Using adoptive transfer experiments , we showed that the selective absence of the P2X7 receptor in CD4 T cells is sufficient to impair Th1 cell differentiation and increase the Tfh cell population . Evidencing the importance of the fine-tuning between Th1 and Tfh cell populations in the control of Pc infection , the balanced Th1/Tfh cell population that differentiated in the presence of P2X7 receptor displayed higher ability to transfer protection to CD4-deficient mice than the increased Tfh cell population developed in its absence . The present study adds novel information on the malaria immunology field by demonstrating the critical role of the P2X7 receptor for the outcome of Pc infection by promoting Th1 cell differentiation to the detriment of Tfh cells .
To investigate the participation of the P2X7 receptor in blood-stage Pc malaria , disease progression was evaluated in C57BL/6 ( B6 ) and P2rx7-/- mice . The absence of the P2X7 receptor led to a worsening of the disease in infected females and males . Infected P2rx7-/- males showed higher parasitemias than their B6 counterparts , and 80% of the animals died during acute infection ( Fig 1A and 1B ) . The disease developed similarly in both female groups up to day 7 p . i . ; however , after this period , P2rx7-/- mice had impaired parasitemia control and limited recovery of clinical parameters ( i . e . anemia , weight loss and hypothermia ) ( Fig 1C ) . P2rx7-/- females also exhibited higher chronic parasitemias than B6 females ( Fig 1A ) . To determine whether the protection induced by the P2X7 receptor is a general phenomenon in murine malaria , we assessed the course of P . yoelii 17XNL infection in B6 and P2rx7-/- mice . It has been shown previously that , in contrast to Pc infection , the lack of IFNγ modestly affects the control of P . yoelii 17XNL infection [35] , whereas antibody deficiency leads to a lethal outcome [32] . Unlike the case of Pc malaria , more control of P . yoelii 17XNL parasites was observed in P2rx7-/- mice than in B6 mice , apparent from day 9 to 20 p . i . when mice of both groups controlled the parasitemia ( Fig 1D ) . Taken together , we showed that Pc infection control relies on P2X7 signaling; instead , the presence of P2X7 signaling results in increased peak parasitemia during P . yoelii 17XNL infection . These results suggest that P2X7-mediated eATP sensing might play a role in controlling the balance between IFNγ- and antibody-mediated immune responses to malaria . To determine whether the amount of ATP released during iRBC lysis is sufficient for P2X7 activation , P2X7-associated pore formation was evaluated by ethidium bromide ( EB ) uptake in blood CD4 T cells , which were obtained before and after iRBC lysis during a synchronized parasite cycle ( S1A Fig ) . This assay explores the P2X7-mediated formation of large transmembrane pores after sustained ATP stimulation , which allows EB to enter the cell and stain the nucleus [13] . The EB staining in B6 CD4 T cells was increased after the rupture of iRBCs at 4 and 5 days p . i . , and this effect was abolished in P2rx7-/- CD4 T cells ( Fig 2A ) . Accordingly , higher ATP serum levels were detected after iRBC lysis ( S1B Fig ) . The effects of acute Pc infection on the splenic CD4 T cell response to eATP were then assessed in vitro at 4 and 7 days p . i . , representing the interval between CD4 T cell proliferation and IFNγ secretion [24] . A marked increase in ATP-induced P2X7-mediated pore formation was observed in CD4 T cells at 4 days p . i . compared with that in non-infected controls ( Fig 2B ) . Although the B6 CD4 T cell response to ATP was drastically reduced at day 7 p . i . compared with that at day 4 p . i . , it remained augmented in relation to that of non-infected controls . The splenic CD4 T cell response to Pc infection was then compared in B6 and P2rx7-/- mice . CD4 T cells at day 4 p . i . were stimulated in vitro with iRBCs , mimicking the in vivo condition in which parasite antigens are available along with ATP released by the iRBCs; the responses were evaluated after 72 h of culture , a time point that corresponds to peak proliferation and IFNγ production [24] . P2rx7-/- CD4 T cells proliferated less than B6 CD4 T cells in the presence of iRBCs and splenocytes from naïve nude mice , as a source of APCs expressing the P2X7 receptor ( Fig 2C ) . IFNγ production was also reduced in iRBC-stimulated P2rx7-/- CD4 T cells , whereas IL-10 was produced at low levels regardless of P2X7 expression ( Fig 2D ) . The treatment with apyrase ( ATP diphosphohydrolase ) or brilliant blue G ( BBG , P2X7 antagonist ) inhibited CD4 T cell proliferation and IFNγ production , confirming the involvement of ATP and P2X7 receptor ( S2A and S2B Fig ) . Accordingly , the lack of the P2X7 receptor drastically impaired the in vivo expansion of the splenic CD4 T cell population during acute Pc infection ( Fig 2E ) . Moreover , at 7 days p . i . , the IFNγ+CD4+ cell number per spleen was higher in B6 mice than in P2rx7-/- mice ( Fig 2F ) . Explaining the gender influence in the susceptibility to acute Pc infection , B6 and P2rx7-/- males had lower IFNγ+CD4+ cell numbers per spleen at day 7 p . i . than female counterparts; the lowest IFNγ levels were observed for P2rx7-/- males . The IL-10+CD4+ cell numbers were comparable in infected B6 and P2rx7-/- mice . The immunoglobulin secretion was also reduced in acutely infected P2rx7-/- mice compared with that in their B6 counterparts , characterized by a predominance of IgM- and IgG2c-secreting cells ( Fig 2G ) as previously reported [36] . Together , this data shows that P2X7 signaling at the time of Pc-iRBC rupture is sufficient to induce pore formation in blood CD4 T cells . They also reveal the importance of the P2X7 receptor for optimal CD4 T cell function during acute Pc infection . The low CD4 T cell proliferation in acutely infected P2rx7-/- mice could explain their reduced IFNγ and antibody responses . However , P2X7 deficiency could also impair Th1 cell differentiation in addition to preventing CD4 T cell expansion . To evaluate this possibility , we next examined whether the P2X7 receptor influences Th1/Tfh cell differentiation during early Pc infection . Splenic CD4 T cells were analyzed according to the expression of the transcription factors T-bet and B cell lymphoma 6 ( Bcl6 ) , which are reciprocal regulators of Th1 and Tfh cell lineage commitment [37] , [38] . At day 4 p . i . , a small percentage increase in CD4 T cells expressing both T-bet and Bcl6 was observed in B6 and P2rx7-/- mice ( S3A Fig ) . By contrast , around 40% of the B6 CD4 T cells at 7 days p . i . showed high amounts of these transcription factors , evidenced by two distinct cell subsets ( Fig 3A and S3B Fig ) . The proportion of T-bethiBcl6lo cells was reduced in the absence of the P2X7 receptor , while that of T-betloBcl6hi cells was augmented . However , considering the low numbers of CD4 T cells per spleen in P2rx7-/- mice at day 7 p . i . ( Fig 2E ) , P2X7 deficiency led to a decrease in the T-bethiBcl6lo cell population , but did not affect the T-betloBcl6hi cell population . Both CD4 T cell subsets showed higher expression of programmed cell death-1 ( PD1 ) and C-X-C chemokine receptor type 5 ( CXCR5 ) than CD4 T cells from non-infected mice ( Fig 3B ) . However , PD1 expression in these cells at day 7 p . i . was lower than that in fully differentiated Tfh cells at 20 days p . i . ( S3C Fig ) . Both T-bethiBcl6lo and T-betloBcl6hi cells from infected B6 mice also expressed high levels of the transcription factor B lymphocyte-induced maturation protein-1 ( Blimp-1 ) ( Fig 3C , left panels ) , which is a strong negative regulator of Tfh cell differentiation [37] . Lower Blimp-1 expression was observed for both CD4 T cell subsets in the absence of P2X7 receptor . We also assessed the expression of IL-2 receptor α ( CD25 ) and β ( CD122 ) chains; effector memory Th1 cells are produced from early CD4 T cell population expressing this receptor in L . monocytogenes infection [39] . Both CD25 and CD122 were preferentially increased in T-bethiBcl6lo cells than in T-betloBcl6hi cells from infected B6 mice; these molecules were expressed at lower levels in T-bethiBcl6lo cells from infected P2rx7-/- mice ( Fig 3C , middle and right panels ) . At 7 days p . i . , the Foxp3+CD4 T cell population showing high CD25 and CD122 levels was also reduced in P2rx7-/- mice compared with that in B6 mice ( S3D and S3E Fig ) . Because the P2X7 receptor significantly affected the Th1 and Tfh cell counterbalance during acute Pc malaria , we then evaluated the P2X7 expression in B6 CD4 T cells . We also assessed the expression of CD39 ecto-nucleotidase , which can down-regulate P2X7 signaling by degrading eATP [18] . The P2X7 and CD39 levels were increased at 4 days p . i . in T-bet+Bcl6+ cells but not in T-bet-Bcl6- cells ( S3F Fig ) . Remarkably , at 7 days p . i . , the P2X7 receptor was mostly expressed in T-betloBcl6hi cells , whereas T-bethiBcl6lo cells exhibited considerably more CD39 ( Fig 3D ) . In fact , two different CD4 T cell populations were identified at 7 days p . i . according to P2X7 and CD39 expression: the P2X7loCD39hi ( T-bethiBcl6lo ) and P2X7hiCD39lo ( T-betloBcl6hi ) subsets ( Fig 3E ) . In summary , P2X7 deficiency impairs the Th1 cell differentiation from the onset of Pc infection . This effect was counterbalanced by a percentage increase of T-betloBcl6hi CD4 T cells . The early increase in the proportion of T-betloBcl6hi CD4 T cells in infected P2rx7-/- mice motivated a careful examination of the development of Tfh cells during chronic Pc malaria . First , we observed hyperplasia of splenic secondary lymphoid follicles in the absence of the P2X7 receptor ( Fig 4A ) . Confocal microscopy analysis of these follicles at 20 days p . i . revealed many CD4 T cells in extensive GL7-stained germinal centers ( Fig 4B ) . Consistently , at 20 and 30 days p . i . , P2rx7-/- mice showed higher B cell numbers per spleen , including germinal center ( Fas+GL7+CD19+ ) B cells ( Fig 4C ) . P2X7 deficiency also led to the development of an increased Tfh cell population that was identified by the expression of inducible T cell co-stimulation ( ICOS ) , CXCR5 , PD1 and Bcl6 ( Fig 4D ) . Remarkably , a prominent Tfh cell response occurred at 14 days p . i . in P2rx7-/- mice yielding an earlier peak of Tfh cell numbers than in B6 mice; the Tfh cell population remained augmented in P2rx7-/- mice until day 100 p . i . ( Fig 4E ) . We also evaluated the IL-21 production , which is a Tfh cell signature and contributes to functional Tfh cell generation in Pc malaria [33] . Following stimulation in vitro with iRBCs , P2rx7-/- splenocytes at 20 days p . i . produced higher amounts of IL-21 than B6 counterparts ( Fig 4F ) . Furthermore , the serum concentrations of anti-parasite IgM at 30 days p . i . and IgG2c at 50 days p . i . were higher in P2rx7-/- mice than in B6 mice ( Fig 4G ) . This data suggests that P2X7 signaling controls the Tfh cell population and anti-parasite antibody production during chronic Pc infection . The higher parasitemias observed in chronically infected P2rx7-/- mice compared with those of their B6 counterparts suggested that the P2X7 receptor is required for the development of acquired immunity to Pc malaria . Because of the key role of Th1 cells in complete parasite elimination [25]-[27] , the splenic CD4 T cell response was also analyzed in chronically infected B6 and P2rx7-/- mice . The amounts of additional ATP required to induce P2X7-mediated pore formation in B6 CD4 T cells were still lower at 20 and 30 days p . i . compared to that in non-infected mice , returning to control levels at 50 days p . i . ( Fig 5A ) . Unlike in acute infection , higher CD4 T cell numbers per spleen were observed in P2rx7-/- mice than in B6 mice during chronic infection ( Fig 5B ) . The analysis of CD4 T cell subsets , as previously defined [26] , revealed larger populations of effector ( TE ) , effector memory ( TEM ) and central memory ( TCM ) cells in chronically infected P2rx7-/- mice ( S4A Fig ) . Despite these increases , the lack of the P2X7 receptor impaired IFNγ secretion , without affecting IL-10 secretion , as observed during acute Pc infection ( Fig 5C ) . Most IFNγ- and IL-10-producing cells exhibited a TEM cell phenotype at 20 days p . i . ( S4B Fig ) . Concordantly , reduced amounts of IFNγ , but not of IL-10 , were detected in the supernatants of iRBC-stimulated P2rx7-/- splenocytes at 30 days p . i . ( Fig 5D ) . Furthermore , ATP stimulation of B6 splenocytes boosted P2X7-dependent IFNγ production and inhibited P2X7-independent IL-10 production . Subsequently , we investigated whether the P2X7 receptor is required to induce and/or maintain T-bet expression in splenic CD4 T cells from chronically infected mice . CD4 T cells from B6 mice maintained high expression of this transcription factor at 20 and 30 days p . i . while P2rx7-/- CD4 T cells did not ( Fig 5E ) . Furthermore , B6 CD4 T cells at 20 days p . i . co-expressed T-bet and P2X7 receptor ( Fig 5F ) . At 30 days p . i . , T-bet up-regulation was observed in B6 CD4 TE/EM and TCM cells but not in P2rx7-/- counterparts ( S5A Fig ) . The B6 TE/EM and TCM cells also exhibited high P2X7 levels at day 20 p . i . ( S5B Fig ) . These results indicate that P2X7 receptor is required for the Th1 cell response during chronic Pc malaria . One explanation for our previous results is that P2X7 signaling promotes Th1 cell differentiation to the detriment of Tfh cell differentiation during blood-stage Pc malaria by inducing the expression of T-bet , rather than Bcl6 . Alternatively , but not mutually exclusive , P2X7 signaling could mediate splenic Tfh cell death as described for Peyer’s patches [16] . Therefore , we examined whether splenic Tfh cells at day 20 p . i . were susceptible to ATP-induced cell death . Using the terminal deoxynucleotidyl transferase dUTP nick end labeling ( TUNEL ) assay [40] , we identified apoptotic cells mostly in the spleen of B6 mice than of P2rx7-/- mice at 20 days p . i . ( Fig 6A ) . Aggregates of TUNEL+ cells were seen in the lymphoid follicles . Furthermore , at 20 days p . i . , lower percentages of hypodiploid nuclei were detected in P2rx7-/- CD4 T cells than in B6 CD4 T cells ( Fig 6B ) . Because phosphatidylserine ( PS ) exposure precedes P2X7-mediated T cell death [15] , annexin V staining was also compared in Tfh and non-Tfh cells . In infected B6 mice , most Tfh cells were labeled with annexin V , whereas non-Tfh cells were mainly negative ( Fig 6C ) . Lower annexin V staining was observed in P2rx7-/- Tfh cells than in B6 Tfh cells , but Tfh cells still presented higher labeling than non-Tfh cells in the absence of P2X7 receptor . Based on the results of Aqua Live/Dead staining , most annexin V+ Tfh cells maintained membrane integrity ( Fig 6D ) . However , after 2 h in culture , a large proportion of B6 Tfh cells died spontaneously as evidenced by the co-expression of annexin V and Aqua Live/Dead staining ( Fig 6E ) . This process was mediated by the P2X7 receptor and increased in the presence of 300 μM ATP . In contrast , most non-Tfh cells remained alive even following ATP stimulation . Additionally , ATP-induced P2X7-mediated pore formation was observed mostly in Tfh cells compared with non-Tfh cells ( Fig 6F ) . In order to determine the mechanism responsible for the greater sensitivity of Tfh cells to eATP , we evaluated P2X7 and CD39 expression in Tfh and non-Tfh cells from chronically infected B6 mice . At 20 days p . i . , Tfh cells exhibited higher levels of P2X7 receptor than non-Tfh cells and TE/EM cells ( Fig 6G ) , which contained both non-Tfh cells and Tfh cells ( S5C Fig ) . Of note , a direct correlation between P2X7 and PD1 expression seems to occur in B6 CD4 T cells . Furthermore , two distinct CD39+ CD4 T cell subsets were identified in chronically infected B6 mice; PD1hiCXCR5+CD39lo ( Tfh ) cells expressed more P2X7 receptor than CXCR5-CD39hi ( non-Tfh ) cells ( Fig 6H ) . Because T-betloBcl6hi cells at day 7 p . i . also expressed a P2X7hiCD39lo phenotype ( Fig 3G ) , we wondered if the early stages of Tfh cell differentiation were also susceptible to ATP-induced cell death . In fact , a higher proportion of T-betloBcl6hi cells died spontaneously after 2 h in culture compared with that of T-bethiBcl6lo cells , a phenomenon that was also observed in the presence of 300 μM ATP . ( Fig 6I ) . These results indicate that P2X7-mediated cell death contributes to control the Tfh cell response during chronic Pc malaria . The high expression of the P2X7 receptor concomitantly with low CD39 levels explains the great sensitivity of the Tfh cell lineage to eATP . To verify whether P2X7 expression in CD4 T cells is required for Th1 cell differentiation during Pc malaria , splenic CD4 T cells were sorted from naïve B6 or P2rx7-/- donors and transferred into Cd4-/- recipients that were infected with Pc parasites a week later ( Fig 7A ) . The treatment with subcurative doses of chloroquine controlled parasitemias at comparable levels in both mouse groups ( Fig 7B ) . A similar increase in CD4 T cell numbers per spleen was observed in these mice at 30 days p . i . ( S6A Fig ) . Nevertheless , P2rx7-/- CD4 T cells generated lower numbers of TEM and TCM cells than B6 CD4 T cells ( S6B Fig ) . Demonstrating the role of P2X7 signaling for the early CD4 T cell activation during acute Pc malaria , higher cytosolic calcium levels were observed at 7 days p . i . in B6 CD4 T cells compared with P2rx7-/- CD4 T cells ( Fig 7C ) . Impaired IL-2 and IFNγ production accompanied the lower cytosolic calcium levels in P2rx7-/- CD4 T cells ( Fig 7D and 7E ) . In line with these findings , at 7 and 30 days p . i . , T-bet up-regulation was observed in B6 CD4 T cells but not in P2rx7-/- CD4 T cells ( Fig 7F ) . Supporting the concept that P2X7 signaling changes the Th1/Tfh cell balance during Pc malaria , the Tfh cell population was significantly increased when CD4 T cells lacked the P2X7 receptor ( Fig 7G ) . Next , we examined how B6 and P2rx7-/- CD4 T cells differentiate in the same environment after Pc infection . Therefore , Cd4-/- mice co-transferred with B6 ( CD45 . 1 ) and P2rx7-/- ( CD45 . 2 ) CD4 T cells were infected with Pc parasites a week later ( Fig 8A ) . At 30 days p . i . , similar numbers of B6 and P2rx7-/- CD4 T cells were observed in the spleen ( Fig 8B ) . Consistent with our previous results , the proportion of PD1loCD39hi CD4 T cells was greater in the B6 population , whereas PD1hiCD39lo CD4 T cells predominated in the P2rx7-/- population ( Fig 8C ) . For the B6 and P2rx7-/- populations , PD1loCD39hi and PD1hiCD39lo CD4 T cell subsets showed higher T-bet and Bcl6 expression , respectively ( Fig 8D ) . In conclusion , P2X7 expression in CD4 T cells is sufficient to promote Th1 cell differentiation over Tfh cells during Pc malaria . Our previous results indicate that P2X7 expression in CD4 T cells is critical for generating a balanced Th1/Tfh cell population during chronic Pc malaria , while an increased Tfh cell population develops in the absence of P2X7 receptor . To assess the consequences of that for protective immunity , splenic CD4 T cells from B6 or P2rx7-/- mice at day 20 p . i . were transferred into Cd4-/- recipients that were infected with Pc parasites ( Fig 8E ) . Cd4-/- mice transferred with naïve B6 CD4 T cells were used as controls . Significantly lower first and second parasitemia peaks were observed in Cd4-/- mice transferred with B6 CD4 T cells at day 20 p . i . compared with those transferred with P2rx7-/- CD4 T cells at day 20 p . i . ( Fig 8F ) . Notably , Cd4-/- mice transferred with naïve B6 CD4 T cells or P2rx7-/- CD4 T cells at day 20 p . i . failed to control recrudescent parasitemia up to a month of infection . We concluded that the balanced Th1/Tfh cell population that develops in infected B6 mice is more efficient in controlling the first and second parasitemia peaks than the increased Tfh cell population generated in infected P2rx7-/- mice
The P2X7 receptor has been implicated in both the protection and exacerbation of infectious diseases by inducing the activation and death of infected macrophages [41–43] . An important concept that emerges from our study is that eATP recognition by P2X7 receptor promotes the differentiation of Th1 cells to the detriment of Tfh cells during Pc malaria and thus contributes to protection . The first evidence in this direction was the susceptibility pattern of P2rx7-/- mice that closely resembled the one previously described for IFNγ deficiency , in which most males died during acute Pc infection and females developed increased acute and chronic parasitemia [23] . The immunosuppressive effects of testosterone in the production of and response to cytokines were implicated in the higher susceptibility of male IFNγ-/- mice to Pc parasites , a phenomenon that was recapitulated in P2rx7-/- males due to their extremely low Th1 response . The eATP concentrations following iRBC rupture achieved , at least in some microenvironments , the threshold required for P2X7 activation in CD4 T cells , as observed for other pathological processes [44] , [45] . Furthermore , splenic CD4 T cells became highly responsive to eATP at the beginning of Pc infection , which may have contributed to amplifying the early lymphocyte response . Indeed , the P2X7 receptor was important for CD4 T cell proliferation and IFNγ production during acute Pc malaria and consequently for IgM and IgG2c secretion that depends , at least partially , on CD4 T cell help [24] . Both IFNγ and low-affinity antibodies participate in the control of acute Pc malaria [22] , [23] , [46] . The P2X7 expression was also critical for IFNγ production during chronic disease . This finding corroborates our previous studies suggesting that a Th1 response is required , together with anti-parasite antibodies , to resolve a persistent Plasmodium infection [25] , [27] . Moreover , the inflammatory response may be further potentiated during Pc infection by the inhibitory effect of eATP on IL-10 secretion , which was shown here to be independent of the P2X7 receptor . In addition to the important contribution of the P2X7 receptor in boosting the CD4 T cell response during acute Pc malaria , P2X7 signaling was critical for driving Th1/Tfh cell differentiation . On day 4 p . i . , CD4 T cells had apparently not yet been committed on Th1 and Tfh cell lineages because there was only a small population expressing both T-bet and Bcl6 . However , the two subsets could clearly be distinguished at 7 days p . i . by the preferential expression of T-bet , CD25 , CD122 and CD39 in Th1-biased cells and Bcl6 and P2X7 receptor in Tfh-biased cells; P2X7 deficiency reduced the proportion of T-bethiBcl6lo cells and augmented that of T-betloBcl6hi cells . In line with these findings , a single-cell RNA seq analysis of Th1/Tfh cell differentiation during acute Pc malaria showed that these subsets emerge in parallel by day 7 p . i . ; Entpd1 ( CD39 ) and P2rx7 expression was identified as Th1 and Tfh signature , respectively [47] . To explain our data showing low antibody production in acutely infected P2rx7-/- mice , we considered the possibility that T-bethiBcl6lo and T-betloBcl6hi cell subsets could provide help for B cells at the early Pc infection; P2X7 deficiency reduced only the T-bethiBcl6lo cell population . Supporting this idea , both subsets displayed an early Tfh phenotype expressing CXCR5 and PD1 at levels below those of fully differentiated Tfh cells . During chronic Pc malaria , there was a sharp expansion of the Tfh cell population in P2rx7-/- mice that resulted in higher serum levels of anti-parasite IgM and IgG2c than in the B6 counterparts . The persistence of Th1 bias in chronically infected B6 mice was evidenced by T-bet and P2X7 co-expression in TE/EM and TCM cells; however , CD4 T cells from P2rx7-/- mice did not express this transcription factor . Of note , the selective absence of the P2X7 receptor in CD4 T cells was sufficient to change the Th1/Tfh cell balance . This effect was shown in mice transferred with either P2rx7-/- or B6 CD4 T cells in which parasitemia was maintained at similar levels by drug treatment , as well as in mice co-transferred with both CD4 T populations . In both experimental conditions , there was no apparent competitive advantage between the P2rx7-/- and B6 CD4 T cell populations that displayed similar sizes a month after infection . P2X7 deficiency in CD4 T cells resulted in a lower increase in the TE , TEM and TCM cell populations a month after infection . The opposing effect obtained in chronically infected P2rx7-/- mice may be a consequence of higher parasitemia found in these animals in relation to chronically infected B6 mice . The P2X7 signaling can influence Th1/Tfh cell differentiation by inducing the T-bet-controlled Th1 cell program , which hinders the development of the Bcl6-controlled Tfh cell program . In fact , P2X7 deficiency led to lower expression of Blimp-1 in CD4 T cells and this transcription factor is a known antagonist of Bcl6 [37] . It is generally accepted that P2X7 signaling in T cells amplifies T cell receptor ( TCR ) -induced calcium influx and thus increases IL-2 secretion [48] , [49] . Accordingly , calcium influx and IL-2 secretion were dependent on P2X7 expression in CD4 T cells from acute Pc malaria . Furthermore , it has been shown that eATP is required for IL-2 and IFNγ secretion by antigen-specific T cells [50] . IL-2 induces the signal transducer and activator of transcription 5 ( STAT5 ) in CD4 T cells that up-regulates the IL-12 receptor β2-chain , T-bet and Blimp-1 [51] , [52] . Moreover , IL-2-mediated activation of the mammalian target of rapamycin complex 1 ( mTORc1 ) kinase axis up-regulates Blimp-1 expression and shifts differentiation away from Tfh cells , instead promoting that of Th1 cells [53] . Another evidence that IL-2 induced by P2X7 signaling influences the CD4 T cell response to Pc malaria was the lower increase of CD4 T cells expressing IL-2 receptor α- and β-chains in infected P2rx7-/- mice compared with the B6 counterparts . Other cytokines can induce T-bet expression , such as IL-12 , IL-27 and IFNα [38] , [54] , but the relationship of these signaling pathways with P2X7 receptor is unclear . An alternative non-exclusive molecular mechanism by which the P2X7 receptor can change the Th1/Tfh balance relies on the high susceptibility of Tfh cells to ATP-induced cell death . Suggesting that this mechanism operates during Pc infection , P2X7 deficiency reduced the apoptotic cell death in germinal centers and PS exposure in Tfh cells . Furthermore , Tfh cells are particularly prone to die spontaneously through the P2X7 receptor . The relatively low CD39 expression may be insufficient to degrade eATP rapidly and thus prevent ATP-induced Tfh cell death , which can be accelerated by the extremely high levels of the P2X7 receptor in these cells . Similarly , in mouse Peyer’s patches and human tonsils , the P2rx7 gene is highly expressed in Tfh cells that are particularly responsive to eATP and undergo P2X7-mediated cell death [16] . In this study , as in ours , Tfh cell death was evaluated in B6 mice , which have an allelic mutation in the predicted death domain of P2X7 receptor that reduces eATP sensitivity [55] . Macrophages and lymphocytes from B6 mice respond to ATP stimulation and undergo P2X7-mediated cell death [16] , [43] , [56] , although they are more resistant than cells from other mouse strains [55] . Therefore , the effects of eATP are expected to be greater on Tfh cells expressing the unmutated P2rx7 gene . Regarding how splenic Tfh cells come into contact with eATP during Pc malaria , it is likely that the absence of compartmentalization between the white and red pulp observed at the acute infection makes available large amounts of locally released ATP from lysed iRBCs [57] . Lymphocytes stressed or dying as a result of the affinity maturation process , as well as lysed iRBCs in the marginal zone and in the vessels that irrigate the germinal centers , can be the sources of eATP during chronic Pc malaria . The detailed analysis of the P2X7 receptor's role in the immune response to Pc infection allowed us to reveal the importance of this molecule in the fine-tuning between Th1 and Tfh cell populations . It is generally accepted that antibody affinity maturation and memory B cells develop properly in Pc malaria [58] , [59] . Additionally , functional Tfh cells are necessary for an efficient antibody response and resolution of chronic Pc parasitemia [33] . Nevertheless , some control of Tfh cell differentiation through the P2X7 receptor appears to be required to generate the Th1 response to Pc infection , thus improving the disease outcome . Supporting this idea , mice transferred with B6 CD4 T cells at 20 days p . i . containing both Th1 and Tfh cells showed a better control of the first and second parasitemia peaks than those transferred with the P2rx7-/- counterparts where there is a marked predominance of the Tfh phenotype . Of note , P2X7 deficiency led to better control of P . yoelii 17XNL malaria whose outcome seems to depend more on anti-parasite antibodies and less on IFNγ than in Pc malaria [32] , [35] . A feasible explanation for this finding is that Pc merozoites , when released , rapidly infect nearby erythrocytes , but free P . yoelii 17XNL parasites are required to remain longer in the extracellular environment to find few reticulocytes , thereby making them easy targets for antibodies . In apparent contradiction to our results , the expansion of the Tfh cell population by inhibiting PD1 and lymphocyte activation gene ( LAG ) -3 , or in mice deficient in IFNα receptor 1 , improves the control of both Pc and P . yoelii 17XNL infections [32] , [34] . However , the effect of Th1/Tfh imbalance cannot be appreciated under these conditions because IFNγ production is also increased . We propose here that elevated eATP in the spleen , which may occur during Plasmodium infection , might modify the CD4 T cell balance toward a more potent inflammatory response and hinder the antibody response . This concept is in line with a recent study showing that the extremely intense inflammatory response during severe Plasmodium berghei ANKA infection drastically inhibits Tfh cell differentiation [60] . Thus , our study goes a step further in understanding malaria pathogenesis , suggesting that continuous ATP release and P2X7 signaling might not only promote the Th1 response but also delay antibody production during Plasmodium infection . Our study provides mechanistic insights into malaria pathogenesis by demonstrating the importance of damage signals for the outcome of the disease . By promoting a balanced Th1/Tfh cell response , the immune system becomes more effective in protecting against Pc parasites whose control is based on both IFNγ and anti-parasite antibodies . It remains unclear whether the development of a robust Tfh cell response as a result of P2X7 deficiency is responsible for increasing the resistance to P . yoelii 17XNL infection . Human CD4 T cells express P2X7 receptor and secrete IL-2 following eATP stimulation [48] , but it is still unknown whether P2X7 signaling influences the CD4 T cell response in malaria patients . Supporting the role of P2X7 signaling in controlling Tfh cell numbers in humans , tuberculosis patients carrying a single loss-of-function P2rx7 gene polymorphism produce more IgG against mycobacteria than control groups [61] . This knowledge raises the possibility that the ATP-P2X7 axis can be manipulated with P2X7 agonists and antagonists to change the Th1/Tfh cell balance , aiming to ameliorate pathological conditions or to improve immunization protocols .
Six-to-eight-week-old C57BL/6 ( B6 ) , B6 . SJL-Ptprca Pepcb/BoyJ ( CD45 . 1+/+ ) , B6 . 129P2-P2rx7 tm1Gab/J ( P2rx7-/- ) , B6 . Cg-Fosn1nu/J ( nude ) and B6 . 129S2-Cd4tm1Mak/J ( Cd4-/- ) female and male mice ( originally from The Jackson Laboratory ) were bred under specific pathogen-free conditions at the Isogenic Mice Facility ( ICB-USP , Brazil ) . The P2rx7-/- mice were generated by Pfizer Inc . ( USA ) ; a single nucleotide polymorphism panel analysis throughout the genome suggested a B6 genetic background . The experiments were performed in female mice with the exception that females and males were compared . Pc and P . yoelii 17XNL parasites were maintained as described elsewhere [62] . Because the Pc schizogonic cycle depends on the host circadian rhythm , the mice were maintained under an inverted light/dark cycle for at least 15 days before infection to access the period adjacent to erythrocyte invasion [63] . Mice were infected intraperitoneally with 1 × 106 iRBCs . All experimental procedures were in accordance with national regulations of ethical guidelines for mouse experimentation and welfare of the Health National Council and Animal Experimentation Brazilian College ( COBEA ) —Brazil , the protocols being approved by the Health Animal Committee of USP , with permit numbers 050/2009 and 175/2011 . Parasitemias were monitored by microscopic examination of Giemsa-stained blood smears . Body weight variation was determined with respect to the day 0 weight with an analytical balance ( Sartorious , USA ) . Axial temperature was assessed with a digital thermometer ( Kent Scientific Co . , USA ) . Hemoglobin serum concentration was evaluated with a hemoglobin kit ( Doles Inc . , Brazil ) . Blood and spleen cells were washed and maintained in cold RPMI 1640 supplemented with penicillin ( 100 U/ml ) , streptomycin ( 100 μg/ml ) , 2-mercaptoethanol ( 50 μM ) , L-glutamine ( 2 mM ) , sodium pyruvate ( 1 mM ) and 3% heat-inactivated fetal calf serum . All supplements were purchased from Life Technologies ( USA ) . Leukocytes were obtained in 70% Percoll gradient ( GE Health Care , USA ) . For the cell proliferation assay , spleen CD4 T cells were magnetically purified by negative selection . Non-CD4 T cells were labeled with biotinylated antibodies and streptavidin-coated magnetic particles and then were separated using an EasySep magnet ( Stem Cell Technologies , Canada ) . For the calcium flux assay , spleen CD4 T cells were magnetically purified ( LS columns ) by positive selection using anti-CD4 microbeads with autoMACS ( Miltenyi Biotec , Germany ) . For adoptive transfer , spleen CD4 T cells were magnetically purified ( LS columns ) by negative selection using anti-CD19 , -IAb and -CD8 microbeads with autoMACS ( Miltenyi Biotec , Germany ) and then were sorted using a FACS Aria device ( BD Biosciences , USA ) . In some experiments , spleen CD4 T cells were magnetically purified by negative selection using an EasySep magnet and then were sorted using a FACS Aria device . Cells ( 1 × 106 ) were stained with FITC- , PE- , APC- , PerCP- , PECy7- , APC Cy7- , Pacific Blue- , or AmCyan-labeled monoclonal antibodies ( mAbs ) ( BD Biosciences ) to CD4 ( H129 . 19 or GK1 . 5 ) , CD19 ( ID3 ) , CD25 ( PC61 ) , CD39 ( 24DMS1 ) , CD44 ( IM7 ) , CD62L ( MEL-14 ) , CD122 ( TM-β1 ) , CD178 ( ICOS ) ( 7E . 17G9 ) , CD127 ( A7R34 ) , GL7 ( GL7 ) , PD1 ( J43 ) , CXCR5 ( 2G8 ) and P2X7 ( 1F11 ) . For the detection of intracellular staining , PE-labeled mAb to Bcl-6 ( K112-91 , BD Biosciences ) , PerCP-labeled mAb to T-bet ( eBio4B10; eBioscience , USA ) , PE-labeled mAb to Blimp-1 ( 6D3; BD Biosciences ) and APC-labeled mAb to Foxp3 ( FJK-16s; eBioscience ) were used according to the manufacturer’s instructions . PE-labeled rat IgG1 ( BD Bioscience ) and PerCP-labeled anti-CD45 . 1 mAb ( A20; BD Bioscience ) were used as isotype controls . Annexin V staining was performed in the appropriate binding buffer ( 10 mM HEPES , 150 mM NaCl , 5 mM KCl , 1 mM MgCl2 , 1 . 8 mM CaCl2 [pH 7 . 4] ) . Cells were analyzed by flow cytometry using a FACSCanto device with DIVA software ( BD Biosciences ) . Data were analyzed with FlowJo software v . 7 . 2 . 2 ( Tree Star Inc . , USA ) . Blood and spleen cells ( 1 × 106 ) were stained with APC-labeled anti-CD4 mAb ( BD Biosciences ) . In some experiments , stained cells were pre-warmed ( 37°C ) in phosphate-buffered saline with 3% bovine serum albumin ( Sigma-Aldrich ) and then incubated with 25–500 μM ATP ( Amersham Bioscience , USA ) , lysed iRBC supernatant or medium alone for 15 min . The non-infected RBCs ( nRBCs ) and iRBCs ( 2 × 108 ) were lysed with 200 μl of lysis buffer ( 40 mM NH4Cl , 4 . 2 mM Tris [pH 7 . 4] ) for 5 min at 4°C . Spleen cells ( 1 × 106 ) were incubated with 200 μl of RBC supernatants diluted 1:5 in cold RPMI with 1% heat-inactivated fetal calf serum ( FCS ) . The fluorescent 2 . 5 μM EB dye ( Sigma-Aldrich ) was added , and the samples were immediately analyzed by flow cytometry . ATP concentrations were determined using an ATP bioluminescence assay kit ( Sigma-Aldrich ) . Serum ( 50 μl/well ) was mixed 1:1 with the luciferase reagent . The bioluminescence was quantified in a temperature-controlled luminometer ( Berthold , USA ) . Purified CD4 T cells ( 3 × 107 ) were incubated for 20 min at 37°C with 5 μM 5 , 6-carboxyfluorescein succinimidyl ester ( CFSE; Molecular Probes , USA ) in phosphate-buffered saline ( PBS ) with 0 . 1% bovine serum albumin ( BSA , Sigma-Aldrich ) . CD4 T cells ( 5 × 105 ) were cultured with iRBCs ( 4 × 106 ) or medium alone , in the presence of spleen cells ( 5 × 105 ) from nude mice as a source of APCs , for 72 h at 37°C in a 5% CO2 atmosphere , stained with PECy7-labeled anti-CD4 mAb and analyzed by flow cytometry . In the experiments using apyrase or BBG , spleen cells ( 3 x 107 ) were stained with CFSE as described above . Cells ( 1 x 106 ) were cultured with iRBCs ( 4 x 106 ) in the presence or absence of apyrase ( 20 U/ml ) or BBG ( 35 μM ) for 72 h at 37°C in a 5% CO2 atmosphere , stained with PE-labeled mAb to CD4 and analyzed by flow cytometry . For intracellular ex vivo detection , spleen cells ( 1 × 106 ) were cultured with GolgiStop reagent ( containing monensin ) according to the manufacturer's instructions for 6 h at 37°C in a 5% CO2 atmosphere . For intracellular in vitro detection , spleen cells ( 1 × 106 ) were cultured with iRBCs ( 4 × 106 ) or medium alone for 72 h at 37°C in a 5% CO2 atmosphere . The GolgiStop reagent was added at the last 6 h of culture according to the manufacturer's instructions . After washing , cells were surface stained with APC- or Pacific Blue-labeled mAbs to CD4 . Cells were then fixed with Cytofix/Cytoperm buffer , stained with PE-labeled mAb to IFNγ ( XMG-1 . 2 ) and APC-labeled mAb to IL-10 ( JESS-16E3 ) diluted in Perm/Wash buffer , and analyzed by flow cytometry . All reagents were purchased from BD Biosciences . The IL-2 cytokine secretion assay was performed according to the manufacturer's instructions ( BD Biosciences ) . This assay uses a bi-functional mAb capable of binding CD45 and IL-2 . Cells ( 2 × 106 ) were incubated with the bi-functional mAb for 45 min at 37°C in 5% CO2 atmosphere . The IL-2 bound to the surface of cells was detected with PE-labeled anti-IL-2 mAb by flow cytometry . For supernatant cytokine detection , spleen cells ( 1 × 106 ) were cultured with 3 × 106 iRBCs in the presence or absence of apyrase ( 20 U/ml ) , BBG ( 35 μM ) or medium alone for 72 h at 37°C in a 5% CO2 atmosphere . Cytokine concentrations were determined using the OptEIA IFNγ kit ( BD Biosciences ) , OptEIA IL-10 kit ( BD Biosciences ) and mouse IL-21 ELISA ( eBioscience ) . The anti-Pc IgM , IgG1 and IgG2c serum levels were quantified by ELISA as described elsewhere [64] . Briefly , 96-well flat-bottom microtest plates ( Costar , USA ) were coated overnight at 4°C with 8 μg/ml of a total Pc extract and saturated with 1% BSA for 3 h . After washing , 100 μl of mouse serum samples ( diluted from 1/10 to 1/1 , 280 ) were added and left overnight at 4°C . Antibody concentrations were determined using Ig standards . The assays were developed by adding goat anti-mouse Ig isotype peroxidase–conjugated antibodies ( Southern Biotechnology Associates , USA ) for 45 min , followed by the addition of 100 μl of tetramethylbenzi-dine ( Invitrogen , USA ) . Absorbance was measured at 650 nm with an Epoch Microplate Spectrophotometer ( BioTek , USA ) . Ig-producing cells were quantified by the ELISPOT assay as described elsewhere [65] . In brief , 96-well flat-bottom microtest plates ( Costar ) were coated overnight at 4°C with 10 μg/ml of goat-anti-mouse total Ig and saturated with 1% gelatin ( Merck , Germany ) in PBS for 120 min . Spleen cells ( 1 × 106 to 5 × 102 cells/well ) were cultured for 6 h at 37°C in a 5% CO2 atmosphere . The spots were developed by adding goat anti-mouse Ig isotype biotinylated antibodies overnight , followed by the addition of a phosphatase alkaline-avidin conjugate . All antibodies and conjugates were purchased from Southern Biotechnologies Associates . 5-Bromo chloro 3-indolyl phosphate ( BCIP; Sigma-Aldrich ) diluted in 2-amino 2-methyl 1-propanol ( AMP , Merck ) was used as a substrate . CD4 T cells ( 1 × 107 ) were loaded with a mixture of 4 μM Fura-3AM ( Molecular Probes , USA ) and 0 . 7 mg/ml of Probenecid ( Sigma-Aldrich ) at 37°C for 30 min . After washing , cells ( 1 × 105 ) were analyzed in a fluorescence microscope ( Nikon Inverted Microscope , Japan ) to determine the fluorescence intensity . Intracellular calcium was determined by calculating the corrected total cell fluorescence ( CTCF ) . CTCF = integrated density— ( area of selected cell × mean fluorescence of background readings ) . Spleens were harvested and frozen in Tissue-Tek O . C . T . ( Sakura Finetek , USA ) . In situ DNA fragmentation in 8-μm-thick slices was performed using the DeadEnd Fluorometric TUNEL System according to the manufacturer’s protocol ( Promega , USA ) . APC-labeled anti-CD4 mAb stained cells ( 1 × 106 ) were fixed with 70% ethanol . After washing , cells were incubated in DNA extraction buffer ( 0 . 2 M Na2HPO4 and 0 . 1% Triton x-100 , Sigma-Aldrich ) at 24°C for 5 min , centrifuged and resuspended in a DNA staining solution ( 20 μg/ml PI , Sigma-Aldrich ) . RNase ( 50 μg ) ( Invitrogen ) was then added to each sample , and the cells were incubated at 24°C for 30 min in the dark . PI incorporation was determined by flow cytometry . Spleens were fixed in buffered formol for 12 h and paraffin-embedded . Splenic tissue sections ( 5 mm ) were hematoxylin-eosin stained using standard procedures . Spleens were harvested and immediately embedded in Tissue-Tek O . C . T . ( Sakura Finetek , Japan ) and snap frozen . Slices measuring 8 μm thick were fixed in acetone and blocked with 3% BSA plus Fc Block ( 1:100 ) ( BD Bioscience ) for 1 h at 24°C . Slides were then stained with anti-CD4-Alexa 700 ( RM4-5 ) ( eBioscience ) , CD19-APC ( ID3 ) , GL-7-Biotin ( GL7 ) and Streptavidin-FITC ( BD Biosciences ) for 2 h at 24°C . After washing , slides were stained with 4' , 6-diamidino-2-phenylindole ( DAPI; Molecular Probes ) and mounted with Vectashield mounting medium ( Vector , USA ) . Sections were analyzed by confocal microscopy ( Zeiss LSM 780 , Germany ) . For CD4 T-cell phenotypic analysis , Cd4-/- mice adoptively transferred i . v . with purified CD4 T cells ( 1 . 5 × 106 ) from naïve B6 or P2rx7-/- mice were treated i . p . from day 7 p . i . with 3 every-other-day doses of chloroquine ( 10 mg/kg of body weight/day ) . Alternatively , Cd4-/- mice co-transferred i . v . with purified B6 and P2rx7-/- CD4 T cells ( 1 × 106/each population ) from naïve mice were treated i . p . from day 7 p . i . with 3 consecutive daily doses of chloroquine ( 10 mg/kg of body weight/day ) . For parasitemia analysis , B6 or P2rx7-/- CD4 T cells ( 1 . 5 × 106 ) at day 20 p . i . were transferred i . v . to Cd4-/- mice . Statistical analysis was performed by the Mann-Whitney U test to compare two groups . For more than two groups , data were analyzed by Kruskal-Wallis test . Survival curves were analyzed by the log-rank test using the Kaplan-Meier method . GraphPad Prism 6 software was used , in which differences between groups were considered significant when p < 0 . 05 ( 5% ) .
|
Malaria still causes the death of approximately half a million people yearly despite efforts to develop vaccines . The ability of Plasmodium parasites to survive the immune effector mechanisms indicates how suitable the immune response must be to eliminate the infection . CD4 T cells have a dual role in protection against blood-stage malaria by producing IFNγ and helping B cells to secrete antibodies . Infected erythrocytes release adenosine triphosphate ( ATP ) , a damage signal that can be recognized by purinergic receptors . Among them , the P2X7 receptor senses extracellular ATP and induces CD4 T cell activation and death . Here , we evaluated the role of P2X7 receptor in the CD4 T cell response during blood-stage Plasmodium chabaudi malaria . We observed that the selective expression of P2X7 receptor in CD4 T cells was required for T helper 1 ( Th1 ) cell differentiation , contributing to IFNγ production and parasite control . In contrast , we found an increase in follicular T helper ( Tfh ) cell population , germinal center reaction and anti-parasite antibody production in the absence of the P2X7 receptor . Our findings provide mechanistic insights into malaria pathogenesis by demonstrating the importance of damage signals for the fine-tuning between Th1 and Tfh cell populations and thus for the outcome of the disease .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"t",
"helper",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"spleen",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"cell",
"differentiation",
"parasitology",
"parasitemia",
"developmental",
"biology",
"protozoans",
"quantitative",
"parasitology",
"immune",
"system",
"proteins",
"white",
"blood",
"cells",
"malarial",
"parasites",
"animal",
"cells",
"proteins",
"biochemistry",
"signal",
"transduction",
"t",
"cell",
"receptors",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"malaria",
"immune",
"receptors",
"organisms"
] |
2017
|
P2X7 receptor drives Th1 cell differentiation and controls the follicular helper T cell population to protect against Plasmodium chabaudi malaria
|
The human pathogen Vibrio cholerae is an aquatic bacterium frequently encountered in rivers , lakes , estuaries , and coastal regions . Within these environmental reservoirs , the bacterium is often found associated with zooplankton and more specifically with their chitinous exoskeleton . Upon growth on such chitinous surfaces , V . cholerae initiates a developmental program termed “natural competence for genetic transformation . ” Natural competence for transformation is a mode of horizontal gene transfer in bacteria and contributes to the maintenance and evolution of bacterial genomes . In this study , we investigated competence gene expression within this organism at the single cell level . We provide evidence that under homogeneous inducing conditions the majority of the cells express competence genes . A more heterogeneous expression pattern was observable on chitin surfaces . We hypothesize that this was the case due to the heterogeneity around the chitin surface , which might vary extensively with respect to chitin degradation products and autoinducers; these molecules contribute to competence induction based on carbon catabolite repression and quorum-sensing pathways , respectively . Therefore , we investigated the contribution of these two signaling pathways to natural competence in detail using natural transformation assays , transcriptional reporter fusions , quantitative RT–PCR , and immunological detection of protein levels using Western blot analysis . The results illustrate that all tested competence genes are dependent on the transformation regulator TfoX . Furthermore , intracellular cAMP levels play a major role in natural transformation . Finally , we demonstrate that only a minority of genes involved in natural transformation are regulated in a quorum-sensing-dependent manner and that these genes determine the fate of the surrounding DNA . We conclude with a model of the regulatory circuit of chitin-induced natural competence in V . cholerae .
The bacterium Vibrio cholerae is a facultative pathogen and the causative agent of the disease cholera . Cholera is far from being extinct and is , in fact , considered a re-emerging disease [1] . The destructive capacity of cholera is demonstrated by its current outbreak in Haiti . According to a recent health bulletin issued by the Ministère de la Santé Publique et de la Population ( MSPP ) of Haiti and the PAHO , 515'699 cholera cases have been reported from Haiti up to November 30th 2011 with more than 6'942 deaths . This epidemic highlights the fact that new modeling approaches are required to allow for the prediction of cholera outbreaks in time and space [2]–[5] . However , it also initiated discussions on the origin of the V . cholerae strain , which appears more closely related to south Asian strains than to Latin American and U . S . Gulf Coast isolates [6] . This study by Chin et al . [6] once more reminded us of the differentiation power that whole genome sequencing provides . Indeed an earlier study by Rita Colwell and collaborators compared 23 V . cholerae strains isolated over the past 98 years using whole genome sequencing [7] . These authors concluded that “V . cholerae undergoes extensive genetic recombination via lateral gene transfer” . It is therefore of major importance to understand the mechanisms underlying horizontal gene transfer ( HGT ) . Natural competence for transformation , as one of the three modes of HGT in bacteria , describes the physiological state that allows a bacterium to take up free DNA from the environment . If the internalized DNA is recombined into the chromosome , the bacterium is considered naturally transformed . V . cholerae commonly occurs in aquatic ecosystems , its true habitat , where it intimately associates with zooplankton and their chitinous exoskeleton [8]–[10] . In this context , it has been shown that chitin , the polymer used as the building block of planktonic exoskeletons , induces natural competence for transformation of V . cholerae [11] . Thus , HGT is tightly linked to the environmental niche of V . cholerae and potentially also to the niche of many other Vibrio species . In fact , three other species of the genus Vibrio , V . fischeri , V . vulnificus and V . parahaemolyticus , are naturally transformable in a chitin-dependent manner [12]–[14] . Transforming DNA can be used to repair damaged genes and , therefore , contributes to genome maintenance or to the acquisition of new alleles/genes , which lead to genetic diversity and evolution . Indeed , experimental laboratory microcosm experiments that simulate aquatic environments have succeeded in recapitulating a V . cholerae O1-to-O139 serogroup conversion by means of natural transformation [15] . This result provides a potential explanation for the devastating occurrence of the O139 serogroup variant of V . cholerae . Today this strain is almost undetectable in endemic regions even though researchers have feared its occurrence as the onset of a new and , therefore 8th , cholera pandemic [16] . However , an important lesson can be learned from the emergence of this new strain: by means of horizontal gene transfer ( HGT ) , Vibrio species may exchange genetic material and become more virulent to mankind . Chitin-induced natural competence and transformation is poorly understood in spite of its importance . Based on a few suggestive experiments on how natural competence could be regulated in V . cholerae , the authors of a previous study proposed a model that involved at least three regulatory pathways [11]: 1 ) induction by chitin , 2 ) catabolite repression , and 3 ) quorum-sensing ( QS ) . We and others followed up on this study and provided further evidence for an involvement of these three pathways [17]–[23] . However , all of these studies have only looked at a population-wide level . This could lead to a lack of information on how competence is regulated within a single cell . This is exemplified in one of the best-studied naturally competent bacterial species , Bacillus subtilis , for which it is known that “a majority of the bacteria being insusceptible and a minority being highly susceptible to transformation” [24] . David Dubnau and collaborators explained why only 10–20% of cells within a B . subtilis population enter the competence state , and demonstrated that such bistability is caused by intrinsic noise in competence gene expression [25] , [26] . Here we show , for the first time , that under homogeneous competence–inducing conditions V . cholerae displays a homogeneous expression pattern as the vast majority of cell within a population scored positive for expression of competence genes . Taking this important finding into consideration , we then moved on to establish an inducible competence system for V . cholerae , which is based on low levels of TfoX production and not on tfoX overexpression as previously done . To date all studies on natural competence and transformation in V . cholerae have only looked at single genes involved in the competence program , at single pathways , and at varying inter-experimental conditions ( e . g . , chitin surface transformation phenotypes compared to artificial competence induction with plasmids in rich medium etc . ) [17] , [20]–[23] . This inducible and chromosomally encoded competence system allowed us to look at different aspects of the regulatory network of natural competence under standardized conditions . Based on these new data , we propose a model of how the regulatory network of natural competence functions in V . cholerae . The goals of our study were to 1 ) investigate whether natural competence is induced in a whole population under natural or optimized conditions , 2 ) establish a homogeneous competence-inducing system to investigate the contribution of separate and interconnected regulatory pathways to competence induction , and 3 ) test whether different competence genes are subject to the same regulatory circuits . We achieved these aims by using transcriptional reporter fusion constructs of representative competence genes . We combined these fluorescent reporter fusions with the following detection methods: epifluorescence microscopy and flow cytometry , which allowed us to visualize the expression of fluorescent reporter genes at the level of single cells and to quantify gene expression accordingly; and fluorescent plate reading , which we used to investigate a plethora of regulatory mutants and regulated genes based on population average fluorescent value measurements .
To better understand whether natural competence of V . cholerae is a developmental program followed by ( almost ) all members of a population or rather a state , which only a subpopulation acquires , we investigated gene expression at the single cell level . Therefore , we transcriptionally fused the promoter regions of competence genes to those genes encoding fluorescent proteins ( FPs ) . Our choice of FPs was GFP-mut3* [27] and DsRed . T3[DNT] [28] , [29] as both of them have been optimized for fluorescence intensity and can be visualized within the same cell ( i . e . , the excitation/emission spectra are adequately separated ) . We initially focused on two promoter regions: the upstream region of the pilA-D operon [30] , hereafter referred to as the pilA promoter , and the region upstream of comEA . Both of these genes , pilA and comEA , are upregulated on chitin ( [11] , [31]; Blokesch and Schoolnik , unpublished ) and essential for natural transformation to occur [11] . PilA encodes a major pilin , which , as a part of a hypothetical type IV-like pilus [30] , is most likely involved in the DNA uptake process . The same involvement in DNA uptake holds true for ComEA , which shows homology to ComEA of Bacillus subtilis [32] as it also contains a helix-hairpin-helix ( HhH ) motif ( pfam12836 ) . HhH motifs have been previously described as short DNA-binding domains that bind DNA in a non-sequence-specific manner [33] . How exactly the DNA uptake , including the involvement of the type IV-like pilus and ComEA , functions is so far unknown for V . cholerae and for other naturally competent bacteria . With these reporter fusions in hand , we moved on to visualize competence gene expression . We first tested the expression in V . cholerae strains after allowing them to colonize chitin beads . Chitin beads mimic the natural environment of V . cholerae in which the bacteria are often found associated with the chitinous exoskeletons of zooplankton [10] . In contrast to other competence-inducing chitin surfaces , such as crab shell fragments or chitin flakes [11] , [19] , chitin beads are amenable to light microscopy . As shown in Figure 1 , no significant green or red fluorescence was detectable by epifluorescence microscopy for V . cholerae cells grown on chitin beads if the bacteria carried the promoter-less FP reporter plasmid ( vector control; panel I ) . In contrast , bright fluorescence signals were visible when the FP genes were driven by either of the two promoters belonging to pilA or comEA ( panel II and reciprocal with respect to the FP-fusions in panel III ) . From the obtained images , it was apparent that not all of the bacteria within the population were fluorescent and , therefore , expressing the competence genes at detectable levels ( Figure 1 ) . Based on the finding that not all bacteria appeared as fluorescent we constructed another reporter as positive control , which consisted of the promoter preceeding the housekeeping gene gyrA ( encoding gyrase ) transcriptionally fused to gfp . This fusion was cloned onto the same plasmid as the PcomEA-dsRed fusion ( see Material and Methods ) . We tested this reporter strain in our chitin bead colonization assay ( Figure 1 , panel IV ) . In contrast to dsRed being driven by the comEA promoter the gyrA promoter led to detectable gfp expression in a significantly larger fraction of cells . The same expression pattern as for gyrA was observable for three additional transcriptional reporter fusions containing the promoter region of recA , clpX , and ftsH , respectively ( Figure S1 ) . We included these reporter strains in our study as the expression of the housekeeping gene gyrA is most likely controlled by DNA supercoiling as it was demonstrated for E . coli [34] and/or might be cell cycle-dependent as shown for Caulobacter crescentus [35] . The rational for choosing recA , clpX , and ftsH as additional positive controls was based on the fact that other researchers have already used these genes to normalize quantitative RT-PCR expression data . Furthermore , expression of none of these genes was significantly changed in microarray expression studies using different V . cholerae mutant strains ( [11] , [36] and Blokesch and Schoolnik , unpublished ) or using different growth conditions ( e . g . comparing rabbit ileal loop grown cells versus exponential in vitro cultures of V . cholerae [37] ) . We considered three different reasons for the finding that competence genes are only expressed at detectable levels in a fraction of the population: 1 ) competence gene expression in V . cholerae is a bistable phenomenon , which is similar to B . subtilis [25]; 2 ) the environment around the chitinous surface is heterogeneous and , thus , does not lead to competence induction in all cells; and 3 ) the fluorescence signal in cells that appear as non-induced for pilA and comEA expression is too weak to be detected with our epifluorescence microscopy settings . To follow up on these three possibilities , we aimed at differentiating between the existence of an intrinsic bistable switch versus the idea of a heterogeneous expression pattern due to heterogeneous conditions and to concomitantly judge whether the seemingly uninduced cells observed on the chitinous surfaces ( Figure 1 ) were the result of experimental limitations in our system . First , we wanted to test if we would observe a bistable competence gene expression pattern under homogeneous growth conditions . Therefore , we changed the chitin substrate from chitin beads ( an insoluble GlcNAc polymer ) to soluble hexa-N-acetylchitohexaose ( from here on referred to as GlcNAc6 ) . This chitin oligomer has been used before to induce natural competence in V . cholerae [11] , [23] . As a control , we grew the same V . cholerae reporter strains under identical minimal medium conditions ( defined artificial seawater medium , DASW; [11] ) but changed the main carbon source from GlcNAc6 to the non-competence inducing chitin monomer N-acetylglucosamine ( GlcNAc ) . All strains were grown to the same optical density before being either visualized by epifluorescence microscopy ( Figure S2 ) or quantified with respect to their fluorescence intensity using flow cytometry ( Figure 2 ) . As shown in Figure S2A and S2B , we did not detect significant fluorescent signals for the promoterless reporter control using microscopy , which is in accordance with the low levels of fluorescent signals measured by flow cytometry ( Figure 2A ) . Thus , the fluorescence intensity of these bacteria ( panel A , flow cytometry graphs ) was considered background . For the strains grown with GlcNAc , FP expression driven by the comEA promoter was not detectable using microscopy ( Figure S2C , S2E , S2G including the respective image analysis ) and quantified as extremely low fluorescent signals using flow cytometry ( Figure 2B–2D upper row ) . However , weak pilA promoter-driven gfp expression in the presence of GlcNAc was observed after an extended exposure time ( Figure S2C ) . We confirmed this basal pilA promoter-driven gfp expression in the presence of GlcNAc by flow cytometry ( mean FU = 7 . 2×102; Figure 2B , upper row ) . Swapping the FP reporter gene behind the pilA promoter from gfp to dsRed resulted in undetectable red fluorescence using our epifluorescence microscopy settings ( Figure S2E ) ; however , an increased pilA promoter-driven expression of dsRed compared to the promoterless reporter plasmid control ( Figure 2A ) was detectable by flow cytometry ( Figure 2C , upper row ) , confirming the low level of pilA expression under non-competence inducing conditions . In cells grown under competence-inducing conditions ( e . g . in the presence of GlcNAc6 ) , the reporter strains displayed strong pilA and comEA promoter-driven fluorescent signals ( Figure S2D , S2F , S2H ) . This was the case for the majority of the cells and only a minority appeared as non-fluorescent under these conditions . There was a distribution of fluorescence intensities as depicted in the flow cytometry graphs ( Figure 2B–2D , lower row ) but the distribution was not bimodal . To further investigate whether the non-fluorescent-appearing cells in the microscopy images ( Figure S2 ) were meaningful with respect to competence expression , we again investigated the behavior of the housekeeping gene reporter strains under these homogenous competence-inducing conditions ( for gyrA see Figure 2D and Figure S2G and S2H; or for recA , clpX , and ftsH see Figure S3 ) . The same expression pattern as for the competence genes was observed for these strains with a minority of cells not displaying any detectable fluorescence using our epifluorescence microscopy and image display settings . Thus , and also based on the image analysis ( Figure S2 ) and on the flow cytometry measurements ( Figure 2 ) , this minority of non-fluorescent-appearing bacteria probably corresponds to cells , which fluoresce at lower levels ( mostly with a good correlation between both FPs; Figure S2 ) . An issue that we had to consider was the fact that our FP reporter constructs were plasmid-encoded . Indeed , plasmid copy numbers can change in V . cholerae according to growth rate [38] . However , in our experiments we mainly looked at different strains but under similar growth conditions ( and the biological replicates were highly reproducible; Figure S4 ) . Furthermore , a recent study by Silander et al . provided evidence that plasmid-based systems are useful to study gene expression in bacteria; they observed that both the mean and the variation of expression correlated well between both settings [39] . Based on the data described above , we concluded that bistability of competence gene expression is unlikely in a population of V . cholerae cells under homogeneous competence-inducing conditions . Therefore , it seemed feasible to further investigate the regulatory circuit at a population-wide level given that homogeneous , competence-inducing growth conditions were provided . Unfortunately , GlcNAc6 , a competence-inducer , has recently been discontinued by the Seikagaku Corporation , and multiple and large-scale experiments using this commercially available compound are extremely costly . Furthermore , shorter GlcNAc oligomers such as GlcNAc2 often result in large variations with respect to transformation frequencies ( M . Blokesch , unpublished ) , which is most likely due to GlcNAc monomer impurities within the preparation that exert catabolite repression on natural transformation [23] . Therefore , we thought of establishing a chitin-independent , competence-inducing system . One possibility was to artificially overexpress the major regulator of transformation , TfoX , from a plasmid as previously performed [11] , [17] , [22] , [23] . However , we observed major disadvantages using this method . First , the morphology of some of the cells changed towards a filamentous form after competence induction due to the plasmid-maintaining antibiotic ampicillin ( Blokesch , unpublished ) . Second , TfoX overexpression from a multi-copy plasmid resulted in the induction of heat shock proteins and chaperones [11] , which might be indicative of stress conditions . Indeed , the toxicity of TfoX overexpression has been previously described for Escherichia coli and Haemophilus influenzae [40] , [41] . Third , we wanted to avoid working with V . cholerae cells containing two different plasmids within the same cell ( tfoX and FP reporter fusions-carrying ) . Therefore , we constructed a chromosomally encoded competence induction system , which is based on inducible low-level TfoX production . The system was composed of tfoX under the control of an arabinose-inducible promoter ( PBAD ) and the gene encoding AraC , which act as a repressor or initiator of gene expression in the absence or presence of L-arabinose , respectively [42] . Both of these elements were cloned into a mini-Tn7 transposon [43] , which integrates into the large chromosome of V . cholerae ( later referred to as TntfoX ) . We transferred this transposon by triparental mating into the V . cholerae wild type strain A1552 and tested the respective strain for natural transformability in LB medium ( Figure 3 ) . By adding a low amount of L-arabinose ( 0 . 02% ) , we obtained transformation frequencies that were two orders of magnitude higher than what has been described for overproduced tfoX in V . cholerae [11] ( Figure 3A ) . Furthermore , the transformation frequency ( 2 . 1×10−4 ) was in the same range as the frequencies we usually obtain using optimized chitin-inducing conditions ( 3 . 1×10−4; [19] ) . We were unable to detect any naturally transformed colony-forming units in the inducer-free control culture ( shown for the wild type in Figure 3A but likewise tested in all other strains ) . We also examined the abundance of TfoX at the protein level . To do so we grew the V . cholerae strain A1552-TntfoX in LB medium in the absence or presence of different L-arabinose concentrations followed by western blot analysis of cellular proteins using antibodies against TfoX ( Figure S5 ) . In parallel we grew a strain containing inducible tfoX on a plasmid similar to the tfoX-overexpression system described earlier [11] . As shown in Figure S5 we observed a major difference in TfoX protein levels comparing the previous [11] and current experimental setup as indicated by the two arrows . This reassured us that this system was not heavily overproducing TfoX and was therefore adequate for further analysis to establish the genetic interactions downstream of TfoX . We first wanted to visualize competence gene expression in this chitin-independent system . We transferred the respective FP reporter fusion constructs into a wild type V . cholerae strain carrying the chromosomally encoded tfoX construct ( A1552-TntfoX ) and visualized FP gene expression by epifluorescence microscopy ( Figure 3B ) . As can be appreciated from the images in Figure 3B ( middle and lower part ) , the pilA- and comEA promoter-driven expression pattern of the FP reporters under such chitin-independent , competence-inducing conditions mirrored what we observed under the GlcNAc6-mediated induction of competence ( Figure S2 ) . As described above for the chitin-dependent experiment , only a minority of cells did not display any detectable fluorescence using this microscopy technique , which was also the case for gyrA promoter-driven FP reporter expression ( data not shown ) . The fluorescent signal was below the detection limit in cells grown in the absence of inducer ( Figure 3B , -ara ) or in cells harboring the promoter-less plasmid as a control ( Figure 3B , upper two rows ) . The fluorescent signal was quantified using a 96-well plate reader ( Figure 3C ) . A statistically significant increase in fluorescence intensity was observed upon induction of competence for all promoter-driven FP reporter fusion constructs ( Figure 3C , middle and right columns ) . No significant difference in fluorescence signals between competence-uninduced and competence-induced conditions was observed for the promoter-less FP reporter control ( Figure 3C ) . In agreement with the chitin data described above , we detected a statistically significant increase in pilA-driven gfp expression compared to the promoter-less construct even in the absence of inducer . Therefore , we conclude that this basal expression of pilA is TfoX-independent . We then moved on to investigate the regulatory network of natural competence in V . cholerae in further detail . The first assay aimed at testing whether carbon catabolite repression ( CCR ) plays a role in this chitin-independent setup . Carbon catabolite repression occurs if preferred phosphoenolpyruvate∶carbohydrate phosphotransferase system ( PTS ) -transported sugars are abundant . However , in their absence the PTS systems are unsaturated and indirectly lead to the activation of the enzyme adenylate cyclase , which subsequently synthesizes cAMP within the cell ( for review on CCR see [44] ) . We were mostly interested in V . cholerae strains that are impaired in this synthesis or in the degradation of cAMP . The concentration of cAMP within cells is accomplished by interplay between adenylate cyclase ( CyaA ) and cAMP-degrading phosphodiesterases ( CpdA ) . A recent study by Kim et al . demonstrated the importance of CpdA in balancing the intracellular cAMP level in Vibrio vulnificus [45] . As the cpdA gene of V . cholerae is located at the same chromosomal locus ( as analyzed using SynTView , a synteny viewer developed by the Genomic and Genetic Department of Institute Pasteur , Paris ) and cpdA/CpdA display 64%/68% identity ( 76%/82% similarity ) at the DNA and protein levels , respectively , the functionality of the protein is most likely identical in both organisms . To disrupt the equilibrium between cAMP production and degradation , the cpdA gene in V . cholerae was deleted ( Table 1 ) . Although the production of cAMP does not change in this mutant , cAMP degradation should be impaired , resulting in the accumulation of cAMP within the cell . We tested this strain in a chitin surface colonization assay [23] and observed a hyper-colonization phenotype consistent with increased intracellular cAMP levels ( data not shown ) . More importantly , we transferred the TntfoX transposon into this V . cholerae strain as well as a strain lacking adenylate cyclase ( ΔcyaA ) and tested both strains with respect to natural transformability ( Table 2 ) and pilA/comEA promoter-driven FP gene expression ( Figure S6 ) . The results confirmed part of what we had previously demonstrated on chitin surfaces [23] , namely that adenylate cyclase is essential for natural transformation even under rich culture medium conditions . However , in this study we extended this knowledge by showing that a statistically significant increase in natural transformability occurred in the newly constructed cpdA mutant compared to the wild type parental strain ( Table 2 ) . With respect to competence gene induction , pilA and comEA promoter-driven FP gene expression was abolished in the absence of cAMP ( Figure S6 ) . This was in contrast to the fluorescent signal measured for the cpdA mutant , that is , both the pilA and the comEA promoter efficiently drove FP gene expression in this genetic background upon competence induction ( Figure S6 ) . These data highlight the necessity of cAMP for competence gene expression even when competence induction is uncoupled from chitin surface colonization and chitin degradation ( e . g . , from metabolism of carbon sources ) . The next question we wanted to address was with respect to QS and the involved autoinducer molecules . We recently showed that the species-specific cholera autoinducer 1 ( CAI-1; [46] , [47] ) plays a major role in natural competence for transformation and suggested that CAI-1 could be considered a competence pheromone [22] . We showed that the absence of the non-species-specific autoinducer 2 ( AI-2; [48] ) had no statistically significant effect on natural transformation on chitin surfaces , whereas strains devoid of CAI-1 synthesis were rarely transformable , and , even then , only at very low transformation frequencies [22] . However , as described above , chitin surfaces appear to be a rather heterogeneous environment , and cells might not all encounter the same autoinducer concentration in time and space , making chitin surface experiments difficult to conclusively judge the involvement of QS in the regulatory circuit of V . cholerae . Therefore , we investigated the role of QS in natural competence and transformation using our homogeneous competence-inducing system ( Figure 4 ) . We constructed TntfoX-containing V . cholerae deletion strains , which were devoid of either or both of the autoinducer-synthesizing enzymes CqsA and LuxS , or which lacked the gene encoding the major regulator of QS , HapR . These strains were grown in LB medium in the presence or absence of the competence-inducer arabinose and scored for natural transformability or competence gene promoter-driven FP expression ( Figure 4 ) . In the absence of inducer , the transformation frequency was consistently below the level of detection in all strains ( Figure 4A ) . In the absence of the AI-2-producing enzyme LuxS , only a minor and statistically not significant decrease in transformation frequency upon competence induction was detectable when compared to the wild type parental strain ( Figure 4A ) . More importantly , the dependency on CAI-1 was even enhanced when compared to our previous study on chitin surfaces , in that a deletion in the gene cqsA completely abolished natural transformation . This was also the case in strains devoid of both autoinducer synthases ( CqsA and LuxS ) and the major regulator of QS , HapR ( Figure 4A ) . We argue that few occasionally detectable transformants in a CAI-1 negative mutant in our previous study on chitin flake surfaces [22] were the result of the heterogeneity of the chitin surface environment in which at least three components are not evenly distributed: autoinducers , transforming DNA and nuclease . However , under homogeneous conditions as tested here , a full dependency on CAI-1 is apparent . We then visualized and quantified competence gene expression in these strains using the above-described FP report fusions ( Figure 4B , 4C ) . The data for comEA promoter-driven FP expression ( Figure 4B ) mirrored the data of the transformation assay ( Figure 4A ) . Under non-competence-inducing conditions , only the background fluorescence was measurable ( in the range of the vector control shown in Figure 3 ) . There was no statistically significant difference between the fluorescence signal detected in the wild type strain and the signal derived from the luxS-deficient strain upon competence-inducing conditions . A highly significant reduction of comEA promoter-driven FP gene expression was detected in the cqsA , cqsA/luxS and hapR negative strains . We also measured for the first time pilA promoter-driven FP expression in the different QS mutants and thus in the presence or absence of the two autoinducers . We observed that the expression pattern looked completely different from the comEA data; though the fluorescent signal increased upon competence induction , the fluorescence units were in the same range for all strains tested ( Figure 4C ) . Therefore , we conclude that , in contrast to comEA , pilA is not subject to QS-dependent regulation . Taken together , we provide evidence that CAI-1 is essential for comEA expression and natural transformation under homogeneous competence-inducing conditions . These data are in slight contrast to another study where a gradual decrease in comEA expression and natural transformation from a wild type V . cholerae strain towards an AI-2- and CAI-1-deficient strain , respectively , was displayed [21] . The authors of this study concluded that not only CAI-1 but also AI-2 contributes to natural transformation . We believe that the discrepancy between studies ( the study described here , [21] and [22] ) could reflect the different V . cholerae O1 El Tor strains used in both studies ( A1552 here and in [22] , versus C6707str in [21] ) . This hypothesis is in excellent agreement with a recent finding by Fong and Yildiz [36] , who showed that the cAMP-CRP-mediated negative regulation of the biofilm regulatory gene vpsR only occurred in three out of four tested V . cholerae strains , namely strains A1552 , N16961 and MO10 . In contrast to this result , V . cholerae strain C6706 displayed no such regulation [36] , highlighting the fact that different regulatory circuits exist in these V . cholerae strains . The involvement of QS in natural transformation has previously been demonstrated by elucidating a role for HapR in the repression of a gene encoding the extracellular nuclease Dns [17] . This conclusion was mainly based on comparing a wild type V . cholerae strain to a ΔhapR mutant with respect to dns gene expression or nuclease activity . However , a direct correlation between HapR protein levels within the cell , dns repression , and comEA induction has never been demonstrated . We addressed this missing information by performing western blot analysis of non-competence-induced and competence-induced cells to detect the HapR protein within different QS mutant strains ( Figure 5A ) . Whereas the amount of HapR did not differ between tfoX-induced and tfoX-uninduced cells significant differences between the tested strains were observable ( Figure 5A ) . That is , whereas the HapR level was only slightly reduced in an AI-2-deficient strain ( ΔluxS ) , HapR was almost undetectable in a strain lacking CAI-1 ( ΔcqsA ) ( Figure 5A ) . Better detection was only possible upon overexposure of the film and such an overexposure did not reveal any HapR protein in the dual autoinducer mutant strain ( ΔcqsAΔluxS ) or the hapR negative control ( Figure S7 ) . This HapR protein pattern directly reflects the comEA-promoter driven FP expression data quantified in Figure 4B and also indicates that CAI-1 is the stronger autoinducer compared to AI-2 in V . cholerae strain A1552 ( consistent with what was shown for strain C6707 using a heterologous read-out [49] ) . We also tested the expression of another QS-dependent but competence-independent gene , hapA , by transcriptionally fusing the hapA promoter to gfp . The hapA gene encodes a hemagglutinin protease ( HA protease ) and is positively regulated by HapR [50] . When we compared the comEA expression data ( Figure 4B ) to those of hapA ( Figure 5B ) a very similar expression pattern was observable under competence inducing conditions . We hypothesize that the low amount of HapR present in the cqsA mutant ( Figure 5A and Figure S7 ) is not sufficient to activate expression of either comEA or hapA . A potential reason for this might be that HapR displays only a weak affinity for these promoters , which is consistent with a LuxR promoter affinity model described for Vibrio harveyi [51] . We also tested the impact of the HapR level on the protein amount of the nuclease Dns ( Figure 5C ) and observed an inverse correlation: Dns repression only occurred in those strains in which we detected high levels of HapR protein ( e . g . WT and ΔluxS in Figure 5A ) . This is in good agreement with the absence of any transformants in a cqsA mutant ( Figure 3A ) , as the abundance of the nuclease in this strain would avoid uptake of intact DNA . This is the first time that a direct correlation between HapR protein levels , nuclease levels and comEA expression has been shown , which is the critical link between QS and natural competence/transformation . Finally , we were curious to determine whether this chitin-independent , competence-inducing system would allow us to investigate other genes that are potentially involved in natural transformation . We initially focused on two potential promoter regions belonging to genes VC0047 and comEC . The first promoter precedes VC0047 , which is part of a four-gene operon ( VC0047-50 ) . The gene VC0048 in this cluster encodes DprA , which is essential for natural transformation of V . cholerae [22] . The function of DprA in Streptococcus pneumoniae and probably also in V . cholerae is to protect the incoming single-stranded DNA from degradation and to convey the DNA to RecA-mediated recombination [52] . As shown in Figure 6A , a tfoX-dependent expression pattern was detectable in our VC0047 promoter-driven gfp reporter strain . Another gene that we have previously shown to be essential for transformation of V . cholerae is comEC ( [22]; annotated as inner membrane transporter ) . As so far nothing was known about its regulation we sought to investigate whether comEC is regulated in a TfoX-dependent manner . Using a comEC promoter-driven gfp reporter strain we were able to measure a slight but statistically significant increase in comEC upon competence induction ( Figure 6A ) . This is the first time that comEC has been shown to belong to the chitin-/TfoX-regulon in V . cholerae and as such being co-regulated with other competence genes . We were also interested in the regulation of the pilM-Q operon as pilQ is also required for natural transformation [11] , [22] . Thus , we fused the pilM promoter region to gfp and determined FP expression under competence inducing conditions . Unfortunately , the signal intensity was too low to allow us to unambiguously judge pilM promoter-driven expression . To overcome this obstacle we established quantitative RT-PCR in our laboratory to further monitor competence gene expression using our chitin-independent system . We first compared competence-uninduced to competence-induced cells with respect to expression of comEA , pilA , pilM , VC0047 , dprA and comEC . As indicated in Figure 6B all of these genes were significantly induced upon competence-induction . This again confirmed the TfoX-dependent regulation of comEC shown in Figure 6A , which was missed in previous chitin/TfoX-dependent expression studies [11] , [31] . We suggest that this was the case , as the change in expression did not pass the significance filter in these microarray expression studies . Indeed , the fold-difference for comEC expression upon competence induction was only 2 . 7 ( Figure 6B ) . Interestingly , this 2 . 7-fold increase in comEC expression is in the same range as what we observed using the comEC FP reporter construct ( 1 . 8-fold change; Figure 6A ) , though the latter system was plasmid-based . Finally , we wanted to test whether any of these other competence genes is also regulated in a QS-dependent manner . We therefore tested and compared the expression of these genes under competence-inducing conditions in a wild type and hapR negative strain , respectively ( Figure 6C ) . Apart from comEA and dns , only comEC turned out as also HapR-dependent ( Figure 6C , P = 0 . 0157 ) . This is in nice agreement with our regulatory model in which the fate of the surrounding DNA is determined by QS ( Figure 7 and conclusion below ) . In this study , we analyzed the regulatory network of chitin-induced natural competence and transformation in V . cholerae in its full complexity ( Figure 7 ) . Our results suggest that under homogeneous conditions bistability is unlikely for V . cholerae . However , the conditions might be less homogeneous in time and space around biotic surfaces ( e . g . , different concentrations of autoinducers and PTS sugars , which interfere with natural competence via QS and CCR , respectively ) . Such “environmental heterogeneity” might foster a non-synchronized response by the chitin-associated bacteria . We hypothesize that due to such a heterogeneity competence gene expression appeared absent in a subpopulation of bacteria grown on chitin beads , whereas housekeeping genes were almost uniformly expressed throughout the population ( Figure 1 and Figure S1 ) . Based on the results obtained in this study and combined with the knowledge from earlier studies by us and others [11] , [17] , [22] , [23] , we developed a model for the regulatory network of natural competence and transformation ( Figure 7 ) . The model predicts the interplay between three pathways for the initiation of competence: chitin sensing followed by TfoX activation , CCR , and QS . The first pathway , the dependency on a chitin surface or on chitin oligomers ( e . g . , GlcNAc2–6 ) for competence induction , was discovered as these compounds lead to an upregulation of potential competence genes . Within these competence genes were the so-called pil genes , which encode for a type IV pilus that is potentially involved in the DNA uptake process [22] , [31] ( Figure 7 ) . Chitin also led to an induction of the gene encoding the main regulator of natural transformation , TfoX [11] , [31] . This finding is supported by recent studies that show that chitin oligomers ( GlcNAc>2 ) lead to an increase of tfoX transcription but also to its enhanced translation [18] . The latter effect could be explained after the discovery of a chitin-induced small RNA TfoR , which activates translation of TfoX mRNA and , therefore , contributes to the induction of natural competence [20] . Attempts from our group to look at transcriptional reporter fusions between tfoX and gfp were unsuccessful , which was most likely due to the low activity of the tfoX promoter ( M . Lo Scrudato , M . Grasser , M . Blokesch , unpublished ) . The flow of information in this part of the regulatory circuit ( e . g . , chitin sensing ) is therefore as follows ( Figure 7 ) : the presence of chitin is sensed by the chitin sensor ChiS , due to chitinase-released GlcNAc di-/oligomers [31] , [53] , and the signal is then transferred via TfoR towards the production of TfoX . As this chitin-dependent pathway is well established , we excluded it in the second part of this study and designed a chitin-independent , competence-inducing system , which is based on artifical tfoX expression ( though not overproduction ) . With this chitin-independent system , we obtained transformants at comparable frequencies to our optimized chitin-induced transformation protocol [19] and at 10- to 10 , 000-fold higher frequencies than recent studies by other groups [18] , [20] , [21] . As the induction occurs under homogeneous conditions in this system , it allowed us to us to better investigate the two other pathways involved in natural competence regulation , CCR and QS . Based on experimental data showing that glucose interferes with chitin-induced transformation , Meibom et al . suggested that catabolite repression might be involved in the competence phenotype [11] . This hypothesis was recently extended as we showed that competing PTS-dependent carbon sources indeed repress natural transformation [23] . Such sugars are known to play a role in the intracellular accumulation of the secondary messenger cAMP , which , together with the cAMP receptor protein CRP , contributes to chitin surface colonization , chitin degradation and natural competence [23] . In this current study , we circumvented the problem that CCR mutants are often impaired for colonization and growth on chitin as a sole carbon source [23] by uncoupling natural competence-induction from chitin . This allowed us to better understand the dependency on cAMP for competence gene expression . Indeed we observed a change in natural transformability upon creating an imbalance in the intracellular cAMP pool , either by inhibiting cAMP production or , alternatively , by avoiding cAMP degradation . The latter effect was accomplished by deleting the gene encoding the cAMP phosphodiesterase , an enzyme that has never been studied in V . cholerae before . We also confirmed that the TfoX-induced expression of pilA and comEA requires cAMP , which is consistent with the idea proposed , but not yet unequivocally demonstrated , for Haemophilus influenzae that TfoX and cAMP-CRP act in concert to induce competence genes [40] , [41] , [54] ( Figure 7 ) . The third pathway that participates in natural competence induction is quorum-sensing ( QS ) . The involvement of QS in competence initiation was initially speculated based on different facts . First , the first sequenced strain of V . cholerae N16961 [55] was non-transformable [11] . In this strain , the indigenous hapR gene contains a frameshift mutation that renders it non-functional [56] . This deficiency could be overcome by providing a functional copy of hapR back in cis [11] . The second line of evidence for the involvement of QS in natural competence and transformation came from the fact that cells were more efficiently transformable after longer growth on chitin surfaces , which is equivalent to higher cell densities [11] . However , such a finding could also be explained by elevated intracellular cAMP levels after extended growth on chitin . The involvement of QS in natural transformation was more directly demonstrated by elucidating a role for HapR in the repression of a gene encoding the extracellular nuclease Dns [17] . The finding that HapR “acts as a negative regulator for dns transcription” was recently also confirmed by others [57] . The study by Blokesch and Schoolnik unambiguously demonstrated that the HapR-induced repression of the nuclease is the main , but not the only , contribution of QS to natural transformability , and the authors proposed that HapR also acts as a positive regulator of comEA [17] . This speculation was based on earlier microarray expression data , which showed that comEA expression upon growth on chitin was significantly reduced in the absence of HapR [11] . A QS-dependent regulation of comEA has recently been demonstrated [21] but differed from the data presented here as discussed above . In the current study , we extended this analysis by studying natural transformation and competence gene expression using the same competence-inducing conditions . Furthermore , we simultaneously monitored the expression of two competence genes , namely comEA and pilA , and compared their expression levels in different QS mutants . Based on the data provided , we conclude that comEA but not pilA is regulated by QS ( Figure 7 ) . We also showed for the first time a direct link between intracellular HapR protein concentrations , nuclease repression and comEA induction ( Figure 4 and Figure 5 ) . Finally , we tested other competence genes , namely pilM , VC0047 , dprA , and comEC , with respect to their expression levels; for these genes less or no information concerning their regulation was known before this study . We conclusively showed that all tested competence genes were dependent on induction by TfoX but only a part of them was also co-regulated in a QS-dependent manner ( Figure 6 ) . In fact only three genes showed a significantly different expression under competence inducing conditions in a wild type strain compared to a hapR mutant , namely dns , comEA and comEC ( Figure 7 ) . The encoded proteins are all directly involved in the fate of the surrounding DNA: whereas Dns degrades free DNA at low cell density , ComEA and ComEC are required for the DNA uptake process once high cell density is reached on the chitin surface ( Figure 7 ) . Further studies will follow to provide a better insight into the DNA uptake process itself .
Bacterial strains and plasmids used in this study are listed in Table 1 . E . coli strains DH5α [58] and One Shot PIR1 or PIR2 ( Invitrogen ) were used as hosts for cloning purposes . E . coli strain S17-1λpir [59] served as mating donor for plasmid transfers between E . coli and V . cholerae . Overnight cultures were grown in LB medium under aerobic conditions . Defined artificial seawater medium ( DASW ) [11] supplemented with vitamins ( MEM , Gibco ) and 0 . 1% casamino acids ( Becton , Dickson and Company ) was used for static growth of V . cholerae on chitin beads ( New England Biolabs ) or for growth under shaking conditions with N-acetylglucosamine ( GlcNAc ) or hexa-N-acetylchitohexaose ( GlcNAc ) 6 ( obtained from Seikagaku Corporation via Northstar BioProducts and LuBioScience , Lucerne , Switzerland ) as sole carbon source . Thiosulfate Citrate Bile Salts Sucrose ( TCBS ) agar plates were prepared following the manufacturer's instructions ( Fluka ) and were used to counter-select E . coli strains after triparental mating with V . cholerae . Experiments for artificial tfoX expression were performed in LB medium with or without the addition of 0 . 02% arabinose . LB medium and LB agar plates were supplemented with antibiotics wherever required . The final concentrations of antibiotics were 75 µg/ml for kanamycin , 50 or 100 µg/ml for ampicillin and 50 µg/ml for gentamicin . V . cholerae cells were always grown at 30°C . The V . cholerae cpdA deletion strain was constructed using plasmid pGP704-28-SacB-ΔcpdA ( Table 1 ) and the gene disruption method described previously [31] . The oligonucleotides used for the construction of the deletion plasmid are indicated in Table S1 . V . cholerae strains carrying artificially inducible tfoX on the chromosome ( e . g . , in cis ) were created by triparental mating between the respective V . cholerae strain ( Table 1 ) , E . coli strain S17λpir/pUX-BF13 ( providing the transposase function; [60] ) , or E . coli strain S17λpir/pGP704-mTn7-araC-tfoX . The latter plasmid consists of the suicide vector pGP704 as the backbone and the mini-Tn7 transposon [43] containing the gene cluster araC-PBAD-tfoX as cargo . This gene cluster was amplified by PCR from the plasmid pBAD-tfoX-stop [22] ( primers indicated in Table S1 ) . The FP reporter constructs are all based on plasmid pBR322 [61] . Initially , this plasmid was modified by ( partial ) deletion of the tetracycline resistance cassette as well as the constitutive promoter PTet , resulting in plasmid pBR-Tet_MCSI ( Table 1 ) . A kanamycin resistance cassette ( aph ) as well as promoter-less versions of the genes gfp and dsRed ( DsRed . T3[DNT] ) , both pointing in the opposite direction , were inserted into pBR-Tet_MCSI to yield plasmid pBR-GFP_dsRed_Kan . Details of all plasmids are included in Table 1 . The primers used for construction of the plasmids are listed in Table S1 ( synthesized by Microsynth , Switzerland ) . PCR mixtures , PCR programs , restriction enzyme digestions , primer phosphorylation , and ligations followed standard protocols recommended by the manufacturers of the enzymes ( Roche , Switzerland and New England Biolabs via Bioconcept , Switzerland ) . V . cholerae strains were grown aerobically in LB medium until an OD600 of ∼0 . 3 . Cells were harvested , washed in DASW medium and mixed with an equal volume of prewashed chitin beads ( New England Biolabs ) and three volumes of DASW ( final volume 1 ml ) . The mixture was supplemented with vitamins , 0 . 1% casamino acids , and kanamycin ( for plasmid maintenance ) . The bacteria were grown as standing cultures in 12-well plates . Bacteria were visualized by epifluorescence microscopy after 24 and 48 hours of growth , respectively . Specifications of the epifluorescence microscope were as follows: Zeiss Axio Imager M2 microscope; Zeiss High Resolution Microscopy Camera AxioCam MRm; Illuminator HXP 120 as fluorescence light source ( metal halide ) ; objective used in this study: Plan-Apochromat 100×/1 . 40 Oil Ph3 M27 ( WD = 0 . 17 mm ) . Filters relevant to this study were Zeiss Filter set 63 HE mRFP shift free; EX BP 572/25 , BS FT 590 , EM BP 629/62 and Zeiss filter set 38 Endow GFP shift free; EX BP 470/40 , BS FT 495 , EM BP 525/50 . Image acquisition was done using the Zeiss AxioVision software . Images were rotated , cropped and uniformly enhanced with respect to contrast and brightness using Zeiss AxioVision and Adobe Photoshop CS3 . Microscopy image analysis was performed using the Matlab-based MicrobeTracker Suite [62] and according to the instructions given by the inventors ( http://microbetracker . org/ ) . Fluorescence intensities were normalized with respect to the area of the cell and the exposure time . Chitin-dependent but surface-independent growth of V . cholerae was performed as previously described [23] using 2 mM of hexa-N-acetylchitohexaose GlcNAc6 ( obtained from Seikagaku Corporation via Northstar BioProducts and LuBioScience , Lucerne , Switzerland ) as sole carbon source . The same strains were grown in parallel with N-acetylglucosamine ( GlcNAc; control ) . Bacteria were harvested at an OD600 of 0 . 8 . Bacteria were either immediately visualized by epifluorescence microscopy or fixed in 2% paraformaldehyde for 30 min . Fixed samples were washed , diluted in PBS ( 1∶5 ) , and analyzed by flow cytometry using a BD LSR II Flow cytometer . BD FACSDiva software was used for data acquisition . GFP signals were excited with a blue laser ( 488 nm ) and detected with a 525/50 filter . DsRed . T3[DNT] was excited with a green laser ( 561 nm ) and detected with a 585/15 filter . For each sample , 100 , 000 events were counted in total . Biologically independent experimental replicates ( three for Figure 1 and Figure 2; two for Figure S1 and Figure S3 ) were performed within two weeks . One representative experiment is depicted in Figure 2 and the averages of the mean fluorescence intensities from three different biological replicates are shown in Figure S4 . Strains used for chitin-independent competence-induction all carried inducible tfoX ( araC-PBAD-tfoX ) on a mini-Tn7 transposon [43] within the chromosome . Induction was accomplished by growth in LB supplemented with 0 . 02% arabinose . For transformation assays , cells were grown until an OD600∼1 . 0 . At that point , aliquots of 0 . 5 ml cultures were transferred to 1 . 5-ml tubes and supplemented with 2 µg/ml transforming DNA ( gDNA of strain A1552-LacZ-Kan; [19] ) . Tubes were shaken horizontally for 5 h . Transformed cells and total colony forming units ( CFU ) were enumerated by a variation of a previously described method [63] . Briefly , the cultures underwent a serial dilution , and 5 µl of each dilution step was spotted in duplicate or triplicate on plain LB or LB containing 75 µg/ml kanamycin plates , respectively . Transformation frequencies were calculated as number of transformants divided by total number of CFUs . Each experiment was repeated at least three independent times , and the averages of all experiments are given in the figures ( ± standard deviations ) . Statistical analyses of transformation frequencies were performed on log-transformed data [64] using a two-tailed Student's t test . Strains were grown in LB medium with or without 0 . 02% arabinose to investigate transcriptional FP reporter fusions . After 24 h , the cells were either visualized by epifluorescence microscopy or measured for relative fluorescence using a Tecan Infinite M200 plate reader . Parameters for detection of GFP were: excitation ( Ex ) at 485 nm ( 9 nm bandwidth ) and emission ( Em ) at 515 nm ( 20 nm bandwidth ) . DsRed . T3[DNT] was detected using Ex 560 ( 9 ) /Em 587 ( 20 ) nm , as previously described for DsRed . T3 [28] . The samples were also measured with respect to their OD600 , and results are given as relative fluorescence units ( RFU ) divided by OD600 values . The averages of three biological replicates are shown . Error bars indicate standard deviations . Statistically significant differences were analyzed using two-tailed Student's t tests . V . cholerae strains were grown in 10 ml LB in the absence or presence of 0 . 02% arabinose until they reached an optical density of ∼1 . 7 . At that time 5 ml of each culture was harvested and lysed in 1 ml Tri Reagent ( Sigma ) . The samples were stored at −80°C . RNA isolation , DNase treatment , and reverse transcription using 1 µg of total RNA as template was done as previously described [23] . The obtained cDNA was diluted 40-fold and served as template in the qPCR . The primers used for the qPCR are indicated in Table S1 . The qPCR mix was based on the Fast Start Essential DNA Green Master Mix ( Roche , Switzerland ) , a ready-to use hot start reaction mix optimized for qPCR using the Light Cycler Nano system from Roche . The qPCR mix further contained 0 . 5 µM of each primer . The qPCR run using the Light Cycler Nano was performed according to these parameters: a denaturation step at 95°C for 10 min followed by 40 cycles of 95°C for 10 s , 60°C for 20 s , 72°C for 20 s . Each run was finished with a melting-curve ranging from 50°C to 95°C to validate specific amplification , which was also initially confirmed for each primer pair by standard PCR and visualization of PCR fragments in agarose gels . For each sample a reverse transcriptase-negative control was also performed while doing the reverse transcription and the respective samples were analyzed with at least three independent primer pairs to exclude residual DNA contaminations . A standard curve was prepared for each primer pair using purified genomic DNA of V . cholerae A1552 diluted in PCR grade water ( from 1000 to 0 . 1 pg gDNA template ) . A negative control lacking any template was also tested for each primer pair . The expression values were normalized against expression of the housekeeping gene gyrA as previously described [65] . However , as discussed above the expression of gyrA might be dependent on DNA supercoiling and the cell cycle as shown for other bacteria [34] , [35] . We therefore compared the relative expression of the four housekeeping genes gyrA , recA , clpX , ftsH as well as comEA in two different V . cholerae strains ( WT and ΔhapR ) . The expression patterns were extremely similar no matter whether we normalized the expression data against gyrA expression ( Figure S8A ) or against recA expression ( Figure S8B ) as internal controls . The results were analyzed using the Light Cycler Nano software . For the preparation of cell lysates , bacteria were grown aerobically in LB medium in the absence or presence of 0 . 02% arabinose . Cells were harvested after reaching an OD600 of ∼1 . 5 , resuspended in SDS-loading buffer ( 2×Laemmli buffer without β-mercaptoethanol and bromophenol blue ) , and boiled at 98°C for 15 min . Total protein concentration was quantified using the Pierce BCA Protein Assay kit ( Thermo Scientific ) before the addition of β-mercaptoethanol and bromophenol blue ( 7 . 5% and 0 . 01% final concentrations , respectively ) . Antibodies raised against the peptides derived from the proteins HapR , Dns , and TfoX were produced by Biomatik ( Canada ) . The polyclonal antibody production service included suggestions for the design of two peptides per protein , peptide synthesis , conjugation of the peptides to the carrier proteins ( keyhole limpet hemocyanin ) , and immunization of two rabbits per peptide mix . Polyclonal antibodies were affinity purified against the antigen and checked by ELISA . Each antibody was first validated for potential cross-reactions at the same size as the target protein using western blot analysis of the respective know-out strains . Separation of proteins under denaturing conditions was conducted by SDS-PAGE using 15% acrylamide gels [66] , [67] . The amount of total protein loaded per lane was 6 µg , 12 µg , and 50 µg for HapR , Dns , and TfoX detection , respectively . For western blot analysis , the proteins were transferred onto PVDF western blotting membranes ( Roche ) , stained with amido black to verify transfer efficiency , incubated in blocking buffer , and reacted with primary antibodies directed against HapR ( 1∶5000 ) , Dns ( 1∶1000 ) , or TfoX ( 1∶2000 ) . Detection of the primary antibody was performed using a secondary goat anti-rabbit IgG antibody conjugated to peroxidase ( Sigma A9169; used at a 1∶20 , 000 dilution ) . Signals were revealed using Lumi-LightPLUS Western Blotting substrate ( Roche , Switzerland ) and were recorded by exposure to chemiluminescence-detecting films ( Amersham Hyperfilm ECL , GE Healthcare ) .
|
The human pathogen Vibrio cholerae is an aquatic bacterium often encountered in rivers , estuaries , and coastal regions . Within this environmental niche , the bacterium often associates with the chitinous exoskeleton of zooplankton . Upon colonization of these chitinous surfaces , V . cholerae switches on a developmental program known as natural competence for genetic transformation . Natural competence for transformation is a mode of horizontal gene transfer that allows bacteria to acquire new genes derived from free DNA , which is released by other members within the same habitat . The evolutionary consequences could be that the bacterial recipient becomes better adapted to its environmental niche or , in a worst-case scenario , more pathogenic for man . The results of this study show that , under optimal conditions , the majority of cells within a V . cholerae population express competence genes . However , in an aquatic environment , a combination of different ecological factors might lead to heterogeneity in the competence phenotype . Therefore , we investigated the role of extracellular and intracellular signaling molecules with respect to competence induction . This report illustrates that at least three interconnected signaling cascades are required for competence induction , which are based on bacterial metabolism and group behavior .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"gene",
"networks",
"gene",
"regulation",
"microbiology",
"microbial",
"evolution",
"molecular",
"genetics",
"microbial",
"growth",
"and",
"development",
"bacterial",
"pathogens",
"biology",
"microbial",
"ecology",
"bacterial",
"physiology",
"gram",
"negative",
"genetics",
"bacterial",
"evolution",
"genetics",
"and",
"genomics"
] |
2012
|
The Regulatory Network of Natural Competence and Transformation of Vibrio cholerae
|
Epithelial homeostasis in the posterior midgut of Drosophila is maintained by multipotent intestinal stem cells ( ISCs ) . ISCs self-renew and produce enteroblasts ( EBs ) that differentiate into either enterocytes ( ECs ) or enteroendocrine cells ( EEs ) in response to differential Notch ( N ) activation . Various environmental and growth signals dynamically regulate ISC activity , but their integration with differentiation cues in the ISC lineage remains unclear . Here we identify Notch-mediated repression of Tuberous Sclerosis Complex 2 ( TSC2 ) in EBs as a required step in the commitment of EBs into the EC fate . The TSC1/2 complex inhibits TOR signaling , acting as a tumor suppressor in vertebrates and regulating cell growth . We find that TSC2 is expressed highly in ISCs , where it maintains stem cell identity , and that N-mediated repression of TSC2 in EBs is required and sufficient to promote EC differentiation . Regulation of TSC/TOR activity by N signaling thus emerges as critical for maintenance and differentiation in somatic stem cell lineages .
Regenerative processes in somatic tissues require coordinated regulation of stem cell proliferation and daughter cell differentiation to ensure long-term tissue homeostasis [1]–[3] . The Drosophila posterior midgut epithelium has emerged as an excellent model system to study this regulation [4]–[7] . It is maintained by Intestinal stem cells ( ISCs ) that divide to self-renew and produce enteroblasts ( EB ) , which undergo differentiation to become either enterocytes ( ECs ) or enteroendocrine cells ( EEs ) [8]–[10] . Differentiation in the ISC lineage is controlled by Delta/Notch ( Dl/N ) signaling . ISCs express Dl and activate N in EBs , thus promoting differentiation into either EEs or ECs . The cell fate decision between ECs and EEs seems to be regulated by the intensity of the Dl signal , i . e . high levels of N activity in EBs result in EC differentiation , while moderate activation of N promotes EE differentiation [10] , [11] . Dl-mediated N activation in EBs increases the activity of the Suppressor of Hairless ( Su ( H ) ) transcription factor , presumably by replacing the Hairless transcriptional repressor from Enhancer of Split ( E ( spl ) ) complex promoters with the Notch intracellular domain ( NICD ) [12] . How this pathway coordinates cell specification with cell growth and proliferation in the ISC lineage remains unclear . ISC proliferation is regulated by growth factor and stress signaling pathways [13]–[32] . These pro-mitotic signals include the Insulin/IGF signaling pathway ( IIS ) , which is sufficient and required for ISC proliferation [18] , [23] , [33] , [34] . Activation of the Insulin Receptor ( InR ) in flies initiates an evolutionarily conserved signaling cascade composed of insulin receptor substrate ( IRS , Chico ) , PI3Kinase ( DP110 ) and Akt , inducing cell proliferation and/or growth and endoreplication [35]–[39] . Interestingly , IIS induces ISC proliferation through both cell-autonomous mechanisms involving the Akt-regulated transcription factor Foxo , as well as through a non-autonomous process in which IIS – induced EB differentiation is critical to allow further ISC divisions [7] , [34] . In EBs , InR is sufficient and required for differentiation into ECs [34] . In most Drosophila tissues , cell growth is regulated downstream of Akt by the evolutionarily conserved TSC/Rheb/TOR pathway [37] , [40] , [41] . As supported by genetic and biochemical studies , this pathway can be activated in response to Akt-mediated phosphorylation of Tuberous Sclerosis Complex 2 ( TSC2; encoded by the gene gigas in Drosophila ) and subsequent inhibition of the TSC1/2 complex [37] , [40] . TSC1 promotes the stability of TSC2 , which is a GTPase activating protein for the small GTPase Rheb , inhibiting Rheb-mediated TOR Kinase activation . TOR , in turn , phosphorylates translational regulators , including ribosomal protein S6 Kinase ( S6K ) and eIF4E Binding Protein ( 4EBP ) , resulting in a net increase of protein production in cells . In addition to Akt-mediated phosphorylation , other signals are likely to play an important role in regulating TSC1/2 activity in vivo [42] , [43] . Accordingly , various other regulatory events , including phosphorylation in response to several growth factors and morphogens , ubiquitination , and degradation , have been reported to influence TSC1/2 activity [37] , [40] , [44] . The TSC1/2 complex thus represents a critical node in signaling networks that arbitrate between cell proliferation and growth in response to increased insulin signaling . Supporting this view , mutations in TSC1/2 result in Tuberous Sclerosis Complex , a rare autosomal dominant disease that is characterized by widespread benign tumor formation [45] . Recent studies suggest that TSC/TOR signaling has an important regulatory role in both vertebrate and invertebrate stem cell lineages . In human embryonic stem cells , activation of S6K by mTOR has been reported to induce differentiation [46] , while a recent study in the mouse has identified an interesting non-autonomous function for mTOR activity in ISC support cells , the Paneth cells . Under conditions of dietary restriction , TOR signaling activity is reduced in Paneth cells , resulting in secretion of factors that promote stem cell maintenance and proliferation [47] . In the Drosophila germline , TSC/TOR signaling regulates proliferation and maintenance of germline stem cells ( GSCs ) [48]–[50] . GSCs mutant for TSC1/2 undergo differentiation , through a so far unknown mechanism [49] , [50] , while GSCs mutant for the TOR kinase exhibit proliferation defects [49] . TSC/TOR signaling is thus likely to mediate , at least partially , the effects of the dietary status of the organism on GSC proliferation and maintenance [48]–[50][4] , [51] . In the Drosophila intestine , TSC/TOR signaling may have a similar function , as ISCs are also regulated according to nutrient availability [33] , [52] . Indeed , a recent report shows that loss of TSC in ISCs causes excessive ISC growth and impairs ISC proliferation [53] . Using the ISC and EB driver esgGal4 , it was shown that TSC2-RNAi expressing ISCs become large , express less cell cycle markers , have reduced DNA replication , and that these phenotypes are Rapamycin-sensitive . These cells further fail to respond to tissue damage by initiating cell divisions , and exhibit increased DNA content , indicating that they are becoming polyploidy [53] . While these characteristics indicated differentiation of TSC deficient cells , it was shown that TSC2-RNAi expressing cells do not express the EC marker Pdm1 , and do not form ECs with brushed borders , suggesting that they may have initiated , but not completed , differentiation . TSC-mediated inhibition of TOR signaling thus seems to be critical to maintain ISC activity and function . It remained unclear , however , what physiological role , if any , TOR activation may have in ISCs or their daughter cells , and how Tor signaling may interact with others pathways regulating ISC commitment and differentiation . Here , to address these questions , we characterize the function and regulation of TSC/TOR signaling in the ISC lineage in more detail . We find that TSC2 is highly expressed in ISCs , but specifically down regulated in EBs . While , consistent with the previous report [53] , high TSC2 expression is required for ISC function , we also find that the down-regulation of TSC2 in EBs , and the resulting TOR activation , are critical for EC differentiation . Our results further suggest that TSC activity promotes lineage commitment of EBs into the EE fate . To characterize the regulation of TSC/TOR signaling in EBs further , we assessed its interaction with the Notch ( N ) signaling pathway . We find that N-induced Su ( H ) activity represses TSC2 expression in EBs . Strikingly , repression of TSC1/2 function is sufficient to commit cells into the EC fate independently of N activity , indicating that TSC2 repression is a central step in N-induced EC differentiation . We also find that food conditions significantly impact the proliferative capacity of TSC-deficient ISCs . We show that TSC mutant ISCs are capable of generating normal clones of daughter cells on a low calorie ( low yeast ) diet , but that these lineages decline over time . Rearing flies on high-yeast food , however , causes growth and proliferation phenotypes similar to the ones observed in [53] , accelerating the decline of TSC mutant clones . TSC activity in ISCs is thus specifically required to maintain ISC function under high nutrient conditions .
While InR/Akt signaling can activate TOR signaling in many Drosophila tissues [35]–[37] , [40] , [41] , previous reports have suggested an opposing role of InR and TOR signaling in the control of ISC proliferation: While Insulin/IGF signaling ( IIS ) is required for ISC proliferation , and activation of IIS ( by InR over-expression ) induces increased proliferation [18] , [23] , [33] , [34] , activation of TOR signaling ( by loss of TSC function ) was found to impair proliferative capacity [53] . Using RNAi-based knockdown of TSC2 in ISCs and EBs , Amcheslavsky et al found that loss of TSC2 increased the size of ISCs . Based on cell cycle markers and EdU incorporation experiments , it was concluded that these cells are not mitotically active . Furthermore , the proliferative capacity of TSC2 homozygous mutant ISCs was assessed using lineage tracing by somatic recombination . However , a mitotic recombination with a repressible cell marker ( MARCM , [54] ) approach was used in which GFP was expressed under the control of esg::Gal4 , which labels only ISCs and EBs ( see Materials and Methods in [53] ) . Clones with more than two cells ( including ECs and EEs ) that may be formed by TSC mutant ISCs ( see below ) can not be observed with this approach ( see for example Fig . 2A in [53] ) , and a full lineage analysis of TSC deficient ISCs was thus not possible . While the results reported in [53] thus clearly identified a critical role for TSC2 in maintaining small , diploid ISCs , it remained unclear whether activation of TOR signaling in ISCs would fully impair their proliferative activity and prevent generation of ISC daughter cells . Interestingly , the reported results indicated that TSC2 function is required in ISCs for IIS-mediated induction of proliferation , suggesting that IIS activation does not result in inactivation of TSC in the ISC lineage . It further remained unclear whether TOR signaling has to be continuously repressed by TSC2 in the ISC lineage , or whether TOR activation occurs naturally in the lineage to regulate proliferation , growth or differentiation of ISCs or their daughter cells . To characterize the relationship between InR and the TOR signaling pathway in ISC lineages in more detail , we generated ISC clones with gain- and loss-of-function conditions for multiple IIS and TOR pathway components ( Figure 1 ) . We used MARCM to generate ISC clones that over-express wild-type or dominant-negative insulin receptor ( InR; [55] ) molecules , that were mutant for the IRS homologue Chico ( carrying the loss of function allele chico1 [56] ) , or that were homozygous for the InR loss-of-function alleles InRE19 or InR353 [57] . Similarly , we generated clones with TOR pathway gain- and loss-of-function conditions by over-expressing Rheb or TSC1 and 2 , introducing the TSC1 loss of function allele Tsc1Q87X [58] , the TSC2 loss of function allele gigas192 [59] , the TOR loss of function alleles Tor2L1 , Tor2L19 and TorW1251R [60] , or the Rheb loss of function allele Rheb2D1 [61] , or expressing dsRNA against TSC2 ( TSC2RNAi ) . Importantly , we used a MARCM approach in which all daughter cells of mutant ISCs are labeled by GFP ( since GFP was expressed under the control of tub::Gal4 ) . The vast majority of GFP+ lineages ( >95% ) in the midgut were induced by mitotic recombination in response to heat shock , as very few marked cell clones could be observed in control animals ( Figure S1C ) . The number of cells in each clone at a given time point after the heat shock thus accurately reflects ISC proliferation . Consistent with previous reports [18] , [23] , [33] , [34] , gain of InR function increased the number of cells produced in a clone , while loss of InR or chico activity significantly reduced the number of cells produced by an ISC in 7 days ( Figure 1A , 1B ) . Loss of TOR pathway activity also reduced clone sizes at 7 days , and increasing TOR pathway activity ( in TSC mutants or Rheb over-expressing clones ) resulted in clones that showed no significant difference in average cell numbers at 7 day after induction . In contrast to the observations reported in [53] , ISCs with increased TOR pathway ( i . e . reduced TSC ) activity were thus capable of generating normal ISC lineages in our studies . However , the variability in clone sizes increased in TOR gain of function conditions compared to wild-type clones ( compare standard deviations in Figure 1B ) , indicating that , consistent with [53] , individual ISCs may lose the ability to generate normal numbers of daughter cells ( see below ) . As expected , we also observed significantly larger Enterocytes ( ECs , defined as the largest polyploid , Dl - negative cell in a clone ) in both IIS and TOR gain-of-function conditions at 7 days after clone induction , and significantly smaller cells in IIS/TOR loss-of-function conditions ( Figure 1C ) . This is consistent with previous findings in developmental contexts and in GSCs , showing that IIS and TOR signaling act in concert to promote endoreplication and growth [39] , [49] . Our results thus support a positive interaction between InR and TOR signaling in the ISC lineage . We tested whether TOR signaling acts downstream of InR in the regulation of proliferation and growth in this lineage by assessing the frequency of mitotic figures and the size of EC nuclei . We co-overexpressed InR with TSC1 and TSC2 , or with dsRNA against S6K ( S6KRNAi ) using the ISC/EB driver esg::Gal4 in combination with the heat-sensitive Gal4 inhibitor Gal80ts ( Figure S1A , S1B; TARGET system [8] , [9] , [62] ) . InR over-expression using this driver dramatically increases ISC proliferation rates ( as represented by the number of phospho-histone H3 ( pH3 ) positive cells [18] ) and increases cell sizes in the gut ( as represented by the size of EC nuclei; Figure S1A , S1B ) . Loss of TOR pathway activity ( over-expression of TSC1 and 2 or knockdown of S6K ) did not affect InR-mediated proliferation , but significantly prevented the increase in EC nuclear size . In these InR gain-of-function conditions , the TSC/TOR/S6K pathway is thus specifically required to promote growth and endoreplication rather than proliferation in the ISC lineage . Since these results contrasted with the observations reported in [53] , we assessed the phenotypes of TSC deficient ISCs in more detail . A timecourse analysis revealed that loss of TSC1 resulted in clones that initially grew faster than wild-type clones , but declined and became heterogeneous in size at later timepoints ( Figure 2A and Figure S2A; see large standard deviations in TSC1 mutant clones at 5 , 7 , and 15 days , and compare with TSC2 mutant clones in Figure 1B ) . This indicated an initial increase of proliferative activity in TSC mutant ISCs , followed by a sporadic loss of proliferation in individual ISCs at a later timepoint . Indeed , while many Tsc1Q87X or gigas192 mutant clones , or clones expressing TSC2RNAi , were recovered that contained a single diploid Dl+ ISC even at 15 days after clone induction , at all ages rare clones could also be observed in which the Dl+ cell became large and polyploid , consistent with the phenotype reported by Amcheslavsky et al ( Figure S2B , S2C ) . In our experiments , TSC mutant ISCs did thus not immediately increase in size , but grew and lost function sporadically . This interpretation is supported by the fact that the number of Tsc1Q87X mutant clones observed in the gut declined over time ( Figure 2B ) . To explore why TSC mutant ISCs exhibited a much less penetrant growth phenotype in our experiments as compared to [53] , we tested whether the rate of the spontaneous growth of TSC mutant ISCs might be influenced by dietary conditions , which can modulate TOR activity independently of TSC [35]–[39] . Indeed , the average number of cells generated by Tsc1Q87X mutant ISCs within 7 days was significantly reduced when flies were reared on high yeast food ( HY , 15% yeast ) compared to our regular food ( RF , 2% yeast ) ( Figure 2C , 2D ) . Dl+ cells in these Tsc1Q87X mutant clones became large and polyploid , similar to the phenotype described in [53] ( Figure 2E ) . Clone sizes were also reduced in wild-type flies reared on high yeast food , but this reduction was less significant than the size reduction of TSC deficient clones ( Figure 2D ) . Two recent studies have reported strong effects of yeast , the only protein source in fly food , on ISC activity . Both studies reported increased ISC activity in yeast-fed flies compared to flies completely starved of yeast [33] , [34] . ISCs thus require a protein source to become fully active , yet our results indicate that they can also lose function when protein levels are too high . This effect is significantly enhanced when TSC is lost , indicating that TSC activity isolates the TOR pathway from dietary stimuli in ISCs , maintaining their function . The role of the TSC1/2 complex in ensuring the long-term maintenance of ISCs is thus reminiscent of its function in GSCs [48]–[50] . Interestingly , Tsc1Q87X mutant , or TSC2RNAi or Rheb over-expressing clones were significantly less likely to contain prospero-labeled EE cells than wild-type clones , suggesting that TOR activation also impaired the commitment of EBs into the EE cell fate , or the terminal differentiation of EEs ( Figure 2F , 2G ) . It remains unclear , however , whether this is a consequence of direct TOR pathway-mediated regulation of prospero expression , or of other events required for EE differentiation . Our observations thus suggest that the TSC complex promotes ISC maintenance in varying nutritional conditions , influences commitment into the EE fate , and regulates EC growth in the intestinal epithelium . We hypothesized that these multiple functions of TSC are coordinated by intricate , cell-type specific regulation of TSC activity in the ISC lineage . To start analyzing this regulation , we examined the expression of TSC2 using an anti-Gigas antibody described in [44] . High expression of TSC2 was detected in ISCs ( Dl+ cells that do not express GFP under the control of the RU486-inducible EB/EC driver 5966::GS [63] , Figure 3A ) and in EEs ( pros+ cells , Figure 3C . These cells show even higher TSC2 expression than ISCs ) , and its expression was significantly weaker in EBs ( cells expressing bGalactosidase from a Su ( H ) -GBE::lacZ construct , Figure 3B ) . Consistent with this expression pattern of TSC2 , we found that in wild-type homeostatic conditions , the TOR pathway is highly active in EBs ( compared to ISCs or ECs ) , as determined using an antibody against phosphorylated 4EBP ( Figure 3D , this antibody reliably detects changes in TOR signaling activity , see S3A and [49] ) . Preventing this activation of TOR signaling in EBs was sufficient to impair the formation of normal EBs: over-expression of TSC1/2 or knockdown of S6K ( S6KRNAi ) specifically in EBs and ECs ( using 5966::GS ) , resulted in the accumulation of small Dl+ cells that also express GFP ( Figure 3E ) . Most of these cells had DNA content that was similar to ISCs , indicating that they are diploid or have not completed endoreplication ( Figure S3B ) . The disruption of the normal asymmetric distribution of Dl in ISC/EB pairs indicates that TOR inactivation in ISC daughter cells inhibits differentiation . A similar disruption of normal EB determination was observed when TSC1/2 were over-expressed in EBs only using Su ( H ) -GBE::Gal4 [64] ( Figure 3F , 3G ) . Interestingly , these guts also exhibited a significant increase in the number of pros+ EE cells , indicating that inhibiting TOR activity in EBs is sufficient to alter their commitment from the EC fate into the EE fate ( Figure S3C ) . Combined , our findings suggested that reduced TSC1/2 function in EBs is critical for differentiation of EBs and for lineage commitment into the EC fate . Importantly , these findings also suggested a potential mechanism for TSC2 regulation in the ISC lineage , as down-regulation of TSC2 expression coincides with the activation of N signaling in EBs . We hypothesized that N activation promotes TSC2 down-regulation and tested this idea by over-expressing the N Intracellular Domain ( NICD ) in ISCs and EBs ( using esg::Gal4 , Gal80ts ) . Expression of NICD is sufficient to force differentiation of ISCs into ECs [8] , [9] . Consistently , we found that TSC2 expression was undetectable in most esg::GFP+ cells expressing NICD , while in wild-type intestines , more than 50% of all esg::GFP+ cells express high levels of TSC2 ( Figure 4A , 4B ) . We further tested whether N signaling is required for TSC2 repression in the ISC lineage by over-expressing a dsRNA against N ( NRNAi ) in ISCs and EBs . Expression of NRNAi under the control of esg::Gal4 prevents EB differentiation and results in the formation of ISC tumors characterized by clusters of small , diploid , Dl+ cells [8] , [9] . Cells in these tumors were also TSC2 positive , confirming the correlation between ISC identity and TSC2 expression , and suggesting that N signaling is required for TSC2 repression ( Figure 4C; TSC2 immunoreactivity was suppressed by TSC2RNAi and enhanced by over-expressing both TSC1 and 2 , confirming the specificity of the antibody . Co-expression of TSC1/2 also moderately increased the size of the stem cell tumors , indicating additional enhancement of the NRNAi-caused phenotype ) . The N-responsive transcriptional regulator Su ( H ) has been reported to bind to a cluster of four sites within 1 . 5 kb in the upstream promoter region of the gigas/Tsc2 gene in Drosophila [65] . Su ( H ) -mediated transcriptional repression of gigas/Tsc2 was thus a plausible mechanism for N-induced repression of TSC2 expression in EBs . To test this idea , we assessed the regulation of gigas/Tsc2 in co-cultures of S2 cells that constitutively express N or Dl [66] ( Figure 4D–4E; N activation in N-expressing cells occurs within minutes of exposure to Dl-expressing cells , Figure S4 ) . We first confirmed that Su ( H ) binds to the upstream promoter region of gigas/Tsc2 using chromatin IP ( ChIP , Figure 4D ) , and found significant enrichment of a region proximal to the transcriptional start site in precipitates from cells with activated N signaling . We further measured transcript levels of gigas/Tsc2 and found reduced expression of this gene within 30 min of N activation ( Figure 4E ) . This repression of gigas/Tsc2 was sustained for at least 24 hours . Protein levels of TSC2 ( measured by Western Blot ) did not decrease significantly in S2 cells in these experiments ( not shown ) , indicating that in addition to transcriptional repression , posttranslational mechanisms have to be involved in reducing TSC2 protein levels in vivo as observed in ISCs expressing NICD ( Figure 4A , 4B ) . Importantly , these results suggested that Su ( H ) is a general transcriptional repressor of gigas/Tsc2 expression in Drosophila cells . Accordingly , TSC2 repression in EBs was mediated by Su ( H ) , as inducing ‘Flp-out’ clones [67] expressing Su ( H ) RNAi was sufficient for the formation of tumors containing small Dl+ and TSC2+ cells ( Figure 4F ) . Consistently , gigas/Tsc2 repression in the S2 co-culture system was prevented when Su ( H ) was knocked down by RNAi ( Figure 4G ) . These findings are consistent with a model in which N activation suppresses TSC2 expression in EBs , inducing growth and endoreplication in response to insulin signals . We asked whether TSC2 repression was sufficient and required for ISC differentiation downstream of N , and found that loss of TSC1/2 indeed rescued the tumor phenotype of NRNAi expressing ISCs ( in both MARCM clones , and when driven by esg::Gal4; Figure 5 , Figure S5 ) . N-deficient ISCs generate tumors because they undergo symmetric divisions and thus generate exponentially growing cell clones . Loss of TSC1/2 prevented this accumulation of Dl+ ISCs in N loss of function conditions and converted NRNAi expressing cells into Dl− , polyploid , EC-like cells . Similar to wild-type ECs , these cells also contained brush borders and expressed the EC marker Pdm1 ( Figure 5A–5E , Figure S5A–S5C; brush borders can be observed by staining for phalloidin; Polyploidy measured by intensity of DAPI fluorescence ) . TSC1 suppression is thus sufficient to fully differentiate N-deficient ISCs into ECs . Consistent with a conversion of these cells into a postmitotic state , the number of cells observed in each cluster of NRNAi expressing cells was significantly reduced when TSC1 or 2 were lost ( Figure 5E , 5F ) . These results confirm that loss of TSC1/2 in N loss-of-function conditions is sufficient to promote differentiation of ISCs towards the EC fate . For most analyzed phenotypes , inhibition of TSC1 elicited stronger effects than inhibition of TSC2 , suggesting that the knockdown of TSC2 is less efficient , or reflecting the fact that loss of TSC1 also results in degradation of TSC2 protein , as TSC1 stabilizes TSC2 . While many TSC1/Notch double mutant cells thus are morphologically indistinguishable from wild-type ECs , it is important to note , however , that Notch activation elicits complex gene expression changes in cells , and it remains unclear whether all functional aspects of ECs can be reconstituted in N/TSC1/2 deficient cells . Furthermore , some of these cells retain Dl expression ( see example in Figure 5E ) , indicating that not all of these cells fully differentiate into normal ECs . Loss of TSC1/2 also rescued the accumulation of pros+ EE cells in N loss-of-function conditions ( Figure S5D , Figure S5E ) , confirming a shift towards the EC fate in TSC-deficient EBs . Furthermore , co-expression of TSC1 and 2 resulted in the maintenance of small , diploid , Dl+ cells even in the presence of NICD , showing that TSC2 repression is required for N-induced ISC differentiation ( in both Flp-out clones and when driven by esgGal4 , Figure 6 ) . Repression of TSC1/2 function is thus a critical step in the regulation of EB differentiation .
Our results establish a new mechanism by which lineage commitment , differentiation and growth are coordinated in an epithelial stem cell lineage ( Figure 7 ) . This mechanism allows for the integration of nutritional signals through the IIS and TOR pathways with Notch-mediated differentiation signals: High expression of TSC2 in ISCs prevents differentiation and is thus critical for stem cell maintenance , while reducing TSC activity in EBs is required and sufficient to promote differentiation into ECs . This dynamic regulation of TSC levels in the ISC lineage intersects with the control of TSC1/2 activity by growth factor signals . Based on current models , we propose that control of TSC2 expression is required to set a threshold for the Akt-mediated inactivation of the TSC1/2 complex downstream of growth factor receptors . The TSC2 expression level would thus determine the cellular response to growth signals in the ISC lineage . Supporting this view , ISCs , which express high levels of TSC2 constitutively , do not differentiate in response to InR over-expression , but rather increase their proliferation rate . EBs , on the other hand , express less TSC2 and respond to InR activation by endoreplicating and growing into ECs . Robust expression of TSC2 in ISCs thus prevents premature differentiation and growth of ISCs . When IIS is chronically activated ( as in high nutrient conditions ) , however , Akt-mediated TSC1/2 complex inactivation may cause sporadic differentiation and loss of ISCs . Conversely , ISC maintenance might be improved in conditions of chronically low IIS and TOR activity . The TSC1/2 complex thus acts as a ‘buffer’ that improves ISC maintenance by isolating these cells from changing nutritional conditions . Accordingly , TSC1 deficiency leads to differentiation and loss of ISC function when flies are reared under high yeast conditions . Interestingly , these results also indicate that TOR pathway activation is not constitutive in TSC deficient ISCs , but is still inducible by nutritional changes . It is likely that amino acid-sensing signaling pathways , involving Rag GTPase complexes and the MAP4K3 and Vps34 kinases , regulate TOR in these situations [38] . An effect of TOR signaling on stem cell maintenance has previously been described in GSCs , and is consistent with recent findings that suppression of TOR activity , through rapamycin or genetic means , increases lifespan [48]–[50] , [68] . While reducing IIS activity in the ISC lineage is sufficient to extend lifespan [23] , additional studies will have to be performed to assess the role of TOR signaling in this context . Three pieces of evidence indicate that transcriptional repression of TSC2 occurs downstream of N activation in the ISC lineage: ( i ) TSC2 expression is significantly reduced in EBs with high levels of N signaling activity , ( ii ) forced expression of NICD suppresses TSC2 expression in ISCs , and ( iii ) loss of TSC2 is sufficient to rescue N loss of function phenotypes in the ISC lineage . This regulatory interaction between TSC2 and N signaling is reminiscent of recent findings in Drosophila sensory organ precursors , mouse embryonic fibroblasts , and mammalian cancer cells [46] , [69]–[73] . However , the results reported in these studies indicated that TOR pathway activation could result in increased N cleavage and N pathway activation . In the fly SOP , activation of the TOR pathway phenocopied N gain of function phenotypes , but it was not tested whether these phenotypes were rescued in N loss of function backgrounds . It thus remained unclear whether N activation is a consequence of TOR pathway activation , or whether TOR activation is a required component of the N-induced differentiation pathway in this lineage . Our results demonstrate that in the ISC lineage , TOR activation is sufficient to drive EC differentiation even in the absence of N signaling , supporting a model in which TOR activation occurs downstream of N . In TSC mutant mouse embryonic fibroblasts , on the other hand , TOR -dependent activation of N can be observed , highlighting the close , evolutionarily conserved relationship between these two signaling pathways in the control of cell differentiation , but also suggesting that multiple , context dependent , interaction mechanisms may exist [69] . Our model implies a novel role for Su ( H ) as a transcriptional repressor of the gigas gene . This interpretation is based on the requirement of Su ( H ) for the N-mediated repression of the gigas gene both in S2 cells and in the ISC linage , as well as on the binding of Su ( H ) to the gigas promoter . A function of Su ( H ) as a transcriptional repressor has not previously been described , and additional studies are needed to explore its mechanism . While binding of Su ( H ) to the gigas promoter indicates a direct role in transcriptional repression of gigas , it is possible that Su ( H ) also acts indirectly to repress gigas expression by inducing or cooperating with transcriptional repressors . A candidate group of such repressors are encoded by Su ( H ) target genes , the classical Hairy and E ( Spl ) complex . These transcription factors are induced by Su ( H ) in response to Notch signaling and have been described as transcriptional repressors in other contexts [74] . Putative E ( Spl ) binding sites are present in the gigas promoter ( not shown ) , and additional studies will therefore be of interest to dissect the requirement for individual E ( Spl ) complex genes in the regulation of gigas . Our data further indicate that transcriptional repression of gigas may not be the only mechanism by which TSC2 repression is achieved in EBs . While we find that activation of N is sufficient and required for repression of TSC2 protein in EBs in vivo , our S2 studies indicate that the turnover rate of the TSC2 protein also has to be increased to achieve rapid reduction of TSC2 levels . It can be anticipated that the control of TSC2 ubiquitination by the cul4/ddb1/fbw5 complex may be an important regulatory mechanism here , and it will be of interest to further dissect the interaction of this complex with the N signaling pathway in the ISC lineage [44] . Characterizing these signaling interactions in ISCs in more detail is of significant interest for our understanding of somatic stem cell maintenance , proliferative homeostasis and lineage commitment . The evolutionary conservation of N and TOR signaling , as well as the similarities in the biology of Drosophila and vertebrate stem cell populations [7] , indicate that such understanding will provide important insight into human regenerative and proliferative diseases .
The following fly stocks were obtained from the Bloomington Drosophila Stock Center: w1118 , UAS-InR , UAS-InRDN , UAS-Rheb , UAS-S6KKQ , tub-Gal80ts , UAS-SuHRNAi ( TRiP . HM05110 ) . UAS-TSC1RNAi ( Transformant ID 22252 ) , UAS-TSC2RNAi ( TID 103417 ) and UAS-S6KRNAi ( TID 104369 ) were obtained from the Vienna Drosophila RNAi Center . The following lines were gifts from: Esg-Gal4 , S . Hayashi; UAS-NICD , UAS-NotchRNAi , and hsFlp; tub-Gal4 , UAS-GFP; FRT82B tubGal80 , N . Perrimon; Su ( H ) -GBE-LacZ , S . Bray; Su ( H ) -GBE-Gal4 , S . X . Hou; UAS-TSC1 , TSC2 , M . Tatar; FRT40A , chico1 , FRT82B , InRE19 , and FRT82B , InR353 , and FRT40 , Tor2L1 , FRT40 , Tor2L19 and FRT40 , TorW1251R by D . Drummond-Barbosa; hsFlp; FRT40A , tub-Gal80; tub-Gal4 , UAS-GFP and 5966-GS , B . Ohlstein; w , hsFLP; actin , FRT , y+ , FRT , Gal4 , UAS::RFP , M . Uhlirova; FRT82 , Tsc1Q87X and FRT82 , Rheb2D1 , K . Harvey . Flies were cultured on yeast-molasses based food at 25°C with a 12 hours light/dark cycle . For TARGET experiments flies were raised at 18°C and shifted to the restrictive temperature ( 29°C ) 3–5 days after eclosion . For clone induction ( MARCM and Flp-out ) , 3–5 day old flies were heat shocked at 37°C for 45 minutes . For 5966GS , flies were maintained for 7 days on RU486 food ( 100 µl of a 5 mg/ml solution of RU486 was deposited on top of a 10 ml food vial and dried for 16 hours ) . Guts were dissected in phosphate-buffered saline ( PBS ) and fixed for 45 min at room temperature in 100 mM glutamic acid , 25 mM KCl , 20 mM MgSO4 , 4 mM sodium phosphate , 1 mM MgCl2 , and 4% formaldehyde . All subsequent washes ( 1 hour ) and antibody incubations ( 4°C overnight ) were performed in PBS , 0 . 5% bovine serum albumin and 0 . 1% Triton X-100 . Staining with Delta antibody was performed following the methanol-heptane fixation method described in ( Lin et al . , 2008 ) . Fluorescent in situ hybridization protocol was adapted from [75] using Tyramide signal amplification ( TSA ) and Digoxigenin ( DIG ) labeled RNA probes . The following primers were used to generate RNA probes for pdm1: F 5′-AGT TTG CCA AGA CCT TCA AGC AGC and R 5′-AGG GAT TGA TGC GCT TCT CCT TCT . Primary antibodies with respective dilutions were: From Developmental Studies Hybridoma Bank: mouse anti- Armadillo and anti-Delta , 1∶100; mouse anti-Prospero , 1∶250; Cell Signaling: rabbit anti-phospho-4EBP , 1∶500; ICN: mouse anti-b-galactosidase , 1∶100; gift from Yue Xiong: rabbit anti-Gigas , 1∶500; gift from Yang Xiao-Hang: rabbit anti-Pdm1; Upstate Biotech: rabbit anti phospho histone H3 , 1∶1000; Invitrogen: Alexa Fluor 568 1∶500 Confocal microscopy was performed on a Leica SP5 system . Image processing was done on NIH Image J and Adobe Photoshop . S2 cell lines stably transfected to express wild type Notch receptor ( S2-Mt-N ) or Delta ligand ( S2-Mt-Dl ) from a Cu-inducible metallothionein promoter were obtained from the Drosophila Genomic Resource Center . Both lines were cultured in M3+BPYE medium with 10% heat inactivated Fetal Calf Serum and grown under permanent selection with 0 . 2 µM Methotrexate ( Sigma ) . N and Dl expression was induced separately with 600 mM CuSO4 for 24 hours and the two cell lines were then co-cultured in 1∶1 ratio for the indicated times . For RNAi experiments , double-stranded RNAs were synthesized against GFP and Suppressor of Hairless using T7 promoters ( Ambion MEGAscript RNAi kit ) . S2-Mt-N and S2-Mt-Dl cells were cultured separately and only S-Mt-N cells were treated with GFP dsRNA ( control ) or with SuH dsRNA for 3 days . After two days of dsRNA treatment both cell lines were induced with 600 mM CuSO4 for 24 hours . After three days of dsRNA treatment , S2-Mt-N cells and S2-Mt-Dl cells were co-cultured at a 1∶1 ratio for the indicated times . Knockdown of Su ( H ) was confirmed by RT-PCR ( not shown ) . ∼1×107 cells were collected from triplicate cell cultures of S2-Mt-N cells ( control ) or of 2 h co-cultures of S2-Mt-N and S2-Mt-Dl . Cells were cross-linked using ∼1 . 1% formaldehyde and ChIP was performed using the abcam ChIP kit ( ab500 ) . Cells were sonicated on ice using a Branson Sonicator ( power 4 , 50×10 second pulses with 30 second intervals; average size of genomic DNA fragments was ∼500 bp ) . Sheared chromatin was incubated with 5 µg of rabbit anti-GFP ( invitrogen; negative control ) and goat anti-SuH ( Santa Cruz Biotechnology ) for 24 hours and then precipitated using Protein G Sepharose ( Fast Flow; Sigma ) . De-crosslinking and DNA purification was performed according to kit instructions ( ab500 ) . DNA from different ChIP samples was analyzed for enrichment using real time PCR using the following primer sets: gig 1: 5′-ACAAACGCAAAGTTGGCGAC-3′ and 5′-GTGTGCAACCAGTAATTCCTAGCC-3′; gig 2: 5′-AAGTTGTTCCTCAAATCGCTGCCG-3′ and 5′-ATTGAAGTTGTGCAGCTGCGTGTC-3′; actin5C: 5′-ATTCAACACACCAGCGCTCTCCTT-3′ and 5′-ACCGCACGGTTTGAAAGGAATGAC-3′ . Total RNA was extracted using Trizol . cDNAs were synthesized using oligo-dT primers and real-time RTPCR was performed on a BioRad iQ5 detection system ( using SYBR Green and ΔΔCt quantification method ) . Gigas and Suppressor of Hairless expression levels were quantified relative to Actin5c expression . Cell samples were resolved using 5% ( for NICD ) or 10% ( for tubulin ) SDS-polyacrylamide gel electrophoresis , transferred to nitrocellulose membranes using semi-dry transfer , and probed with the following primary antibodies: mouse anti-NICD ( DSHB , 1∶10 , 000 ) , mouse anti-alpha-tubulin ( Sigma , 1∶5000 ) . Antibodies were detected using horseradish peroxidase-conjugated secondary antibodies and the ECL detection system ( Amersham ) .
|
Stem cells maintain tissue homeostasis in metazoans . A productive model to study the regulation of stem cell function is the Drosophila posterior midgut . Notch ( N ) signaling controls intestinal stem cell ( ISC ) differentiation in this tissue , while ISC proliferation is regulated by growth factor signaling pathways , including Insulin/IGF signaling ( IIS ) . In this study , we explore the interaction between growth signals and N signaling in the control of ISC proliferation and differentiation . We show that TOR signaling , which promotes growth and can be activated by the IIS pathway , is maintained in ISCs in an inactive state by high expression of the TOR inhibitor TSC2 . TSC2 expression shelters ISCs from nutritional cues , ensuring their long-term maintenance . In response to N pathway activation in enteroblasts ( EB ) , the ISC daughter cells , TSC2 is transcriptionally repressed and TOR is activated . We demonstrate that this negative interaction between N and TSC2 is required and sufficient for differentiation of EBs into enterocytes ( ECs ) , the absorptive cells of the epithelium . Our findings establish a critical role for TSC in ISC maintenance and provide a mechanism by which N promotes differentiation into the EC fate . The human homologue of TSC2 is an important tumor suppressor , and our study provides new insight into how its regulation controls regenerative processes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"insulin-dependent",
"signal",
"transduction",
"mitogenic",
"signaling",
"tor",
"signaling",
"cell",
"differentiation",
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"stem",
"cells",
"cell",
"growth",
"signaling",
"pathways",
"adult",
"stem",
"cells",
"signaling",
"in",
"cellular",
"processes",
"biology",
"signal",
"transduction",
"cell",
"biology",
"cellular",
"types",
"molecular",
"cell",
"biology",
"cell",
"fate",
"determination"
] |
2012
|
Notch-Mediated Suppression of TSC2 Expression Regulates Cell Differentiation in the Drosophila Intestinal Stem Cell Lineage
|
Synthetic lethality is a genetic interaction wherein two otherwise nonessential genes cause cellular inviability when knocked out simultaneously . Drugs can mimic genetic knock-out effects; therefore , our understanding of promiscuous drugs , polypharmacology-related adverse drug reactions , and multi-drug therapies , especially cancer combination therapy , may be informed by a deeper understanding of synthetic lethality . However , the colossal experimental burden in humans necessitates in silico methods to guide the identification of synthetic lethal pairs . Here , we present SINaTRA ( Species-INdependent TRAnslation ) , a network-based methodology that discovers genome-wide synthetic lethality in translation between species . SINaTRA uses connectivity homology , defined as biological connectivity patterns that persist across species , to identify synthetic lethal pairs . Importantly , our approach does not rely on genetic homology or structural and functional similarity , and it significantly outperforms models utilizing these data . We validate SINaTRA by predicting synthetic lethality in S . pombe using S . cerevisiae data , then identify over one million putative human synthetic lethal pairs to guide experimental approaches . We highlight the translational applications of our algorithm for drug discovery by identifying clusters of genes significantly enriched for single- and multi-drug cancer therapies .
Synthetic lethality ( SL ) occurs when two nonessential genes cause cellular inviability after being knocked out simultaneously . [1] Although SL was originally studied and described in yeast , it can be a powerful tool for studying drug action in humans; for example , SL may guide the development of cancer combination therapy[2 , 3] and inform drug-drug interactions . SL interactions may differ between cellular contexts;[4] a gene pair that is SL in one cell type may not be SL in another . This can provide a tremendous therapeutic boon when two drugs targeting two gene products mimic an SL interaction in cancer cells and leave healthy cells unaffected . However , drug-induced SL interactions may also cause adverse events via unexpected cell death . Thus , mapping SL in humans is necessary to understanding mono- and polypharmacological effects . Most gene pairs have not been interrogated for SL in humans , and several factors impede a species-wide evaluation of SL . These include the ethical implications of studying SL directly , the inability to discern state-specific SL interactions from global ones in experimental cell lines ( e . g . cancer[4 , 5] ) , and–most significantly–the heavy experimental burden . Over 200 million assays would be required to determine the SL status of all human gene pairs in just a single cellular context . In silico methods are therefore necessary to guide the identification of SL in humans . Previous work on leveraging model organisms to predict human SL has focused in particular on genetic homology , under the hypothesis that SL status will be maintained between orthologous gene pairs . [6] This approach has two major limitations . First , there are only ~2 , 000 genes that genetically homologous between S . cerevisiae and humans ( NCBI Homologene[7] ) . These homologues account for a mere 1% of all possible human pairs , leaving the majority with no predictive data regarding SL status . Second , genetic redundancies that developed independently in each species since deviation from a common ancestor may affect synthetic lethal status . For example , 228 gene duplication events have been suggested between S . cerevisiae and S . pombe[8]in the ~400 million years of evolution between the two species;[9] this number is likely even higher between S . cerevisiae and humans . Each of these events may introduce a functional redundancy that alters SL relationships in the organism by causing a gain or loss of SL . Focusing solely on genetic homology does not account for these complexities . In this work , we first evaluate the performance of genetic homology in predicting SL . We also consider structural similarity using protein structure families , domain similarity using protein domains , and functional similarity with gene ontology annotations . We additionally consider information centrality , a univariate network-based model . We show that homology , structural similarity , and information centrality are limited in their ability to predict SL . We then introduce the concept of connectivity homology , a measure of relatedness between genes that is independent of structure , function , or genetic homology . Relationships between genes and proteins , including redundancies , may be illustrated through the use of biological networks , and we hypothesize that the network connectivity profiles between two genes will better characterize their potential for an SL relationship . In order to compute these profiles , we use well-known graph properties , such as degree centrality and shortest path . [6 , 10–12] We next use machine learning and data from a well-studied organism , S . cerevisiae , to train a model of synthetic lethality that can be applied to any species of interest . We show that our algorithm , Species-INdependent TRAnslation ( SINaTRA ) , significantly outperforms previously published models of predicting SL in translation . Importantly , we can predict synthetic lethality in species without any known ( i . e . experimentally validated ) synthetic lethal pairs . The only necessary information is an experimentally derived protein-protein interaction ( PPI ) network . We also show that the method is robust to network incompleteness . We then use SINaTRA to predict SL in humans and assign each human gene pairs a score between 0 and 1 , indicating the likelihood that the two genes exhibit a synthetic lethal relationship . As a post-processing step to enrich our predictions , we use databases of population genetic variation in humans to filter out likely false positives . Finally , we evaluate of the biomedical implications of our human SL gene pairs by discovering “hot zones” of putative SL pairs that suggest novel cancer combination therapies .
We began our study by considering two published methods of predicting SL , protein homology[13] and bi-nodal information centrality , [8 , 14] and implemented the algorithms as described by the authors . In addition , we hypothesized that structural homology , domain homology , and functional homology may be able to predict SL and designed models based on these parameters for comparative analysis . In Wu et al . , [9 , 13] the authors constructed a model to predict SL in S . cerevisiae , then hypothesized that human gene pairs homologous to SL pairs in S . cerevisiae would also be SL in humans . We implemented the latter part of the approach and evaluated it by predicting SL in S . pombe . By restricting our analysis to only genes that are homologous between S . cerevisiae and S . pombe , we find a significant predictive effect ( OR = 145 , 95% CI: 93–219 , p < 2 . 2e-16 , Fisher’s exact test ) , corresponding to an area under the receiver operating characteristic curve ( AUC ) of 0 . 60 . Model performance decreased to OR = 45 . 9 ( p<2 . 2e-16 ) and an AUC = 0 . 52 when expanding the model to include all gene pairs ( Materials and Methods ) . We next hypothesized that structural , domain , and functional similarity may be predictors of SL . We trained these models in S . cerevisiae and applied them to S . pombe . We used SCOP protein classifications to describe the former , and assigned each gene pair a value between 0 ( no similarity ) and 4 ( same class ) based on their products’ structural similarity . The model was trained and tested only on pairs with SCOP data associated with both genes . Only 399 SL pairs and 109 , 357 NSL pairs had SCOP data for S . cerevisiae ( 16 , 765 , 399 pairs skipped ) and 2 SL/298 NSL for S . pombe ( 1 , 840 , 021 pairs skipped ) . The SCOP-based model had an AUC of 0 . 62 . We additionally created a domain-based model from PFam[15] to predict SL . Domain data exists for a larger number of proteins ( 9 , 424SL/10 , 280 , 492 NSL in S . cerevisiae; 514/1 , 431 , 764 for S . pombe ) , allowing us to score more pairs than the SCOP-based model ( Materials and Methods ) . The AUC in the domain-based model was 0 . 56 . We described functional homology using annotations from Gene Ontology ( GO ) ( Materials and Methods ) . Functional similarity attained an AUC of 0 . 81 . Finally , we calculated the pairwise information centrality[14] in S . pombe and found no significant predictive performance identifying SL pairs ( AUC = 0 . 46 , Logistic Regression ) . Bi-nodal information centrality did not require interspecies translation . We hypothesized that multivariate , network-based models of synthetic lethality would be able to capture SL interactions both within and between species more successfully . We define two proteins as being connectivity homologous if they share similar connectivity profiles in their respective networks . A connectivity homologous relationship may exist between two proteins in the same species , or between proteins of different species . This concept can be generalized for pairs of proteins , or even groups of proteins ( i . e . modules ) . For example , two pairs of proteins may be connectivity homologous because both pairs are connected to each other in a similar way . We illustrate this concept in Fig 1A , where we present two networks of different sizes and topologies . Two network parameters are used to describe the network: degree ( deg . ) and betweenness centrality ( bet . cent . ) . Each node contains its connectivity profile as a vector depicting the degree and betweenness centrality as ‘low’ ( blue ) , ‘medium’ ( white ) , or ‘high’ ( red ) . Although the values may not be immediately comparable between networks , it is obvious that certain nodes share similar connectivity profiles , while some nodes do not have an interspecies , connectivity homologous pair ( high deg . /medium bet . cent . ) . In this paper , we represent connectivity profiles using vectors of network parameters . Each gene is represented by a vector of eight parameters ( Tables 1 and S1 ) . Each gene pair is represented by a vector of four node-pair parameters ( Tables 1 and S1 ) as well as the individual profiles for each gene in the pair , leaving each pair with a connectivity profile defined by 20 network parameters . For the purposes of this investigation , we chose to use protein-protein interaction ( PPI ) networks because of the wide availability of data across many species . PPI data was downloaded from BioGRID[16] to construct graphs of one connected component ( Materials and Methods ) . We computed the connectivity profiles for 5 , 810 proteins in S . cerevisiae , 1 , 919 in S . pombe , 4 , 233 in M . musculus , and 14 , 820 proteins in humans as well as for 16 . 8 million , 1 . 8 million , 8 . 9 million , and 109 . 8 million pairs of proteins for S . cerevisiae , S . pombe , M . musculus , and humans , respectively . We found that the distributions and ranges of network parameter values differed significantly between species ( S1 Fig; S2 Table ) . To correct for these differences ( S2 Fig ) , we evaluated four normalization strategies ( Table 2 ) and chose to use rank normalization to rescale the values of each parameter between 0 and 1 . Rank normalization makes parameter values comparable between species . We refer to normalized data as being “translated . ” We found that proteins with similar connectivity profiles ( i . e . those that are connectivity homologous ) were more likely to share functional annotations . We used the Euclidean distance between connectivity profiles as a measure of connectivity homology ( Materials and Methods ) . We compared this distance between genes that shared genetic homology ( orthologs ) and specific functional annotations ( Gene Ontology [GO] ) between S . cerevisiae and S . pombe ( Sc/Sp ) ( S3A Fig ) and between S . cerevisiae and humans ( Sc/H ) ( S3B Fig ) . We found that proteins annotated with the same function had significantly had significantly lower distances ( Sc/Sp median = 1 . 04 , Sc/H median = 0 . 92 ) than those annotated with different functions ( Sc/Sp median = 1 . 08 , p<2 . 2e-16; Sc/H median = 1 . 04 , p<2 . 2e-16 ) . This result holds even when orthologs are not considered . Non-orthologous genes annotated with the same function had significantly lower distances than non-orthologous genes annotated with different functions ( S3 Fig , p<2 . 2e-16 ) . We also found that orthologs had significantly lower distances than non-orthologous pairs ( S3 Fig , p<2 . 2e-16 ) . These differences were consistent across all levels of functional specificity ( S4 Fig ) . These results suggest that network substructure , and therefore network signals , are conserved between species based on both homology and function .
Previous work on interspecies SL prediction has focused on the use of genetic homology . [33] We found that the method has fairly high predictive power between S . cerevisiae and S . pombe when considering only gene pairs with known homology ( Fig 2B ) . Unfortunately , many genes have no known homology information and , because of this , the model performance suffers when considering all interspecies gene pairs . An additional complication stems from genes with multiple homologues , resulting in ambiguous predictions . In an effort to address some of these challenges with using established orthologs , we also implemented two additional methods: one using shared structural domains , and one derived from structural families . Neither method outperformed SINaTRA . The most successful comparison method was the number of shared functional annotations in the Gene Ontology ( AUC = 0 . 81 ) , which performed almost as well as SINaTRA ( AUC = 0 . 86 ) . We additionally found that the information contained in the functional annotations and SINaTRA was not redundant , suggesting that a model that combines connectivity profiles with functional annotations may yield better performance . In this paper , we introduce the idea of connectivity homology , which exists when two genes share a similarity connectivity patterns quantified by network and graph theoretic parameters . We performed a small exploration of connectivity homology and its relation to genetic homology and function , and found that homologous genes and genes that share function exhibit higher connectivity homology ( S3 Fig ) . We hypothesized that there are connectivity patterns between pairs of genes that are indicative of a synthetic lethal relationship . These patterns are discovered using supervised machine learning in a source species–one where synthetic lethality has been well-characterized–and then identify these patterns in a target species to predict synthetic lethal pairs of genes . We performed a small exploration of connectivity homology and its relation to genetic homology and function , and found that homologous genes exhibit higher connectivity homology; in turn , interspecies gene pairs that share the same specific function have higher connectivity homology than interspecies gene pairs of different functions ( S3 Fig ) . We validated our approach in two species where SL has been experimentally explored ( S . cerevisiae and S . pombe ) . We found that our approach , called SINaTRA , significantly outperformed published methods at predicting SL genes in the target species and we achieve precision up to 150 times higher than expected by chance . This precision increased to over 250 times higher than chance when using additional biological priors . Several mechanisms of synthetic lethality have previously been proposed;[34] these include within complex , parallel pathways , and essential linear pathways . Hints regarding the mechanism driving a given gene pair may be provided by our connectivity parameters . Our data suggest that function-specific network substructures are different , and may be related to trends of SL mechanism within a function . For example , metabolism has a much higher proportion of ‘unknown’ pathway annotations than does apoptosis ( S13 Fig ) . This suggests that putative metabolic SL pairs act through parallel pathways , while apoptotic pairs may act through within-complex mechanisms . Further , gene pairs in apoptotic pathways are farther apart and have lower communicability than gene pairs in metabolic pathways , which may also change how many SL pairs are likely to exist with that function . We also observe that a fraction of the predicted SL pairs had between-pathway interactions , where members of an SL pair do not share any single function ( Fig 4 ) . The respective gene products may act at an interface between two related functions; the putative SL pair may be a false positive; or–most interestingly–one ( or both ) genes have previously unidentified functions that cause their SL behavior . One such example is the putative SL pair , BAIAP2 ( insulin receptor signaling; UniProt DB ) and ALDH7A1 ( protection from oxidative stress; UniProt DB ) ( SINaTRA score: 0 . 957 ) . Oxidative stress is associated with insulin resistance , [35] and knocking out both of these genes may mimic or exacerbate insulin resistance , leading to complications and adverse events . For very rare biological phenomena , it is essential to consider the false positive rate of any experimental or computational approach . An unbiased random selection of gene pairs would yield approximately 1 synthetic lethal pair for every 1 , 000 tested . If biased by biological priors , such as limiting the analysis to pairs of genes whose products are partners in protein complexes , this yield may increase 8-fold , to 1 out of every 125 pairs tested . The SINaTRA score we present can also be used as a biological prior . In this case , it is the connectivity pattern of the pair of proteins that makes them more likely to participate in a synthetic lethal interaction . For example , a score of 0 . 85 or greater would yield approximately 1 SL for every 10 pairs tested . If this is coupled with protein complex prior , this could improve to 1 out of every 3 or 4 pairs tested . Combined with other biological priors , the SINaTRA score can be a powerful tool for directing experimental exploration of synthetic lethality . Fig 2D illustrates this expected hit rate versus the number of experiments that would be necessary . These scores can be used to guide experimental exploration depending on the throughput and cost of the experimental approach . Biological contexts , such as tissue type and disease state , can influence synthetic lethal interactions . [4] At this time , cellular and tissue specificity are not captured by the SINaTRA model . However , we can customize our predictions for a given cell or tissue by pruning away any predicted genes that are known not to be expressed in the given context . We used the Protein Atlas[36] to perform this customization and found that certain tissues and cell types had significantly more or fewer SL pairs filtered . These deviations may suggest tissue or cell types that are particularly robust , or susceptible , to SL interactions . For example , respiratory epithelial cells and endothelial cells have many more SL pairs filtered out than expected by chance; this suggests that the tissues are not as susceptible to SL reactions–a hypothesis that requires further investigation . For nearly a decade , leveraging synthetic lethal relationships specific to cancer cells has been a strategy in drug discovery . Therefore , we applied our predictions of synthetic lethality to the study of pharmacology . We found that many cancer combination therapies currently in the clinical pipeline target genes with high SINaTRA scores , suggesting that they use mechanisms of synthetic lethality as their modes of action . Clustering reveals hotspots of high SINaTRA scores that are significantly enriched for combination therapies under investigation . Importantly , our algorithm was able to identify these without any a priori knowledge of the drug combination . Gene pairs found in these hotspots that have not been previously investigated may be promising leads for novel polyphamacological treatments . Our method for predicting SL relies on the availability of protein-protein interaction data . Due to the high-throughput experimental techniques , such as tandem affinity purification and yeast two-hybrid , these are some of the most widely available–omic data . However , comprehensive networks are only available for a handful of species . Future expansions of the approach will focus on integrating other available data , such as genetic sequence or gene expression . These other data sources may help address the issue of context-specificity in our predictions . In this study , we used 12 distinct graph theoretic parameters to describe each gene pair . The choice of these parameters was based on what was available and has been used in prior work , and is not an exhaustive list . Other methods for computing connectivity may be incorporated in future versions of the algorithm , such as spectral methods . In summary , the methodology presented in this paper can help to inform a wide variety of studies in human health by fully utilizing information gathered in model species . In particular , the differential mechanistic analysis that highlights how biological functions may be targeted using synthetic lethality and the “hot spots” of drug synergy highlighted by our cancer therapy analysis indicate promising areas for novel therapeutics . We provide the SINaTRA scores for almost 110 million human gene pairs as a freely available resource for basic and translational science .
We downloaded protein homology data from Homologene , [37] protein structure data from SCOP , [29 , 38 , 39] and GO data from Entrez . [4 , 40 , 41] We used PFam[15 , 42] data for protein domain similarity; IDs were mapped to Entrez gene IDs for S . cerevisiae and S . pombe using DAVID . [32 , 33 , 43 , 44] We calculated binodal information centrality for each gene pair based on Kranthi et al . [14] In order to create the homology-based model , we replicated a previous paper[45] that defined a gene pair as SL if its homologous pair in another species is SL . Gene pairs were defined as SL if the homologous pair in the source species was SL . In the case of multiple homologous pairs in the source species , gene pairs described by the fraction of homologous pairs defined as SL . A pair score >0 resulted in a classification of SL . Homology-based models use only genes with known homologs between the two species of interest . Whole-genome , homology-based models are the union of all genes in the homologous dataset with all genes that appear in our protein-protein interaction network . Genes with no known homologs are given a feature value of 0 . Protein similarity was defined using values between 0 ( no match ) and 4 ( same class ) according to SCOP annotations . Functional similarity was defined using GO process and function terms , excluding “molecular_function” and “biological_process . ” Gene pairs were assigned a value based on the number of overlapping GO terms assigned to each gene . Using PFam domain data , we used the size of PFam ID overlap ( range: [0 , 8 ) ) for within-species gene pairs . For SCOP- , GO- , and PFam-based models , we trained the logistic regression model on S . cerevisiae and applied it to S . pombe . The homology-based model was already “translated , ” and the model was trained and tested in S . pombe alone using logistic regression and five-fold cross-validation . Information centrality does not require translation and was calculated in S . pombe alone; the model was constructed using logistic regression and tested with five-fold cross-validation . Fig 1A was drawn with , and network parameters of each network were calculated using Cytoscape . [46] Regular normalization of a parameter returns each value divided by the maximum value of that parameter , such that each value is between 0 and 1 . To rank-normalize data for a given species , we calculated all individual single- and two-node parameters . Then , for each parameter , we ranked all calculated values from smallest to largest , resolving ties at random . We then divided all values by the total number of genes in the network ( for single-node parameters ) or the total number of gene pairs ( for node-pair parameters ) . This resulted in all genes or gene pairs having all parameter values be a value between 0 and 1 . Tied-rank normalization assigns the median rank to all equal values , then normalizes single-node parameters by the number of genes in the network , and node-pair parameters by the total number of pairs . Quantile normalization is described in previous work , [47] where networks with fewer nodes/edges are up-sampled . Rank-normalized translation and construction of models is illustrated in S16 Fig . We defined a vector of single-node network parameters ( see Table 1 ) for each gene in the S . cerevisiae , S . pombe , and human networks . We calculated the connectivity homology of each interspecies node pair using Euclidean distance . A lower distance implies greater connectivity homology ( similarity ) . We first divided all gene pairs into same specific function or different specific function . We then further divided these groups into homologous/non-homologous . Specific functions were defined as all GO terms related to process or function ( excluding molecular_function or biological_process ) where the number of genes annotated with that GO in each species was less than or equal to a given cutoff . This cutoff was set to 100 at first , then expanded to cutoffs of 10 , 15 , 20 , 25 , 50 , 75 , 100 , 150 , 200 , 250 , 500 , and 750 . We generated PPI networks using data gathered from BioGrid;[16 , 36] each node represents a gene , while edges represent a physical interaction between gene protein products . We pruned all networks to contain one connected component . BioGrid additionally provided SL data used in this investigation . Saccharomyces cerevisiae had over 14 , 000 unique SL pairs and Schizosaccharomyces pombe have over 700 , while Mus musculus and Homo sapiens have 14 and 1 pairs , respectively . Gene pairs may have one of two classes: SL or non-SL . Because of the scarcity of SL pairs , pairs not explicitly labeled as SL are considered non-SL . We used the NetworkX ( version 1 . 8 . 1 ) package for Python[19 , 37] to calculate all network parameters except shared neighbors , shared non-neighbors , and shared 2nd-degree neighbors , which were elucidated from adjacency matrices for each network . All single-parameter classifiers employ logistic regression due to its high interpretability and simple nature . We implemented multi-parameter classifiers using random forests , [17] which are accurate and efficient on large datasets , as well as resistant to over-fitting data . We used five-fold cross-validation in classifier construction , where training occurs with 80% of the data , and classifier evaluation uses the remaining 20% . Finally , to avoid positional bias in case of a single node having exceptionally high values , we shuffled the order in which each single-node parameter appears . We calculated parameter importance using the built-in function from Python’s sklearn package . We predicted SL within a species using the network parameters defined in Table 1 without any normalization ( raw ) as the features of the classifier , and experimental data from BioGrid[16] as the known classes . From these , we performed five-fold cross-validation by randomly selecting 1/5 of the data on which to train our classifier , and testing it on the remaining 4/5 . We trained models using logistic regression or random forest . To predict synthetic lethality , we trained classifiers on raw and translated parameters of our source species , using SL status downloaded from BioGrid as labels . We then applied the classifier to data from our target species . Here , S . cerevisiae is the source species , and we used its network parameters to train classifiers . S . pombe is the target species . Classifier inputs were vectors of network parameters . Synthetic lethality is expected to occur in 1/1000 gene pairs in diploid organisms; therefore , the PPV expected by chance is 0 . 001 . We calculated positive predictive value ( PPV ) , the fraction of true positives out of all called positives , on all S . pombe gene pairs , and on all gene pairs in the same complex . We selected 1000X the number of NSL pairs as SL pairs and bootstrapped the 99% CI of the PPV for both untranslated and SINaTRA-based predictions . To calculate PPV at each cutoff C , gene pairs with SINaTRA ≥ C were considered to be SL , while pairs with SINaTRA < C were considered NSL . Complex membership was identified by using the Entrez GO database , and filtering all GO terms that contained the word “complex” and were in the “component” category . This amounted to 8 , 365 pairs , of which 5 , 806 appeared in our network . 46 of these were experimentally known SL pairs , leaving a ration of approximately 3:400 SL:NSL . We estimated that , because many SL pairs are unknown in S . pombe , the ratio of SL:NSL in within-complex pairs will be approximately 1:50 , and selected SL:NSL pairs in a ratio of 1:50 in order to estimate within-complex PPV . This simulation was performed 1 , 000 times to identify the 99th percentile CI . We additionally plotted the PPV of SL prediction using genetic homology , structural similarity , functional similarity , and information centrality . The expected PPV of all of these were calculated using SL:NSL gene pairs in ratios of 1:1000; because the cutoffs occurred in a range significantly smaller than [0 , 1] , we selected the cutoff that would provide the optimal PPV for the given model ( all pairs ) , then calculated the PPV when adjusting for SL:NSL ratio . The PPV of genetic homology was calculated using only S . pombe pairs that have homologs in S . cerevisiae . We identified the true and false positives and negatives for homology and whole-genome homology as follows: if the input score was >0 and the target species pair was SL , it was a true positive; else it was a false positive . If the input score was 0 and the target species was NSL , it was a true negative; else , it was a false negative . In whole-genome models , all node pairs with no homology information for at least one node were given a score of 0 . Odds ratios were calculated using confusion matrices of form [[TP , FP] , [FN , TN]] and Fisher’s exact test . For whole-genome SINaTRA methods , if the gene pair SINaTRA score ≥ given cutoff and the target species pair was SL , it was a true positive; else , it was a false positive . If the gene pair SINaTRA score < given cutoff and the target species pair was NSL , it was a true negative; else , it was a true positive . In a whole-genome SINaTRA model , nodes that appeared in the Homologene database but not in the network were assigned SINaTRA scores of 0 . We identified the expected number of unidentified SL pairs in S . pombe by taking the PPV at each SINaTRA cutoff and multiplying it by the number of putative hits at that cutoff . We then transformed this cumulative plot into bins , such that for cutoff C , the number in that bin represents all expected pairs with C ≤ SINaTRA < C+0 . 05 . We ablated the S . pombe network to 90 , 80 , 70 , 60 , and 50% of its original side by removing ( 100-N ) % edges at random . We trained a random forest classifier on the complete S . cerevisiae network and tested it on the ablated S . pombe networks and measured classifier success again using AUROC . We plotted the median SINaTRA score of genes in S . cerevisiae , S . pombe , and humans by the node’s degree , popularity ( the number of times it appeared in the BioGRID database ) , and normalized popularity ( degree/popularity ) . We calculated the Spearman correlation coefficient for all plots , for all species . We predicted SL pairs in mice as we did with S . pombe , using S . cerevisiae as the source species . We calculated network parameters using the NetworkX version 1 . 8 . 1 . We performed statistical analysis in R version 3 . 0 . 2 . De Long’s test for comparing ROC curves was implemented using the pROC library . [49] Scripts use Python version 2 . 7 . 5 . Graphics were generated using Python’s Matplotlib . [50] BioGrid release 3 . 2 . 104 was used in all analyses .
|
Synthetic lethality is a genetic interaction that has promising implications for informing novel cancer therapies . Over 200 million pairwise tests would be required to identify all pairwise synthetic lethal interactions in humans–currently , an impossibly large experimental burden . To simplify the process , we have developed a method to predict human synthetic lethal pairs in translation from a well-studied species to one in which synthetic lethality is understudied using both species’ protein-protein interaction networks . Here , we explore the model’s success in translation from S . cerevisiae to S . pombe . We then predict human synthetic lethality and suggest novel areas of inquiry for cancer therapies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Connectivity Homology Enables Inter-Species Network Models of Synthetic Lethality
|
Yeast WHI2 was originally identified in a genetic screen for regulators of cell cycle arrest and later suggested to function in general stress responses . However , the function of Whi2 is unknown . Whi2 has predicted structure and sequence similarity to human KCTD family proteins , which have been implicated in several cancers and are causally associated with neurological disorders but are largely uncharacterized . The identification of conserved functions between these yeast and human proteins may provide insight into disease mechanisms . We report that yeast WHI2 is a new negative regulator of TORC1 required to suppress TORC1 activity and cell growth specifically in response to low amino acids . In contrast to current opinion , WHI2 is dispensable for TORC1 inhibition in low glucose . The only widely conserved mechanism that actively suppresses both yeast and mammalian TORC1 specifically in response to low amino acids is the conserved SEACIT/GATOR1 complex that inactivates the TORC1-activating RAG-like GTPases . Unexpectedly , Whi2 acts independently and simultaneously with these established GATOR1-like Npr2-Npr3-Iml1 and RAG-like Gtr1-Gtr2 complexes , and also acts independently of the PKA pathway . Instead , Whi2 inhibits TORC1 activity through its binding partners , protein phosphatases Psr1 and Psr2 , which were previously thought to only regulate amino acid levels downstream of TORC1 . Furthermore , the ability to suppress TORC1 is conserved in the SKP1/BTB/POZ domain-containing , Whi2-like human protein KCTD11 but not other KCTD family members tested .
The understudied Whi2 protein of Saccharomyces cerevisiae and its fungal homologs share predicted domain structure and sequence similarity with the family of human KCTD proteins [1] ( S1 Fig ) . Despite primary sequence divergences , a homologous SKP1/BTB/POZ domain is identifiable ( IPR011333 ) in the N-terminal portion of both protein types . KCTD family members have been associated with several types of cancer , epilepsy and other disorders [2–5] . However , the functions of these proteins in any species are not understood and the mechanisms of disease caused by KCTD mutations are unknown . Early studies revealed that yeast Whi2 is required to halt the cell cycle , to suppress cyclin expression , and for entry into stationary phase [6–9] . Another study found that Whi2 is important for handling general environmental stresses by interacting with the protein phosphatase Psr1 , which has been suggested to dephosphorylate and activate the general stress response transcription factor Msn2 [10] . More recently , Psr1 and Psr2 were reported to inhibit plasma membrane transporter activity in response to TORC1 inhibition [11] . However , Whi2 , Psr1 and Psr2 were not previously reported to function as upstream regulators of TORC1 . Diverse functions of mammalian KCTD family proteins have been reported , including KCTD13 and TNFAIP1 binding to Rho GTPases Rnd2/3 during brain development [12] , KCTD10 binding to PCNA in DNA repair [13] , Drosophila Insomniac ( KCTD5 ) interacting with cullin 3 in sleep regulation [14] , and KCTD8 , KCTD12 , and KCTD16 interacting with GABAB G-protein coupled receptors in neurons [15] . In addition , mouse Kctd11 was identified as a tumor suppressor gene near TP53 in an in vivo mouse screen [16] , and human KCTD11 deficiency is suggested to contribute to several cancers including medulloblastoma [2] , hepatocellular carcinoma [17] , and prostate adenocarcinoma [18] . We identified yeast WHI2 by sequencing the genomes of several knockout strains of the mitochondrial fission factor Fis1 to identify the spontaneous mutations responsible for a robust growth phenotype . Remarkably , three independent Δfis1 strains each had a different premature stop codon in WHI2 [19] . Spontaneous WHI2 mutations are not responsible for the mitochondrial fission defects due to FIS1 deletion , but are responsible for two seemingly contradictory phenotypes . WHI2-deficiency causes sensitivity to multiple stresses , including heat , acetic acid , killer viruses and ROS/H2O2 , and paradoxically also causes excessive overgrowth compared to wild type in media containing reduced amino acid levels [1 , 19–21] , Our findings are consistent with previous cell cycle and cell stress studies [6–9 , 22] . We also identified yeast WHI2 in a genome-wide screen for genes required to suppress cell growth/division under low amino acid conditions [1] . This screen also identified the deletion strains of NPR2 and NPR3/RMD11 [1] , which are components of the conserved protein complex SEACIT ( human GATOR1 ) known to actively suppress TORC1 kinase activity and cell growth when the availability of environmental amino acids declines [23 , 24] . Under low amino acid conditions , the conserved GTPase activating ( GAP ) function of SEACIT ( human GATOR1 ) inhibits the TORC1-activating RAG-like GTPase Gtr1 ( human RAGA/B ) [25 , 26] . Additional negative regulators of mTORC1 in this same amino acid signaling pathway have been identified in mammals , such as Sestrin2 [27] , the KICSTOR complex [28] and CASTOR1/2 [29] , but lack obvious orthologs in fungi with rare exception [30] . An alternative negative regulator of TORC1/mTORC1 in some yeast species and in mammals is the TSC complex , a GAP for the TORC1-activating GTPase Rheb [31] , but TSC1/2 are not found in Saccharomyces cerevisiae . Therefore , we investigated a potential role for Whi2 in this SEACIT–Gtr1/Gtr2–TORC1 pathway . Instead , we found that Whi2 suppresses TORC1 independently of the SEACIT-Gtr pathway and independently of the PKA pathway . However , both Whi2 and SEACIT-Gtr can work in concert to suppress cell growth specifically in response to low amino acids . Although the detailed molecular mechanisms are not known , we show that Whi2 suppresses TORC1 activity through its binding partners , protein phosphatases Psr1 and Psr2 , revealing a new role for Psr1/Psr2 upstream of TORC1 . Furthermore , both exogenous and endogenous human KCTD11 , a Whi2-like protein harboring a homologous SKP1/BTB/POZ domain , suppress mTORC1 activity in mammalian cells in response to low amino acids , indicating an evolutionarily conserved function .
We previously reported that wild type and whi2 mutant strains of Saccharomyces cerevisiae grow similarly on synthetic medium containing high amino acid levels ( SCCSH ) ( Fig 1A ) [19] . However , on medium with ~30% lower total amino acids ( SCME ) , whi2 mutants grow significantly more robustly than wild type BY4741 ( Fig 1B ) [19] . Thus , under these relatively small changes in nutrient levels compared to other studies , wild type is apparently capable of sensing the reduced amino acid levels and limiting its growth , in contrast to strains with an engineered WHI2 gene deletion ( Δwhi2 ) or a spontaneous mutation in WHI2 ( Δfis1/whi2-1 , E153X ) [1 , 19] . These findings are consistent with original studies in late-stage cultures showing that WHI2-deficiency causes a failure to enter quiescence as overall nutrient levels decline [6–9] . To determine if TORC1 , the major regulator of cellular responses to amino acid availability in yeast and mammals [25] , is responsible for this overgrowth by whi2 mutants , a low concentration of the TORC1 inhibitor rapamycin ( 2 . 5 ng/mL ) was added to the solid plate medium . Rapamycin reduced the growth of whi2 mutants to wild type levels ( Fig 1C ) , similar to the effects of reintroducing WHI2 under its native regulatory sequences on a plasmid ( without rapamycin ) ( Fig 1D ) , confirming involvement of WHI2 . Thus , WHI2 appears to be required to restrict TORC1-dependent cell growth/division when amino acid levels dwindle and before supplies are exhausted . To further investigate the role of WHI2 in suppressing TORC1 , we monitored TORC1 activity based on the phosphorylation status of endogenous 40S ribosomal subunit S6 ( Rps6 ) at Ser232/Ser233 without requiring reporters or gel-shift assays . Rps6 is phosphorylated by Ypk3 , which is believed to be a direct target of TORC1 in response to amino acids and nitrogen availability [32–34] . Antibodies directed against these same sites in mammalian S6 ( Ser235/Ser236 ) , which are also phosphorylated downstream of mammalian mTORC1 in response to amino acids [35 , 36] , readily detected phosphorylated yeast Rps6 in a TORC1-dependent manner based on inhibition by rapamycin ( Fig 1E and 1F ) . As expected , TORC1 activity assessed by Rps6 phosphorylation is significantly reduced in wild type cells by 3 h and nearly abolished at 6 h after switching from high ( SCCSH ) to low amino acids ( SCME ) ( Fig 1G ) . In contrast , the whi2 knockout has sustained TORC1 activity with only partial diminution by 6 h after media switch . Although the absolute levels of TORC1 activity can shift between independent experiments , the relationship between wild type and the whi2 knockout is highly consistent at each time point , including at baseline ( Fig 1G ) . Furthermore , the ability to suppress TORC1 was restored in Δwhi2 by constitutively expressing HA-Whi2 on a plasmid , and protein expression was verified on anti-HA immunoblots ( Fig 1G ) . Yeast TORC1 and its mammalian counterpart mTORC1 are particularly responsive to amino acid levels , but have also been reported to respond to low glucose in an Snf1/AMPK-dependent manner [37–40] . However , we found that TORC1 activity ( phospho-Rps6 ) was suppressed normally in ∆whi2 after switching from standard 2% glucose to 1% or 0 . 2% glucose ( in high amino acid SCCSH ) ( Fig 2A ) . Overall colony growth density ( colony size ) was detectably reduced on low glucose and obviously reduced without glucose supplementation , but independently of WHI2 ( Fig 2B ) . These findings are in contrast to other studies identifying WHI2 as a general stress response gene , including for low glucose conditions [10 , 41] . If WHI2 has a specific role in communicating low amino acid status , then it is expected that protein expression levels of Whi2 may be sustained or induced even in low amino acid conditions . Consistent with a role in responding to low amino acids , but not to low glucose , endogenous Whi2 protein levels ( detected with a knockin C-terminal TAP-tag [42] ) are consistently increased by 1 h after switching to low amino acids ( SCME 2% glucose ) , but decreased by 1–2 h in low glucose ( SCCSH 1% glucose ) ( Fig 2C ) . These inverse trajectories of Whi2 protein levels continued over the 7 h time course ( Fig 2D ) despite equivalent shut-off of TORC1 activity ( phospho-Rps6 ) in both low amino acids and low glucose conditions ( Fig 2C ) . Notably , induction of Whi2 protein reproducibly precedes TORC1 suppression , consistent with a causal role for Whi2 . Furthermore , the sustained phospho-Rps6 status observed in whi2 mutants grown in low amino acids is abolished by a 30 min treatment with TORC1 inhibitor rapamycin ( 20 ng/mL ) , verifying a role for TORC1 ( Fig 3A ) . These findings are consistent with the model that Whi2 responds specifically to low amino acid levels by restricting TORC1 activity . Therefore , we further investigated a potential role for Whi2 in known amino acid signaling pathways to TORC1 [24 , 43] . WHI2 was the top hit in our genome-wide screen of 4 , 847 gene deletion strains for overgrowth on low amino acids ( SCME ) [1] . This same screen also identified NPR2 and NPR3 [1] , components of the TORC1-suppressing SEACIT complex ( mammalian GATOR1 ) . Several other hits in this screen were false positives due to spontaneous WHI2 mutations [1 , 19] . However , backcrosses and tetrad analysis revealed no secondary mutations to explain the phenotypes of Δnpr2 and Δnpr3 ( S2 Fig ) . NPR2 and NPR3 were first linked to the TORC1 pathway when they were identified as the top hits in a nitrogen starvation reporter screen while in search of genes that communicate amino acid depletion to TORC1 [43] . Yeast Npr2 and Npr3 together with their catalytic subunit Iml1 form the SEACIT complex and , like their mammalian counterparts NPRL2 , NPRL3 and DEPDC5 in the GATOR1 complex [23] , negatively regulate TORC1 in response to low amino acids [24] . Therefore , we directly compared ∆npr2 and ∆npr3 to Δwhi2 for dysregulation of TORC1 activity in our mild amino acid depletion assay . Similar to Δwhi2 , both Δnpr2 and Δnpr3 strains have sustained phosphorylation of Rps6 at 3 h after switching to low amino acids ( SCME ) , which is dependent on TORC1 based on sensitivity to rapamycin ( 20 ng/mL , 30 min ) ( Fig 3A ) , consistent with the original study [43] . Sustained TORC1 activity is consistent with sustained cell growth on low amino acid plates ( SCME ) as both ∆npr2 and Δnpr3 exhibit strong overgrowth ( albeit reduced compared to Δwhi2 ) that reverts to wild type levels with rapamycin treatment ( 2 . 5 ng/mL ) ( Fig 3B ) . This low concentration of rapamycin does not cause general growth inhibition of tested strains on rich media containing high amino acids ( Fig 3B ) . Thus , Whi2 appears to be a potent suppressor of TORC1 and cell growth in response to specific signals ( Fig 3C ) . Two additional readouts for TORC1 activity were used to further confirm that Whi2 suppresses TORC1 . The transcription of DAL80 mRNA is known to be downregulated when TORC1 is active and upregulated when TORC1 is inactive [44] . Using a DAL80 promoter-driven GFP reporter plasmid to monitor the expression of DAL80 [43] , we observed significantly reduced prDAL80-GFP levels in Δwhi2 , Δnpr2 and Δnpr3 compared to wild type , indicating higher TORC1 activity in the mutants ( S3A Fig ) . We also tested the phosphorylation status of Npr1 , suggested to be a direct target of TORC1 [45] . Based on the characteristic up-shifted migration of phosphorylated Npr1 on immunoblots , Npr1 is hyperphosphorylated in Δwhi2 and partially hyperphosphorylated in Δnpr2 compared to wild type cells at time 0 . Npr1 became further shifted in Δnpr2 over 3–6 h in low amino acids , approximately co-migrating with Npr1 in Δwhi2 ( S3B Fig ) , consistent with hyperphosphorylation of Npr1 in low nitrogen in Δnpr2 and in Δnpr3 [43] . Thus , in addition to rapamycin sensitivity , three independent readouts for TORC1 indicate that WHI2 is required to suppress TORC1 activity in low amino acids . Whi2 could potentially regulate TORC1 by several different mechanisms ( Fig 3C ) . To determine if WHI2 is required for the Npr2/3-containing SEACIT complex to suppress TORC1 activity , we first asked if the catalytic subunit of the SEACIT complex , Im1 , which has GAP activity for Gtr1 [24] , can suppress the sustained TORC1 activity in Δwhi2 . Indeed , enforced expression of an enzymatically active Iml1 with C-terminal His-TAP-tags from a plasmid [24] suppressed phospho-Rps6 levels in Δwhi2 following a switch to low amino acids , although less efficiently than HA-Whi2 ( Fig 4A and 4C ) . This indicates that Whi2 is not essential for SEACIT to suppress TORC1 ( ② in Fig 3C ) , and raises the possibility that Whi2 may act in a parallel genetic pathway independently of the SEACIT complex ( ④ or ⑤ in Fig 3C ) . Although HA-Npr2 and HA-Npr3 had no effect on TORC1 activity in the absence of WHI2 , these non-catalytic subunits , in contrast to Iml1 , may be unable to enhance SEACIT function independently ( Fig 4B and 4C ) . The failure of expressed Npr2 and Npr3 to rescue whi2-deficiency is not an inherent defect of these constructs , as each could fully rescue its respective deletion strain ( Fig 4D and 4E ) . However , because we did not express all SEACIT components simultaneously , these results alone cannot exclude the possibility that Whi2 might regulate SEACIT ( ① in Fig 3C ) . To further address this point , we used the inverse approach . We found that enforced expression of HA-Whi2 suppresses overactive TORC1 ( phospho-Rps6 ) in both ∆npr2 and ∆npr3 after switching to low amino acids , but again only partially ( Fig 4D and 4E ) . This is evident because HA-Whi2 was consistently more effective than empty vector but also consistently less effective than re-expression of Npr2 or Npr3 in their respective knockouts by 3 h after switching to low amino acids ( Fig 4D and 4E , compare to empty vector in each case ) . Thus , Npr2 and Npr3 are not essential for Whi2 to affect TORC1 , indicating that Whi2 does not act upstream of the SEACIT complex ( ① in Fig 3C ) . Taking together the observed reciprocal partial rescues , these results are consistent with a model where Whi2 and the Npr2-Npr3-Iml1 SEACIT complex function in parallel pathways to negatively regulate TORC1 . To further extend these studies , we investigated the relationship between Whi2 and the downstream target of SEACIT , the RAG-like GTPases Gtr1/Gtr2 . To signal low amino acid status to TORC1 , the Npr2-Npr3-Iml1 SEACIT complex ( mammalian GATOR1 ) is known to inhibit the Gtr1 GTPase ( mammalian RAGA/B ) to suppress TORC1 activity [24] . To first determine if Whi2 is required for the RAG-like GTPases to regulate TORC1 , constitutively inhibitory Gtr1GDP ( S20L ) and constitutively active Gtr1GTP ( Q65L ) [46] mutants were expressed via plasmids in yeast lacking the WHI2 gene . Both the inhibitory and activated forms of Gtr1 were still capable of modulating TORC1 activity ( phospho-Rps6 ) even in the absence of Whi2 through 6 h after switching to low amino acids ( Fig 5A and 5B ) . That is , Gtr1GDP ( S20L ) dramatically decreases TORC1 activity by 3 h in low amino acids , while Gtr1GTP ( Q65L ) still maintains a high TORC1 activity at 6 h compared with wild type Gtr1 ( Fig 5B ) . Thus , Whi2 does not appear to inhibit TORC1 downstream of the Gtr1-Gtr2 complex as Whi2 is not essential for Gtr1 to modulate TORC1 activity ( ③ in Fig 3C ) . Conversely , Whi2 does not require Gtr1 to regulate TORC1 . Endogenous Whi2 suppresses TORC1 in cells overexpressing Gtr1 or its constitutively active Gtr1GTP ( Q65L ) mutant based on higher TORC1 activity in whi2 knockouts relative to wild type cells at 3 and 6 h after switching to low amino acids ( Fig 5C and 5D ) . Consistent with this finding , co-expression of HA-Whi2 with Gtr1GTP ( Q65L ) suppresses the overzealous TORC1 activity ( phospho-Rps6 ) induced by constitutively active Gtr1GTP ( Q65L ) and modestly suppresses the effects of expressed wild type Gtr1 ( Fig 5E and 5F ) . Thus , because Whi2 can inhibit the effects of constitutively active GTP-bound Gtr1 , any inhibitory effect of Whi2 on Gtr1 would presumably be by a mechanism different from the GAP activity of SEACIT ( ④ in Fig 3C ) , or alternatively , Whi2 could act in a pathway parallel to the Gtr complex ( ⑤ in Fig 3C ) . In the latter case , Whi2 could co-regulate TORC1 in conjunction with the Gtr1-Gtr2 GTPase complex . To further distinguish a parallel Whi2 path ( ⑤ in Fig 3C ) from a Gtr1-Gtr2-dependent path ( ④ in Fig 3C ) , we tested if Whi2 can act independently of these TORC1-activating RAG-like GTPases . We found that expression of HA-Whi2 potently suppresses TORC1 activity in both ∆gtr1 and ∆gtr2 ( Fig 5G and 5H ) . This indicates that Whi2 does not require the Gtr complex to suppress TORC1 activity , and therefore is not upstream of the TORC1-activating RAG-like GTPases ( ④ in Fig 3C ) . Thus , Whi2 appears to function in a parallel independent path to communicate low amino acids signals to TORC1 ( ⑤ in Fig 3C ) . In the event that Whi2 acts to suppress TORC1 in a pathway parallel to the GTPases , then deletion of both the TORC1 activator Gtr1 together with the TORC1 suppressor Whi2 would be expected to yield a neutralizing phenotype if both pathways are active simultaneously . Indeed , deletion of WHI2 in ∆gtr1 reduces but does not abolish TORC1 activity ( phospho-Rps6 ) after switching to low amino acid conditions ( Fig 5I ) . Importantly , the double knockout ∆whi2∆gtr1 also exhibits an intermediate growth phenotype , consistent with the observed TORC1 activity ( Fig 5J ) . Thus , Whi2 appears to reflect a novel alternative amino acid sensing pathway that signals amino acid insufficiency to TORC1 ( ⑤ in Fig 3C ) . Whi2 is reported to interact with the protein phosphatase Psr1 using recombinant proteins [10] and to physically interact with both Psr1 and Psr2 in several high-throughput screens [47–50] . We thus tested if Psr1 and its partially functionally redundant paralog Psr2 is involved in regulating TORC1 activity . Although single gene deletions of PSR1 or PSR2 had little or no overgrowth on low amino acids , the Δpsr1Δpsr2 double knockout overgrew similarly to Δwhi2 ( Fig 6A ) . Consistent with this finding , Δpsr1Δpsr2 also has sustained TORC1 activity after switching to low amino acid media , highly similar to Δwhi2 tested in parallel ( Fig 6B ) . These results suggest that Psr1 and Psr2 are required to suppress TORC1 activity under low amino acid conditions . To determine if Psr1/Psr2 are also required for Whi2 to suppress TORC1 , we overexpressed Whi2 in Δpsr1Δpsr2 . Although expression of either Psr1 or Psr2 was sufficient to suppress overgrowth and TORC1 activity in Δpsr1Δpsr2 under low amino acid conditions , expressed Whi2 had no effect ( Fig 6C and 6D ) , indicating that Whi2 requires Psr1 or Psr2 to suppress TORC1 activity . Conversely , we found that the overexpressed enzymes Psr1 or Psr2 only weakly suppressed the overgrowth and the TORC1 activity of Δwhi2 ( Fig 6E ) . Importantly , expressed Psr1 reduced TORC1 activity in wild type cells much more efficiently than Δwhi2 under low amino acid conditions , indicating that Psr1 also relies on Whi2 to fully suppress TORC1 ( Fig 6F and 6G ) . Together , these reciprocal findings are consistent with Whi2 acting in a complex with Psr1 to suppress TORC1 activity . Supporting this model , we found that Whi2 and Psr1 can be co-immunoprecipitated ( Fig 6H ) , consistent with the earlier study [10] . WHI2-deficiency is reported to cause inappropriate Ras-cAMP-PKA pathway activation [22] . However , when PKA activity was suppressed in Δwhi2 by further deleting TPK3 , which encodes a catalytic subunit of PKA [51] , the Δwhi2Δtpk3 double knockout was indistinguishable from Δwhi2 for both overgrowth and TORC1 activity under low amino acid conditions ( S4A and S4B Fig ) . As an alternative strategy , we overexpressed the high affinity cAMP phosphodiesterase Pde2 , which inhibits PKA activity by hydrolyzing cAMP to AMP [51] . However , overexpressed Pde2 had no effect on the elevated growth or the TORC1 overactivity of Δwhi2 ( S4C and S4D Fig ) . Thus , at least under our low amino acid conditions , Whi2 suppresses TORC1 activity independently of the PKA pathway . Although originally thought to be fungi-specific , Whi2 shares amino acid sequence similarity with a family of poorly characterized human proteins designated potassium channel tetramerization domain proteins ( KCTDs ) [1] . However , like fungal Whi2 proteins , human KCTD family proteins lack predicted channel domains and have very divergent C-termini , but share an N-terminal BTB domain subtype present in fungal Whi2 . Among the 25 human KCTD family members [52] , several have been linked to human cancers , including KCTD8 [53] , KCTD12 [54] , TNFAIP1 [3] and most notably KCTD11/Ren/KCASH1 , a reported tumor suppressor in medulloblastoma , possibly by suppressing Hedgehog signaling [2 , 55] . To determine if mammalian KCTD proteins are also involved in regulating TORC1 , we first tested their ability to suppress TORC1 activity in yeast , despite their limited overall sequence similarity with Whi2 . N-terminal HA-tagged mammalian KCTD proteins KCTD7 , KCTD8 and the extended version of KCTD11 containing the complete BTB domain [2] were expressed in yeast Δwhi2 . Only KCTD11 suppressed TORC1 activity ( phospho-Rps6 ) at baseline and further suppressed TORC1 activity in low amino acids similarly to HA-Whi2 , while KCTD7 and KCTD8 had no effect or slightly increased TORC1 activity ( Fig 7A ) . Human KCTD11 also mimicked yeast Whi2 by increasing expression of prDAL80-GFP , an alternative reporter for TORC1 inhibition [43] ( S5A Fig ) , and also can suppress the overgrowth phenotype of Δwhi2 on low amino acid plates similar to Whi2 ( S5B Fig ) . Strikingly , expressed human KCTD11 can rescue the actin aggregation phenotype of Δwhi2 , resulting in an actin-staining pattern similar to re-expressed Whi2 as previously reported [22] ( S5C and S5D Fig ) . However , the specificity of KCTD11 in yeast is challenging to verify , therefore we tested KCTD11 in mammalian cells . In mammalian cells ( COS7 ) , expression of KCTD7 , KCTD8 , KCTD12 and KCTD16 failed to suppress mTORC1 assessed by the phosphorylation status of mammalian S6 kinase ( S6K ) at T389 , a direct target of mTORC1 [27] , in amino acid-free medium ( lacking serum , which contains amino acids ) and after readdition of amino acids ( Fig 7B ) . Again , only KCTD11 could suppress mTORC1 activity in mammalian cells after 50 min in amino acid-free medium and after readdition of amino acids ( in the absence of serum ) compared to control ( Fig 7C ) . Importantly , endogenous KCTD11 also appears to negatively regulate mammalian TORC1 as transiently transfected shRNA2 , and less effective shRNA1 , consistently enhanced TORC1 activity based on the phosphorylation status of mTORC1 target S6K ( Fig 7D upper ) , and its downstream target S6 ( Fig 7D lower ) . Knockdown by shRNA2 , in contrast to shRNA1 , was confirmed in cells co-expressing human HA-KCTD11 ( Fig 7E ) . These findings indicate that KCTD11 and yeast Whi2 may share an evolutionarily conserved role in suppressing TORC1 activity in response to reduced amino acid levels .
We found that WHI2 is required to suppress TORC1 activity under low amino acid conditions . Because the GATOR1/SEACIT-Gtr1/2 pathway is the major regulator of TORC1 in response to both high and low amino acid levels [24 , 26 , 56] , several strategies were used to probe the potential role for Whi2 in this pathway . Instead , we uncovered a parallel mechanism mediated by Whi2 that transmits low amino acid status to inhibit TORC1 activity independently of the SEACIT–Gtr GTPase axis ( Figs 4 and 5 ) . Thus , it appears that Whi2 and SEACIT-Gtr pathways are independent of each other , but when acting in concert , together they robustly suppress TORC1 activity ( Fig 5 ) . Thus , each could theoretically communicate non-overlapping information about nutrient status to TORC1 . Yeast TORC1 can respond to specific amino acids including leucine [56] , glutamine[57] and methionine [58] , and mammalian mTORC1 was reported to respond to leucine [35] , glutamine [59] and arginine [60] . The specific amino acids responsible for eliciting Whi2-dependent responses are not yet known , but leucine is the major difference between the two media used in this study [19] . The glucose-responsive Ras-cAMP-PKA pathway [61] appears not to be involved in Whi2-mediated suppression of TORC1 ( S4 Fig ) , and Whi2 had no detectable role in suppressing TORC1 activity under low glucose conditions , at least in the presence of amino acids ( Fig 3 ) . This is in contrast to current opinion , although previous studies on low glucose conditions were performed in conjunction with nitrogen deprivation [9 , 62] . The mechanism for Whi2-mediated TORC1 inhibition is not known , but requires the under-characterized phosphatases Psr1/Psr2 ( Fig 6 ) . Psr1 together with its binding partner Psr2 have been reported to interact with Whi2 in pull-down assays and/or high throughput screens [10 , 47–50] , confirmed for Psr1 by our co-IP with Whi2 ( Fig 6H ) . Whi2 is suggested to activate the Psr1/Psr2 phosphatases , which are localized to the plasma membrane and have been implicated in mounting general stress responses by activating Msn2 [10] . Msn2 is reported to be suppressed when TORC1 is active [63] , which is consistent with our model that Whi2 is an upstream negative regulator of TORC1 . How or where Whi2-Psr1-Psr2 might regulate TORC1 directly or indirectly is not known . How could the effect of Whi2–Psr1/Psr2 at the plasma membrane potentially connect to TORC1 ? One potential connection between Whi2 and TORC1 via Psr1/Psr2 is by controlling the activity of plasma membrane transporters . The ammonium transporter Mep2 was reported to be inactivated by Psr1/Psr2 [11] . In this model , Psr1/Psr2 removes the activating phosphorylation on Mep2 installed by Npr1 kinase when nutrients are low and Npr1 is not inactivated by TORC1 [11] . Consistent with their findings , an alternative possibility is that Whi2-Psr1/Psr2 acts upstream of TORC1 to suppress nutrient uptake . Another potential mechanism connects plasma membrane amino acid transporters to TORC1 . The general amino acid permease Gap1 and the arginine-specific transporter Can1 appear to function as a type of amino acid sensor by fluxing protons during amino acid uptake [64] . This proton flux triggers compensatory export of protons by the plasma membrane H+-ATPase Pma1 . By an unknown proton-activated Pma1-dependent signaling mechanism , Pma1 activates TORC1[64] . However , any role for Whi2 in this pathway is not known and unlike Whi2 , the ability of Pma1 to regulate TORC1 is at least partially dependent on the GATOR1/SEACIT-Gtr pathway [64] . Furthermore , the proton-Pma1 pathway to activate TORC1 was postulated only to respond to nutrient availability and not to be involved in the suppression of TORC1 activity in low nutrients [64] . The requirement for Whi2 to suppress TORC1 in low amino acid conditions implies that Δwhi2 cells will be defective for autophagy induction due to sustained TORC1 activity , consistent with a recent study [5] . Mendl et al . reported that Whi2 is required for the degradation of mitochondria via autophagy ( mitophagy ) induced by high concentrations of rapamycin ( 1 μM ) for 7 . 5–24 hours [65] . However , a conflicting report from Klionsky and colleagues concluded that Whi2 is not required for mitophagy when stationary cells grown in lactate are switched to nitrogen starvation medium plus 2% glucose to induce mitophagy [66] . Whi2-dependent amino acid sensing could be important for fungal pathogenesis . Down regulation of TORC1 in pathogenic strains of yeast such as Candida albicans and Candida glabrata is suggested to play a role in the persister state [67] . Consistent with our finding that Whi2 suppresses TORC1 , the Whi2 ortholog in the fungal pathogen Colletotrichum orbiculare was recently suggested to have a role in plant pathogenesis by inhibiting TOR signaling [68] . The evidence presented for Whi2 being a negative regulator of TORC1 is supported by several different readouts for TORC1 activity . We detected TORC1 activity using a specific antibody for highly conserved phosphorylation sites in the ribosomal S6 proteins of both yeast ( Rsp6 Ser232/Ser233 ) and mammals ( Ser235/Ser236 ) , which are known to be phosphorylated downstream of mTORC1 in response to amino acids in mammals [35 , 36] . This antibody permits quantification of yeast TORC1 activity by monitoring an endogenous substrate without reporters or gel-shift assays . TORC1-dependent phosphorylation of yeast Rps6 is further supported by rapid dephosphorylation at Ser232/233 upon rapamycin treatment ( Fig 3 ) . Rps6 phosphorylation is regulated by RAG-like GTPases Gtr1-Gtr2 in wild type cells ( Fig 5 ) , verifying the utility of this strategy . Two alternative TORC1 activity assays ( DAL80 expression and Npr1 phosphorylation ) , further confirmed that Whi2 is required to suppress TORC1 activity under low amino acid conditions ( S3 Fig ) . It is puzzling that WHI2 was not identified in several screens for regulators of low amino acid-sensing in the TORC1 pathway [43 , 69 , 70] . Possibly the death-sensitivity of Δwhi2 to multiple stimuli [1 , 10 , 19] results in loss of whi2 mutants during screening under more harsh conditions . In contrast to other reports [9 , 41] , we found that yeast Whi2 is not required to suppress TORC1 activity in response to low glucose in the presence of amino acids ( Fig 2 ) . WHI2 was identified in one other screen for suppressors of TORC1 [41] . However it was concluded that the apparent suppressive effects of WHI2 on TORC1 activity following glucose or nitrogen depletion was more likely due to indirect consequences of impaired cell cycle arrest by whi2 mutants in a chemostat environment [41] . Conversely , we and others observed spontaneous whi2 mutations in specific knockout strains that appear to arise in part due to the loss of specific knockout gene functions [1 , 19 , 71] . Perhaps the selection for whi2 mutations is related to the role of Whi2 in regulating TORC1 as amino acids become depleted during normal culturing . In further support of a specific role for Whi2 in sensing low amino acids , the protein level of Whi2 rapidly increases upon shifting to low amino acid conditions , followed by suppression of TORC1 activity ( Fig 2D ) . In contrast , glucose limitation has the opposite effect , reducing Whi2 protein levels . Interestingly , we also observed that Iml1 ( component of SEACIT ) levels increase following switch to low amino acids ( Fig 4A ) . The fact that induction or stabilization of Whi2 and Iml1 protein levels is an early event in response to low amino acids suggests that Whi2 and Iml1 are both effectors of an upstream signal . Given the high prevalence of secondary mutations in knockout strains [1] , we performed tetrad analysis on the Δwhi2 , Δnpr2 , Δnpr3 , Δgtr1 and Δgtr2 strains used in this study , which verified that the phenotypes described here are appropriately attributed . However , the Δgtr2 strain in the BY4741 YKO collection is a mix of two prominent phenotypes with dramatically different sensitivities to stress [1] . Although the yet unidentified single secondary gene mutation responsible for stress sensitivity of Δgtr2 did not affect amino acid sensing [1] , a substrain lacking this secondary mutation was used for these studies . No secondary mutations that contribute to the growth phenotypes were identified in multiple colonies tested of Δwhi2 , Δnpr2 , Δnpr3 , Δgtr1 and Δgtr2 strains used in this study . In contrast , the tetrads generated from knockouts strains for Ego1 and Ego3 , which anchor Gtr1-Gtr2 to membranes [56] , exhibited complex genetics/phenotypes and therefore were not further studied . It remains possible that the SEACIT-Gtr and the Whi2 pathways converge near or within the TORC1 protein complex . Other Gtr/RAG-independent mechanisms involving different GTPases have been reported in both yeast and mammals to activate TORC1 in response to amino acid abundance . The mammalian Rab family GTPase Rab1A and its yeast homolog Ypt1 , an essential gene , appear to activate TORC1 independently of RAG/Gtr GTPases [72] . However , it is not known if yet unidentified negative regulators of the Ypt1/RAB1A GTPases might transmit low amino acid status to suppress TORC1 , analogous to Whi2 or SEACIT/GATOR1 . Another RAG GTPase-independent mechanism in Drosophila and mammalian cells requires the ARF1 GTPase for acute TORC1 activation induced by glutamine [59] . This ARF1-dependent mechanism has not been demonstrated in yeast . To the contrary , we identified Δarf1 in the same screen that identified Δnpr2 and Δnpr 3 [1] , implying that yeast Arf1 may be a negative rather than positive regulator of TORC1 . However , it is not known if this phenotype is due to deletion of ARF1 or to a secondary mutation . Yeast Arf1 also activates the Ras-cAMP-PKA pathway particularly in response to glucose [61] . However , our findings indicate that whi2-deficient cells respond normally by suppressing TORC1 activity and cell growth in glucose-free conditions , and that the sustained TORC1 activity and cell growth are not dependent on the PKA pathway ( S4 Fig ) . These Gtr/RAG-independent amino acid-sensing paths involving Ypt1/RAB1A and ARF1 could potentially be connected . Unlike the RAG/Gtr complexes , which localize to the lysosome/vacuole membrane [73] , RAB1A and ARF1 are both predominantly localized on Golgi membranes , supporting a model for TORC1 to sense amino acids in different subcellular regions [74] . Whi2 physically interacts with the phosphatase Psr1/2 , which localizes to the plasma membrane [10] , and endogenous Whi2 fluorescently tagged ( either N- or C-terminus ) has also been shown to localize to the cell periphery [75] . Our study shows that Whi2 inhibits TORC1 activity through Psr1 and Psr2 , raising the possibility of Whi2 extending amino acid signaling to additional subcellular sites . Importantly , Whi2 contributes to the suppression of TORC1 activity that occurs in the absence of Gtr1 ( Fig 5G–5I ) . We provide the first evidence that an amino acid-sensing function of yeast Whi2 is conserved in a mammalian KCTD family protein , KCTD11 . The evolutionary histories of Whi2 and KCTD family proteins have not yet been studied in detail . At this stage therefore , no reliable homology can be inferred beyond the SKP1/BTB/POZ domain shared between yeast and human protein types ( note that subsets of the human KCTD family also have sequence similarities concentrated only in their SKP1/BTB/POZ domain ) . Yet , it is remarkable that KCTD11 can functionally substitute for yeast Whi2 to suppress TORC1 activity and cell growth under low amino acid conditions , and to rescue the actin filament aggregation phenotype of Δwhi2 . Importantly , both expressed and endogenous KCTD11 also negatively regulates mTORC1 activity under amino acid depletion conditions in mammalian cells , confirming the conserved function of TORC1 suppression . It was somewhat unexpected that KCTD7 did not affect mTORC1 activity given that fibroblasts from EPM3 patients with bi-allelic KCTD7 mutations have defective autophagy responses [5] . Biallelic KCTD7 mutations define the diagnosis of EPM3 ( progressive myoclonic epilepsy-3 ) , a severe neurodegenerative disorder with onset in early childhood [4 , 76] . It remains possible that KCTD7 and other family members modulate this pathway in other conditions or cell types . However , KCTD family proteins have received little attention and remain poorly characterized despite their disease associations . KCTD11 was reported to suppress the Hedgehog signaling pathway in medulloblastoma [2] , and crosstalk between the Hedgehog and PI3K/AKT/mTORC1 pathways via Gli1 activation has been reported to occur in several types of cancer models including medulloblastoma , prostate cancer and breast cancer cell lines [77] . KCTD11 also has been implicated as a tumor suppressor in several other cancers including prostate adenocarcinoma [18] , and hepatocellular carcinoma [17] , although its role in cancer remains to be confirmed . In summary , we investigated the possibility that Whi2 is a new upstream negative regulator of TORC1 . Indeed , whi2-deletion strains have sustained TORC1 activity ( phospho-Rps6 ) following a switch to medium with lower amino acids . Both the overgrowth and sustained phospho-Rps6 levels are dependent on TORC1 as both are blocked by low concentrations of rapamycin . However , we found that Whi2 suppresses TORC1 activity independently of the RAG-like Gtr complex conserved in yeast , but through protein phosphatases Psr1 and Psr2 , implying a novel mechanism . Despite amino acid sequence divergence from yeast Whi2 , the human protein KCTD11 , but not other KCTD family members tested under our conditions , suppress TORC1 activity in yeast . More importantly , endogenous KCTD11 suppresses mTORC1 activity in mammalian cells as KCTD11 depletion leads to higher mTORC1 activity . Thus , it is conceivable that failure to suppress mTORC1 in tumors lacking functional KCTD11 may be an important contributor to its proposed role in tumorigenesis [2 , 17 , 18] .
The Saccharomyces cerevisiae strains and the yeast and mammalian expression plasmids used in this study are listed for each figure ( Tables 1 and 2 , respectively ) . Yeast cultures were grown for 48 h in liquid YPD ( 2% peptone , 1% yeast extract , and 2% dextrose ) or SCCSH [19] ( 0 . 67% yeast nitrogen base w/o amino acids , 0 . 2% CSH amino acid mix , 2% glucose ) minus uracil or/and histidine for strains transformed with URA3 or/and HIS3 plasmids . Saturated cultures were washed and serially diluted fivefold in sterile ddH2O , and 5 μL of each dilution were spotted onto solid SCCSH and SCME [19] ( 0 . 67% yeast nitrogen base w/o amino acids , 0 . 124% ME amino acid mix , 2% glucose ) with/without rapamycin , or SCCSH containing indicated concentration of glucose , and incubated at 30 oC for two to three days . Yeast strains were grown overnight in liquid SCCSH , and refed for 1 h in fresh SCCSH medium at 1 OD/mL to allow diauxic phase cells to recover from nutrient deprivation overnight . Yeast cultures were washed once and equal cell numbers were resuspended in low amino acid medium SCME . Lysates were prepared as reported [20] , separated on 12% SDS-PAGE gels and analyzed on immunoblots with antibodies against phosphorylated Rps6 ( mammalian phospho-S235/236 S6 antibody , Cell Signaling Technology , 1:1000 ) , yeast Pgk ( Abcam , 1:1000 ) , HA-epitope and β-actin ( Santa Cruz Biotechnology , 1:1000 ) , followed by HRP-conjugated anti-rabbit and anti-mouse secondary antibodies ( GE Healthcare , 1:20 , 000 ) . TORC1 activity was quantified as a ratio of the intensity of phosphorylated Rps6 relative to loading control for each sample in ImageJ . Yeast strains transformed with prDAL80-GFP plasmid were cultured and lysed as described in the above section . Cell lysates were resolved on 12% SDS-PAGE gels and analyzed on immunoblots with antibody against GFP ( Santa Cruz Biotechnology , 1:1000 ) and yeast Pgk ( Abcam , 1:1000 ) , followed by HRP-conjugated anti-rabbit and anti-mouse secondary antibodies ( GE Healthcare , 1:20 , 000 ) . TORC1 activity was quantified as a ratio of the intensity of GFP relative to loading control for each sample in ImageJ . Yeast strains transformed with HA-NPR1 plasmid were cultured and lysed as described in the above section . Cell lysates were resolved on 7 . 5% SDS-PAGE gels and analyzed on immunoblots with antibody against HA-epitope ( Santa Cruz Biotechnology , 1:1000 ) and yeast Pgk ( Abcam , 1:1000 ) , followed by HRP-conjugated anti-rabbit and anti-mouse secondary antibodies ( GE Healthcare , 1:20 , 000 ) . TORC1 activity was measured by the shift of the HA-Npr1 band . Yeast strains were grown in liquid SCCSH for 24 h before staining F-actin with Rhodamine-phalloidin as reported [82] . Cells were viewed with a Nikon TE-2000 fluorescence microscope . Human KCTD11 cDNA consisting of the KCTD11 ORF ( NM_001002914 . 2 ) plus 120 5’ in-frame nucleotides required to complete the conserved BTB domain [83] were PCR amplified from HEK293 cells ( 5’ primer GGGAGATCGAAGATCTAAAATTTCTCCTCCTCCTGTGCCCTCTTCG , 3’ primer CTTGGTGACCAGATCTTCAGTGCCGGACAAAGCGCAGAGAC ) , and subcloned into the BglII site of pSG5-based vector pDB59 containing a Kozak sequence and N-terminal HA-tag and verified by Sanger sequencing . cDNA of human KCTD7 ( NM_153033 . 4 ) , KCTD12 ( NM_138444 . 3 ) , and KCTD16 ( NM_020768 . 3 ) were cloned similarly into the BglII site of pDB59 vector . Primers used to amplify each gene are listed below . KCTD7: 5’ primer AGCTAGATCTATGGTGGTAGTCACGGGG , 3’ primer AGCTAGATCTTCACCACCATGTGATCTTGAA; KCTD12: 5’ primer AGCTAAGCTTGGATGGCTCTGGCGGACAG , 3’ primer AGCTAAGCTTTCACTCCCTGCAGAAGACG; KCTD16: 5’ primer AGCTAAGCTTATGGCTCTGAGTGGAAACTGTAG , 3’ primer AGCTAAGCTTTTATAGATGATACTTCCTTAAAAGTTCAGATTGCCAA . Expressed KCTD family proteins were tested in monkey COS-7 cells ( from ATCC ) cultured in DMEM containing 10% fetal bovine serum ( FBS ) and Penicillin/Streptomycin antibiotics ( Hyclone ) , and endogenous KCTD11 was evaluated in human embryonic kidney 293 cells ( from ATCC ) cultured in DMEM plus 10% FBS and P/S . Approximately 2x104 cells were seeded into 12-well plates with 1 mL media , and transfected 24 h later with 0 . 75 μg DNA for 24 h ( protein expression plasmids ) or for 48 h ( shRNA plasmids ) using Lipofectamine 2000 according to manufacturer’s instructions . Plasmids containing shRNAs targeting KCTD11 from GeneCopia ( [psi-LVRH1GP] prH1-shRNA-EGFP ) have the following target sequences: All transfected cells were rinsed once with amino acid-free RPMI ( US Biological ) , incubated in 1 mL amino acid-free RPMI for 50 min , and stimulated by adding 20 μL of a 50× amino acid mixture ( RPMI 1640 Amino Acids Solution , Sigma ) for 10 min . Cell lysate were prepared in lysis buffer ( 62 . 5 mM Tris-HCl , pH = 6 . 8 , 2% w/v SDS , 10% glycerol , 0 . 01% w/v bromophenol blue ) , separated on 12% gels by PAGE and analyzed on immunoblots using antibodies against phospho-S235/235 S6 , S6 , phospho-T389 S6K , S6K , ( Cell Signaling Technology ) at 1:1000 dilution , followed by HRP-conjugated anti-rabbit secondary antibodies ( GE Healthcare , 1:20 , 000 ) . To test the efficiency of KCTD11 shRNA knockdowns , HEK293 cells were co-transfected with HA-KCTD11 at a 1:9 ratio of hHA-KCTD11:shRNA ( 0 . 75 μg total DNA per well ) and harvested for immunoblot analysis after 24 h . mTORC1 activity was quantified as a ratio of the intensity of phosphorylated S6K/S6 relative to total S6K/S6 for each sample in ImageJ . The values were normalized to the average value of the control .
|
Yeast and human cells respond to declining levels of available nutrients to prepare ahead for leaner times . The detailed mechanisms of nutrient sensing are not well understood , but defects in these processes have key roles in diseases such as cancer . The evolutionarily conserved protein complex TORC1 is the control hub for responding to both high and low nutrients , particularly amino acids . We identified yeast Whi2 and the human tumor suppressor KCTD11 as novel suppressors of TORC1 activity in low amino acid conditions , and we investigated the detailed mechanisms for Whi2 . Unexpectedly , Whi2 works differently from the usual mechanism where TORC1 is controlled by the SEACIT-Gtr complex ( mammalian GATOR1-RAG complex ) . Furthermore , both the Whi2 and the SEACIT-Gtr pathways work independently and together in parallel to suppress TORC1 . For this function , Whi2 requires its binding partners , the yeast protein phosphatases Psr1 and Psr2 , which were previously thought to function downstream of TORC1 in amino acid signaling . These studies have important implications for human KCTD11 to help advance the understanding of its pathological role .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"phosphorylation",
"chemical",
"compounds",
"molecular",
"probe",
"techniques",
"enzymes",
"enzymology",
"immunoblotting",
"carbohydrates",
"organic",
"compounds",
"glucose",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"amino",
"acid",
"analysis",
"saccharomyces",
"immunoblot",
"analysis",
"proteins",
"chemistry",
"guanosine",
"triphosphatase",
"molecular",
"biology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"yeast",
"biochemistry",
"eukaryota",
"hydrolases",
"organic",
"chemistry",
"post-translational",
"modification",
"monosaccharides",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"physical",
"sciences",
"saccharomyces",
"cerevisiae",
"organisms"
] |
2018
|
Whi2 is a conserved negative regulator of TORC1 in response to low amino acids
|
China has the highest incidence of hemorrhagic fever with renal syndrome ( HFRS ) worldwide . Reported cases account for 90% of the total number of global cases . By 2010 , approximately 1 . 4 million HFRS cases had been reported in China . This study aimed to explore the effect of the rodent reservoir , and natural and socioeconomic variables , on the transmission pattern of HFRS . Data on monthly HFRS cases were collected from 2006 to 2010 . Dynamic rodent monitoring data , normalized difference vegetation index ( NDVI ) data , climate data , and socioeconomic data were also obtained . Principal component analysis was performed , and the time-lag relationships between the extracted principal components and HFRS cases were analyzed . Polynomial distributed lag ( PDL ) models were used to fit and forecast HFRS transmission . Four principal components were extracted . Component 1 ( F1 ) represented rodent density , the NDVI , and monthly average temperature . Component 2 ( F2 ) represented monthly average rainfall and monthly average relative humidity . Component 3 ( F3 ) represented rodent density and monthly average relative humidity . The last component ( F4 ) represented gross domestic product and the urbanization rate . F2 , F3 , and F4 were significantly correlated , with the monthly HFRS incidence with lags of 4 months ( r = −0 . 289 , P<0 . 05 ) , 5 months ( r = −0 . 523 , P<0 . 001 ) , and 0 months ( r = −0 . 376 , P<0 . 01 ) , respectively . F1 was correlated with the monthly HFRS incidence , with a lag of 4 months ( r = 0 . 179 , P = 0 . 192 ) . Multivariate PDL modeling revealed that the four principal components were significantly associated with the transmission of HFRS . The monthly trend in HFRS cases was significantly associated with the local rodent reservoir , climatic factors , the NDVI , and socioeconomic conditions present during the previous months . The findings of this study may facilitate the development of early warning systems for the control and prevention of HFRS and similar diseases .
Hemorrhagic fever with renal syndrome ( HFRS ) is a natural focal disease characterized by fever , hemorrhagic manifestations , and acute renal dysfunction . HFRS is mainly transmitted by rodents [1] . In China , HFRS is primarily caused by one of two types of hantaviruses , Hantaan virus ( HTNV ) and Seoul virus ( SEOV ) [2] . China has the highest incidence of HFRS worldwide . Reported cases account for 90% of the total number of global cases . Approximately 1 . 4 million HFRS cases were reported in China between 1950 and 2010 [3] . HFRS incidence has decreased in China in recent years . However , HFRS still causes significant morbidity and mortality and is a serious public health threat [4] . Hunan Province is one of the most highly endemic areas in China; 2670 cases of HFRS were diagnosed from 2006 to 2010 . Chenzhou , a subtropical city in Hunan Province , is noted for epidemics of HFRS . During a 5-year period ( 2006 to 2010 ) , 321 cases were reported in Chenzhou . HFRS is widely transmitted from rodents to humans through contact with saliva , urine or excreta from infected rodents [5] . It is closely associated with rodent density and the virus-carrying rates of the animal hosts [6]–[9] . These factors are characterized by observable seasonal and regional variation . Variations in HFRS incidence is associated with the growth and decline of rodent population density . Human hantavirus epidemics can be accurately predicted from an analysis of the population dynamics of the rodent hosts [9]–[11] . HFRS incidence is also affected by natural environmental factors such as land use , elevation , vegetation type , crop production and area , and the El Niño-Southern Oscillation ( ENSO ) [12]–[15] . In particular , HFRS incidence closely correlates with meteorological factors that include temperature , rainfall , and humidity [1] . Past evidence has demonstrated that outbreaks of diseases such as schistosomiasis , malaria , tuberculosis and plague are affected by environmental factors ( e . g . , geography , climate , and zoology ) , and are affected and restricted by socioeconomic factors ( e . g . , social institution , economic status , and population mobility ) [16] . For example , social factors , such as population mobility , have an important influence on the transmission of infectious disease . Economic development and land transformation by human activities also affect infectious disease prevalence [17] . In the midstream and downstream of China's Yangtze River , schistosomiasis incidence decreased with the development of local economic conditions . This change could be attributed to increased migration from villages to cities , resulting in a reduction in exposure to cercariae in village rice paddies [18] . The distribution and population of rodent hosts and vectors are directly affected by urbanization , deforestation , irrigation works and road construction . These changes lead to high densities of rodent populations in some areas and outbreaks of some diseases ( e . g . , plague and Lyme disease ) [19] . Few studies have considered the relationship between HFRS and socioeconomic factors and the effect of natural and socioeconomic factors on HFRS incidence . The aim of this study was to analyze the quantitative relationship between HFRS transmission and environmental variables , to forecast the trend in prevalence of HFRS transmission , and to reveal the transmission pattern from data on HFRS cases , rodent host populations , and environmental variables ( natural variables and social variables ) in Chenzhou from 2006 to 2010 . Natural variables included rodent density , the normalized difference vegetation index ( NDVI ) for cultivated land , monthly average temperature , monthly average rainfall and monthly average relative humidity . Social variables included gross domestic product ( GDP ) and the urbanization rate . The results may lead to the discovery of epidemic factors that are important for control of HFRS .
The study area covers Chenzhou , located in a subtropical region of Hunan Province in Central China . Chenzhou is located between latitude 24°53′ and 26°50′ north , and longitude 112°13′ and 114°14′ east . It is 217 km wide and 202 km long , with a total land area of 19 , 400 km2 . The region consists of two municipal districts ( Beihu and Suxian ) and nine counties ( Guiyang , Yizhang , Yongxing , Jiahe , Linwu , Rucheng , Guidong , Anren and Zixing ) , and a total population of about 4 . 6 million people ( Figure 1 ) . From 2006 to 2010 , data from cases of HFRS in Chenzhou were obtained from the Hunan Center for Disease Control and Prevention ( CDC ) . All of the cases were initially diagnosed based on clinical symptoms using diagnostic criteria from the Ministry of Health of the People's Republic of China . Blood samples were collected for serologic identification from all suspect cases . Samples were analyzed at the Hunan CDC laboratory . Detailed procedures can be found in published articles [20] . Surveillance of hantavirus infections among rodent hosts from 2006 to 2010 was conducted once per month for three consecutive nights . At least 300 medium-sized steel traps were set each night ( baited with peanuts ) and were recovered in the morning . More than 100 of these traps were placed indoors at approximately 12- to 15-meter intervals , and more than 200 traps were placed outdoors ( every 5 meters in each row , with 50 meters between rows ) . A total of 698 rodents were captured out of 36 , 243 effective traps . “Relative rodent density” , used as an indicator of abundance , was calculated as the number of rodents captured , divided by the number of traps ( Table 1 ) . Meteorological data ( monthly average temperature , monthly average relative humidity , and monthly precipitation ) for the 2006 to 2010 period were obtained from the China Meteorological Data Sharing Service System ( http://cdc . cma . gov . cn/index . jsp ) . GDP and the urbanization rate of the population were obtained from the Hunan Statistics Yearbook . Immigrant population data were obtained from Hunan Public Security Department . Data for the estimation of the NDVI was obtained from the International Scientific Data Service Platform ( 1 km spatial resolution; http://datamirror . csdb . cn/ ) . The NDVI was generated from a transformation of the near infrared ( NIR ) and red wavelengths ( RED ) using the equation: ( 1 ) Land use data were obtained from the Second National Land Survey and were categorized as to cultivated land , forest , grass , or residential land . The data set was analyzed in ArcGIS 9 . 3 ( ESRI Inc . , Redlands , CA , USA ) and included a digital map of Chenzhou ( 1∶50000 ) , geocoding , case information , and population information . The data for each variable were converted to the same geographic projection and clipped to the study area . The present study was reviewed by the research institutional review board of the Hunan CDC . The review board determined that utilization of disease surveillance data did not require oversight by an ethics committee . Because the data were publicly available secondary data and were analyzed anonymously , no ethics statement was required for the work . The methods did not include animal experimentation , so it was not necessary to obtain an animal ethics license from the Animal Experiment Board . The species captured in this study were not protected wildlife and were not included in the China Species Red List . The principal component analysis was performed using the 2006–2010 data on natural factors ( relative rodent density , NDVI , monthly average temperature , monthly rainfall and monthly average relative humidity ) and social factors ( GDP and urbanization rate ) . Four principal components that included three natural components ( F1: rodent density , NDVI for rice paddies and temperature; F2: rainfall and relative humidity , and F3: rodent density and relative humidity ) and one social component ( F4: GDP and the urbanization rate of the population ) were extracted . Cross correlation analysis , adjusted for seasonality , was performed to infer the time-lag effects between variables . Each sequence of variables was filtered to convert it to white noise before proceeding with the cross correlation analysis . The correlation between the residual sequence of HFRS incidence and the residual sequences of the environmental variables ( F1 , F2 , F3 , F4 ) , lagged 0∼6 months , was then calculated . To confirm the correlation between lagged variables and HFRS incidence , the polynomial distributed lag ( PDL ) model with a lagged dependent variable was used to examine the contribution of various variables to HFRS incidence . The PDL model was: ( 2 ) ( 3 ) ( 4 ) where Y is the dependent variable , α is the regression coefficient of the independent variable , n and m are the lag phases , β and γ are the lagged regression coefficients , and K is the random disturbance term .
A total of 321 HFRS cases were reported in Chenzhou , and yearly average HFRS incidence remained stable , during the study period . Yearly average HFRS was 1 . 53/100 , 000 ( 71 cases ) in 2006 , 1 . 59/100 , 000 ( 74 cases ) in 2007 , 1 . 29/100 , 000 ( 61 cases ) in 2008 , 1 . 41/100 , 000 ( 67 cases ) in 2009 , and 0 . 96/100 , 000 ( 48 cases ) in 2010 , Analysis of monthly HFRS cases revealed that HFRS incidence was higher from November to January and lower in March , April , July , and August ( Figure 2 ) . A total of 251 rodents were captured at specific industry monitoring sites ( catering industry and processing industry in Beihu and Suxian Districts ) . Captured rodents consisted mostly of the species Rattus norvegicus , Rattus flavipectus , and Mus musculus , which are known hosts of hantavirus [21] . The capture rate was 2 . 06 ( per 100 trap-nights ) . A total of 125 rodents were captured in residential monitoring sites in Beihu District; the capture rate was 1 . 04 ( per 100 trap-nights ) . A total of 322 rodents ( mainly R . flavipectus and M . musculus ) were captured at rural monitoring sites ( Table 2 ) . The capture rate was 2 . 68 ( per 100 trap-nights ) . There was an annual peak of rodent density from April to September . The maximum capture rate was 4 . 64 , which was recorded in April , 2006 . The minimum capture rate was 0 . 81 , recorded in January , 2009 ( Figure 2 ) . The monthly NDVI for cultivated land ranged between 0 . 3 to 0 . 8 . The NDVI increased from January to July , and then decreased each month after the peak of variation from August to October . The peak HFRS incidence was preceded by the peak NDVI and the peak monthly average temperature and monthly average rainfall , with a 3∼4 month lag ( Figure 3 , Figure 4 ) . The results of the analysis of the relationship between HFRS incidence and socioeconomic factors ( GDP and urbanization rate ) indicated that the urbanization rate was significant negatively correlated with HFRS incidence ( r = −0 . 903 , P<0 . 001 ) and GDP ( r = −0 . 627 , P<0 . 05 ) . HFRS incidence declined as urbanization rate and GDP increased . The results of the principal component analysis revealed that component 1 ( F1 ) , component 2 ( F2 ) and component 3 ( F3 ) accounted for 91 . 66% of the total variability in natural factors . F1 was closely associated with rodent density , NDVI , and temperature . F2 was closely associated with rainfall and humidity . F3 was closely associated with rodent density and humidity ( Table 3 ) . Component 4 ( F4 ) accounted for 85 . 58% of the total variation in socioeconomic factors ( Table 3 ) . The correlation between HFRS incidence , the variables , the three natural components ( F1 , F2 , F3 ) , and the socioeconomic component ( F4 ) were calculated with a lag of 0∼6 months . Monthly HFRS incidence was positively correlated with rodent density with a 6-month lag ( r = 0 . 354 , P = 0 . 009 ) . HFRS incidence was preceded by NDVI ( r = 0 . 49 , P<0 . 001 ) , temperature ( r = 0 . 515 , P<0 . 001 ) and rainfall ( r = 0 . 414 , P = 0 . 002 ) , with a 5-month lag ( Table 4 ) . Monthly HFRS incidence was correlated with F1 ( rodent density , NDVI for rice paddies and temperature; r = 0 . 179 , P = 0 . 192 ) and F2 ( rainfall and relative humidity; r = −0 . 289 , P = 0 . 032 ) , with a 4-month lag; HFRS incidence was preceded by F3 ( rodent density and relative humidity ) with a 5-month lag ( r = −0 . 523 , P<0 . 001 ) . HFRS incidence was correlated with F4 ( GDP and urbanization rate of population; r = −0 . 376 , P = 0 . 003 ) ( Table 4 ) . The PDL model yielded the best fit based on the R-squared and AIC ( Akaike information criterion ) . First , three principal components based on natural factors were used to build Model 1 ( R2 = 0 . 656 , AIC = 5 . 023 ) . Socioeconomic factors ( F4 ) were included in Model 2 ( R2 = 0 . 677 , AIC = 5 . 106 ) . Finally , 2nd-order autoregression was considered in Model 3 , which indicated that the number of notified HFRS infections in the current month was related to the numbers of cases occurring in the previous 1 and 2 months ( R2 = 0 . 857 , AIC = 4 . 799 ) . The results of the optimal model ( Model 3 ) indicated that HFRS incidence was affected not only by the natural factors but also by the socioeconomic factors ( Table 5 ) . In addition , monthly HFRS incidence was strongly autocorrelated . The estimated/expected number of cases from the regression model fits very well to the observed number of HFRS cases , including the peak values ( Figure 5 ) .
To the best of our knowledge , little is known about the combined effect of environmental variations ( animal reservoirs , natural and socioeconomic factors ) on the transmission and persistence of HFRS . In general terms , rodent density and the extent of high-risk behaviors depend on natural factors . Changes in the animal reservoir may lead to emergence of new epidemics , and threats to human health . However , economic development may improve the residential environment , which could inhibit disease transmission from rodent vectors to humans through decreased contact . In our analysis , the optimal model revealed that HFRS incidence was positively correlated with rainfall and relative humidity in Chenzhou . Rainfall is an important factor in HFRS morbidity , because increased rainfall provides better growth conditions for vegetation that directly or indirectly provides rodents with food , which leads to increases in rodent populations [13] . There is also a very close association between wet or very humid habitat types , and rodent population size [22] , [23] , because the moist environment provides suitable conditions [21] . HFRS epidemic areas are mostly distributed in low-lying moist regions or sub-humid regions [12] , [21] . From 2006 to 2010 , monthly average relative humidity ranged between 60% to 85% , which is conducive for the transmission HFRS . The maximum average relative humidity occurred in January and February . The minimum average relative humidity was in July and August . Temperature and NDVI were important factors for HFRS epidemics and were positively associated with HFRS incidence . Temperature can affect rodent pregnancy rate , litter size , birth rate , and survival rate , and is an important factor in the fluctuation of rodent population size [24] . Rodent survival is greater during warmer winters than in colder winters , which leads to greater rodent population densities [25] , [26] . There was a peak in temperature from June to August , and the peaks in HFRS cases were in May and June , and from November to January , indicating that HFRS incidence lagged behind temperature by approximately 4∼5 months . Temperature can also directly affect the geographic distribution of rodents , because they prefer warmer areas . Temperature is positively correlated with vegetation , which provides food for rodents . Thus , rodent density directly and indirectly depends on the local temperature [27] . Temperature ranged between 5°C to 30°C during the study period , which may have increased the rodent population size and indirectly increased HFRS morbidity in Chenzhou . The NDVI reflects the level of vegetation coverage [28] , which is a good indicator of food and living conditions for rodents . It is correlated with the amount and productivity of vegetation and crops . Most rodent species responded directly to fluctuations in food availability , and population densities are driven by changes in food resources [29] , [30] . Vegetation also provides shelter and safety ( e . g . , from predators ) . HFRS incidence was positively correlated with rodent density . This result indicates that fluctuations in rodent populations had an important effect on HFRS incidence . Rodent population density peaked in March , April , August , and September . The peaks in HFRS cases were in May and June , and from November to January , indicating that HFRS incidence lagged behind rodent density by approximately 2∼3 months . Hantavirus infection rates can increase with increased rodent density if the infected rodents increase their contact with humans [31]–[33] . Therefore , our results suggest that rodent density fluctuations could be used to forecast changes in HFRS incidence and that HFRS transmission could be controlled by reducing the number of rodents in residential areas . HFRS incidence was negatively correlated with GDP ( r = −0 . 627 ) and the urbanization rate ( r = −0 . 903 ) . HFRS incidence decreased with the increase in per capita GDP and urbanization rate . These results suggest that economic development may reduce HFRS transmission , which is consistent with the findings of a previous study [13] . Rodent density decreases with the development of economy and culture . It is generally low in developed countries . Rodent density is lower in the developed areas than in less developed areas in China [34] . Economic development has led to improvements in living conditions , because the size of polluted , disorganized , and poor areas have been reduced . Meanwhile , there have been improvements in deratization methods and in public awareness about rodent prevention and control . Variation in the ecology of the environment that results from extensive construction will certainly have direct or indirect effects on the living conditions and food for rodents , thus leading to variation in disease transmission intensity [35] . There has been increased awareness of diseases prevention and control measures as health care services have improved . However , the environment of most villages in China is suitable for rodent survival and development , and high rodent densities persist in many fields and villages [34] . Peasants work long hours in areas where rodents are active , and increased contact with the animals' secretions ( e . g . feces , urine and saliva ) means that farmers are the main high risk group for HFRS [4] . People in rural areas come into contact with rats more frequently than in urban areas , and the large number of rural residents migrating into cities may be another explanation for the year-by-year decrease in HFRS . Pollution of the fields and other vegetation areas by fertilizers , pesticides , and heavy metals also affects the living conditions and food availability for rodents , and the toxic effects of these substances has a negative effect on population growth [36] . Compared to our previous work in Changsha [11] , this analysis indicates many similarities between these two areas . HFRS was positively correlated with rodent density and the NDVI , and was influenced by temperature and rainfall in Chenzhou and Changsha . This similarity is likely a reflection of comparable host behaviors and habitats . The primary difference between this study and the Changsha study is that we incorporated socioeconomic factors into this analysis . Few studies are currently examining the relationship between HFRS incidence and socioeconomic factors . In this study , we comprehensively considered the effects of natural and socioeconomic factors on HFRS incidence . The addition of socioeconomic variables improved the model fit . Models based on natural and social variables had better performance ( R2 = 0 . 677 ) than models based on only natural variables ( R2 = 0 . 656 ) . In this study , the PDL model used component variables , which reduced the effect of higher order multicollinearity among variables and improved the fit . There were some limitations of this study . First , the monthly average temperature , which was measured in the air , was different from the surface temperature . Surface temperature has a more direct effect on rodents , so it would be more informative to incorporate surface temperature into models of HFRS incidence . Second , HFRS cases were from a passive , instead of active , surveillance system , so some cases may not have been identified . Patients with less serious or less obvious symptoms may not seek medical care , which would result in an underestimate of the true incidence . Finally , the effects of extreme weather conditions ( e . g . , high temperature , torrential rain , and drought ) on the survival and reproduction of rodents , and on HFRS transmission , needs further study . Furthermore , this was a population-level study , and the potential of the ecological fallacy to affect the results is unavoidable in a study of this kind . In conclusion , changes in the risk of HFRS may be the result of changes in contact between humans and the rodent reservoir , which are caused by changes in natural and socioeconomic factors . The results of our analysis provide theoretical support for this hypothesis and indicate that further study of variation in HFRS incidence would be beneficial for the prevention and control of this disease .
|
Hemorrhagic fever with renal syndrome ( HFRS ) , a rodent-borne disease caused by hantaviruses , is characterized by fever , haemorrhage , headache , back pain , abdominal pain , and acute kidney injury . China has the highest incidence of HFRS worldwide . Reported cases account for 90% of the total global cases . Approximately 1 . 4 million HFRS cases were reported in China between 1950 and 2010 . During the same time period , >46 000 people died from HFRS , and the fatality rate was 3 . 29% . A great deal of interest and excitement has developed recently for understanding the role of the environment in the transmission of HFRS . Our article provides evidence that rodent density and behavior depend on natural factors . Changes in animal reservoirs may lead to the emergence of new epidemics and threats to human health . However , economic development may promote a more residential environment , which could inhibit disease transmission from animals to humans by limiting their contact . We combined data about the rodent reservoir , the natural environment , and socioeconomic factors in the model . The results will be helpful for making and prioritizing preventive measures .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"environmental",
"epidemiology",
"epidemiology",
"neglected",
"tropical",
"diseases"
] |
2014
|
Animal Reservoir, Natural and Socioeconomic Variations and the Transmission of Hemorrhagic Fever with Renal Syndrome in Chenzhou, China, 2006–2010
|
The Escherichia coli mazEF module is one of the most thoroughly studied toxin–antitoxin systems . mazF encodes a stable toxin , MazF , and mazE encodes a labile antitoxin , MazE , which prevents the lethal effect of MazF . MazF is an endoribonuclease that leads to the inhibition of protein synthesis by cleaving mRNAs at ACA sequences . Here , using 2D-gels , we show that in E . coli , although MazF induction leads to the inhibition of the synthesis of most proteins , the synthesis of an exclusive group of proteins , mostly smaller than about 20 kDa , is still permitted . We identified some of those small proteins by mass spectrometry . By deleting the genes encoding those proteins from the E . coli chromosome , we showed that they were required for the death of most of the cellular population . Under the same experimental conditions , which induce mazEF-mediated cell death , other such proteins were found to be required for the survival of a small sub-population of cells . Thus , MazF appears to be a regulator that induces downstream pathways leading to death of most of the population and the continued survival of a small sub-population , which will likely become the nucleus of a new population when growth conditions become less stressful .
Toxin-antitoxin modules consist of pairs of genes on the bacterial chromosome [1]–[5]: the downstream gene encodes a stable toxin which causes cell death and the upstream gene encodes a labile antitoxin which counteracts the activity of the toxin . In the E . coli chromosome , seven toxin-antitoxin modules have been identified [3] , [6]–[10] . Among these , one of the most studied is the mazEF system , which was the first to be described as regulatable and responsible for bacterial programmed cell death [11] . mazF encodes the stable toxin MazF and mazE encodes for the labile antitoxin MazE . MazE is degraded by the ATP-dependent ClpAP serine protease [11] . MazF is an endoribonuclease which cleaves mRNAs at ACA sequences in a ribosome-independent manner [12] , [13] . As long as MazE and MazF are co-expressed , MazE counteracts the toxic activity of MazF [11] . Under stressful conditions [11] , [14]–[17] that inhibit mazEF expression , the de novo synthesis of both MazE and MazF is prevented: because MazE is much more labile than MazF , the cellular amount of MazE decreases faster than that of MazF , permitting MazF to act freely , eventually causing cell death [11] . Note that mazEF-mediated cell death is a population phenomenon requiring a quorum-sensing factor called EDF [18] , [19] . Here , we found that the process of mazEF-mediated cell death is more complex than has previously been understood . We show that , as previously reported [12] , [20] , MazF induction causes the inhibition of protein synthesis . But we were particularly interested to find that this inhibition was not complete: though MazF led to the inhibition of the synthesis of most proteins , it selectively enabled the synthesis of other specific proteins . Some of those specific proteins were required for the death of most of the population . Surprisingly , we also found that MazF enabled the synthesis of proteins that permitted the survival of a small sub-population under those stressful conditions that cause mazEF-mediated cell death for the majority of the population . These findings further support our understanding that mazEF-mediated cell death is a population phenomenon .
It has been previously reported that MazF inhibits protein synthesis [12] , [20] . Here , we performed similar studies on the effect of MazF on protein synthesis . We compared the rate of incorporation of [35S]methionine into the acid insoluble fraction in MazF-induced and uninduced bacterial cell cultures . Our careful analysis revealed that , after MazF-induction , though most protein synthesis was inhibited , a low level of protein synthesis remained . Even as long as 30 minutes after MazF induction , about 10% protein synthesis was observed compared to the level in the control culture ( Figure 1A ) . We asked: Was the synthesis of all of the proteins reduced to a basal level ? Or perhaps a small selected group of proteins continued to be synthesized ? Using 1D-gels , we analyzed the mobility of the proteins that were synthesized after MazF induction: within fifteen minutes , while most protein synthesis was prevented , some clear , sharp radioactive bands appeared ( Figure 1B and 1C ) . These results suggest that while MazF-induction lead to the inhibition of synthesis of most proteins in E . coli , the synthesis of an exclusive group of proteins was still permitted . It should be noted that the results shown in Figure 1 were obtained by MazF induction in E . coli strain MC4100 relA1 . This because in our previous studies we have shown that MazF induction causes an irreversible loss of viability in this strain [21] . In addition , we also used 1D-gels to examine the effect of MazF induction on E . coli strain MC4100 relA+ . We found that MazF induction affected both strains identically ( data not shown ) . To better resolve the differences between the profiles of protein synthesis in cultures in which MazF had been induced or not , we took samples which we had previously applied to 1D-gels ( Figure 1B and 1C ) , and subsequently applied them to 2D-gels . Superimposing the autoradiograms of gels of these two cultures revealed that the presence of MazF led to a dramatic change in the profile of protein synthesis in E . coli ( Figure 2A ) . This change is reflected in the size of the synthesized proteins . Clearly , the synthesis of proteins whose molecular weight was greater than ∼20 kDa tended to be inhibited ( Figure 2A ) , while the synthesis of proteins whose molecular weight was less than ∼20 kDa tended to be increased . We verified this observation by computer analysis ( Figure 2B and 2C ) : the molecular weights of most of the proteins whose level of synthesis was increased by two times were less than ∼20 kDa ( Figure 2B ) ; the molecular weights of most of the proteins whose level of synthesis was decreased by two times were more than ∼20 kDa ( Figure 2C ) . To exclude the possibility that the observed increase in the level of small proteins was a result of degradation of larger proteins , we performed a pulse-chase experiment . During the period examined after MazF induction , we found no change in the general stability of the cellular proteins ( Figure S1 ) . We wondered if proteins whose level of synthesis was not reduced after MazF induction were required for cell death . From our 2D-gels ( Figure 2 ) , we chose to examine 13 proteins that correspond to this criterion . We extracted these selected proteins from a 2D-gel of an unlabeled MazF-induced culture that we had prepared in parallel with the labeled culture . We identified the nature of those proteins by mass-spectrometry; their positions in the gel are shown in Figure 3A . The identified proteins whose synthesis was increased after MazF induction were: ClpP , Crr , ElaC , NfnB , RsuA , SlyD , YajQ , and YfbU ( see Table 1 for the increment in the level of synthesis of each protein ) . The proteins whose level of synthesis did not change significantly after MazF induction were: AhpC , DeoC , EF-P , YfiD , and YgcR ( see Table 1 for the level of synthesis of each protein ) . To examine the involvement in cell death of each of these proteins , we deleted each of the genes encoding them individually from the E . coli MC4100 relA+ chromosome . Under stressful conditions , we compared the viability of these deleted mutants to that of the WT and its ΔmazEF derivative . We chose stressful conditions that we had previously shown to cause mazEF-dependent cell death [14] , [17]: ( a ) brief inhibition of translation by spectinomycin or ( b ) DNA damage caused by nalidixic acid . As mentioned above , the effect of MazF induction on protein synthesis was identical in both strains E . coli MC4100 relA1 and E . coli MC4100 relA+ . Because mazEF-mediated cell death under stressful growth conditions requires the presence of the relA gene [17] , we only examined the effect of the deleted mutants in strain E . coli MC4100 relA+ . With respect to cell survival under the stressful conditions that we used , we found three types of mutants . The mutants in the first group behaved like the ΔmazEF derivative , that is , most of the population survived . The second group surprised us because the mutants in this group were significantly less viable than was the WT strain . The mutants in the third group behaved like the WT strain and thus , at least under the stressful conditions examined , were irrelevant to our study of cell death . The mutants in the first group were ΔclpP , ΔslyD , ΔyfiD , ΔelaC , ΔygcR , and ΔyfbU . Among these , only ΔclpP , ΔslyD , and ΔyfiD behaved like the ΔmazEF derivative under both stressful conditions , inhibition of translation ( Figure 3B ) and DNA damage ( Figure 3C ) . The mutants , ΔygcR and ΔyfbU behaved like ΔmazEF only under conditions causing DNA damage ( Figure 3C ) . The mutant ΔelaC behaved like ΔmazEF only under conditions causing the inhibition of translation ( Figure 3B ) . We called clpP , slyD , yfiD , elaC , ygcR and yfbU “Death Genes” , noting that some were involved in cell death under both conditions of inhibition of translation and DNA damage , and some were involved only when the DNA was damaged or when translation was inhibited . When translation was inhibited only briefly , the mutants in the second group , ΔyajQ , ΔdeoC , and ΔrsuA , were significantly less viable than the WT strain ( Figure 3B; for a logarithmic-scale view of the results see Figure S2 ) . When we deleted each of these genes individually , the level of survivors in the population was dramatically reduced from about 10% ( for the WT strain ) to about 2% ( for the deleted mutants ) . Thus , we called yajQ , deoC , and rsuA “Survival Genes” . Note that we observed no correlation between the growth rates of these mutants and their relevance to mazEF-mediated cell death ( Figure S3 ) . Here are some examples: ( i ) on one hand , the growth rates of the mutants of elaC and slyD , which encode “Death Genes” , resembled that of the ΔmazEF strain . On the other hand , the growth rates of the mutants of yfiD , yfbU and clpP , which also encode “Death Genes” , were much slower than the growth rate of the WT strain ( Figure S3A ) ; ( ii ) the growth rate of the mutant of yfiD , which encodes a “Death Gene” , resembled that of the mutant of ahpC , which did not show any relevance to mazEF-mediated cell death ( Figure S3A ) ; ( iii ) the growth rates of any one of the mutants of the genes rsuA , yajQ or deoC , which encode “Survival Genes” , were much slower than the growth rate of the WT strain ( Figure S3B ) . However , the growth rate of the mutant of efp , which appeared not to be involved in mazEF-mediated cell death , was much slower than the growth rates of those “Survival Genes” . In addition , we compared the CFUs of the above mentioned mutants to the CFUs of the WT and the ΔmazEF strains . The CFU was determined at OD600 0 . 6 , the stage where we examined the viability of each strain as shown in Figure 3B and 3C . We did not observe any significant difference between the CFUs of those strains ( data not shown ) . We have recently discovered that there are two mazEF-mediated cell death pathways - an ROS-dependent and ROS-independent [22] . The ROS-dependent pathway is induced by the inhibition of transcription and/or translation , and the ROS-independent patheway is induced by DNA damaging agents . Here we have shown that there are additional genes to mazEF that are involved in mazEF-mediated cell death . Moreover , we have shown that during this death process , some of those genes function as “Survival Genes” . Therefore , we asked whether the genes soxS and soxR , known to be involved in ROS detoxification [23] , might function as “Survival Genes” in the ROS-dependent mazEF-mediated cell death pathway . To this end , each of the genes soxS and soxR were individually deleted from the chromosome of E . coli MC4100 relA+ . Once again , we compared the viability of those deleted mutants to that of the WT and its ΔmazEF derivative under the following stressful conditions: ( a ) brief inhibition of translation by spectinomycin or ( b ) DNA damage caused by nalidixic acid . As we expected , the mutants ΔsoxS and ΔsoxR were significantly less viable than the WT strain upon a brief inhibition of translation ( Figure 3B ) – a stressful condition which induces a ROS-dependent mazEF-mediated cell death pathway [22] . The level of survivors in the population was dramatically reduced from about 10% ( for the WT strain ) to about 1–2% ( for the deleted mutants ) . In contrast , the viability of ΔsoxS and ΔsoxR strains resembled that of the WT strain when DNA was damaged ( Figure 3C ) - a stressful condition which induces a ROS-independent mazEF-mediated cell death pathway [22] . Therefore , we suggest that soxS and soxR function indeed as “Survival Genes” in ROS-dependent mazEF-mediated cell death pathway .
Until now , it has been understood that MazF causes the complete inhibition of protein synthesis [12] , [24] . Here , when we performed incorporation experiments similar to those previously done by others [12] , [20] , we indeed observed a dramatic reduction in the level of protein synthesis . However , in contrast to previous reports , we found that the inhibition of protein synthesis was incomplete: a basal level of about 10% protein synthesis remained ( Figure 1A ) . Comparing MazF-induced and MazF-uninduced cultures in 1D-gels revealed that this basal level of protein synthesis remaining after MazF induction represented an exclusive group of proteins ( Figure 1B and 1C ) . More thoroughly analyzing those results on 2D-gel revealed that MazF induction led to a clear change in the pattern of protein synthesis ( Figure 2 ) . After MazF induction , we observed an increase in the level of synthesis of proteins whose molecular weight was smaller than ∼20 kDa ( Figure 2A and 2B ) , but a decrease in the level of synthesis of proteins whose molecular weight was greater than ∼20 kDa ( Figure 2A and 2C ) . MazF is an endoribonuclease that cleaves mRNAs at ACA sequences in a ribosome-independent manner [12] , [13] . For this research we used mass-spectrometry to identify 13 proteins that were synthesized within a period of 15 minutes after MazF induction ( Figure 3A ) . We observed that each of the mRNA sequences encoding these proteins carried at least one ACA sequence ( data not shown ) . Since the mRNAs of these proteins carry the MazF's target site , how could those proteins be synthesized after MazF induction ? A possible explanation is that there is an as yet unknown mechanism that protects those mRNAs from cleavage by MazF , or at least reduces the rate of cleavage in comparison to the other mRNAs in E . coli . We are currently searching for such a mechanism that would allow the selective synthesis of those proteins . We also found that some of the proteins selectively synthesized after MazF induction were required for cell death ( Figure 3B and 3C ) . Thus , while inhibiting bulk protein synthesis ( Figure 1 ) , it seems that MazF also enabled the selective synthesis of proteins essential for cell death ( Figure 3 ) . The genes encoding the proteins , which are essential for cell death , can be divided into three groups: ( a ) ygcR and yfbU are involved in cell death only when triggered by DNA damage ( Figure 3C ) but not in cell death triggered by the inhibition of translation ( Figure 3B ) ; ( b ) elaC is involved in cell death only when triggered by the inhibition of translation ( Figure 3B ) but not in cell death triggered by DNA damage ( Figure 3C ) ; ( c ) clpP , slyD , and yfiD are involved in cell death triggered by both the inhibition of translation ( Figure 3B ) and DNA damage ( Figure 3C ) . These results suggest that there may be at least two separate death pathways that may share some common steps . What are the roles of the genes that were found by us to be required for mazEF-mediated cell death in E . coli ? ( i ) slyD encodes a peptidyl prolyl cis/trans-isomerase [25] , [26] which also functions as an E . coli chaperone [27] , [28] . SlyD is also involved in the insertion of Ni2+ during the maturation of hydrogenases [29] . Moreover , SlyD is required for phage φX174-induced cell lysis [25] , [30] where it appears to stabilize the φX174 lysis protein E [27] . We have not yet tested if these functions of SlyD also contribute to mazEF-mediated cell death . However , the involvement of SlyD in cell lysis is very intriguing and is currently under our investigation . ( ii ) yfiD encodes a glycyl radical protein that can replace a pyruvate formate-lyase subunit that has been damaged by oxidation [31] . Our recent discovery that ROS is produced during mazEF-mediated cell death [22] may provide a clue how the product of yfiD is involved: YfiD may enable the ROS-sensitive protein pyruvate formate-lyase to function during the death process . ( iii ) clpP has already been shown to be involved in mazEF-mediated cell death [11] . The ATP-dependent ClpAP serine protease degrades MazE antitoxin . When mazEF expression is inhibited by specific stressful conditions , there is no de novo synthesis of MazE and MazF . Then , ClpAP degrades MazE and the concentration of MazE is reduced . In the absence of MazE , the stable MazF can act freely and cause cell death . In addition , the ATP-dependent ClpXP protease is involved in the synthesis of the communication signaling peptide EDF which is required for mazEF-mediated cell death [19] . Here we show that MazF induction causes an increase in the amount of the intracellular ClpP . This may be a part of a positive feedback loop in which the increase in ClpP will cause both a decrease in the level of MazE and an increase in the level of EDF . However , we cannot exclude the possibility that ClpP has an additional role in the cell death process , downstream from MazF activity . ( iv ) Generally , CCA is the consensus sequence required for a tRNA to be charged with an aminoacyl group . elaC encodes RNase BN that cleaves the 3′-terminal portion of tRNA if it differs from CCA [32] . In fact , in E . coli , the contribution of RNase BN as a 3′-terminal nuclease remains elusive since E . coli has no tRNAs lacking the CCA sequence at their 3′-termini [32] , [33] . Recently , it has been suggested that RNase BN may also be responsible for cleaving unstructured RNAs [34] . At this stage we cannot determine whether these functions of RNase BN are connected to mazEF-mediated cell death or whether this enzyme may have additional functions essential for mazEF-mediated cell death . ( v ) yfbU and ygcR encode for proteins of unknown function . Here we show , for the first time , that those genes are required for at least one cellular process in E . coli – programmed cell death . The possible roles of the identified “Death Proteins” are summarized in Table 1 . Note that not all of the genes that encoded proteins that were selectively synthesized after MazF induction were part of the death pathway ( s ) . We found that the proteins encoded by yajQ , rsuA , and deoC were not at all involved in the death of the greater part of the cell population . Instead , we found that these genes , whose gene products were selectively synthesized after MazF induction , supported the survival of a small sub-population ( Figure 3B ) . These results indicate that MazF enabled the simultaneous synthesis of specific proteins essential for the death of most of the population and of specific proteins essential for the survival of a small sub-population . How could these “Survival Genes” contribute to the survival of a small sub-population under stressful conditions causing mazEF-mediated cell death ? We will discuss each of these genes separately: ( i ) soxS and soxR are involved in ROS detoxification [23] . We have recently discovered that there are two mazEF-mediated cell death pathways - an ROS-dependent and ROS-independent [22] . The first is induced by the inhibition of transcription and/or translation and the second by DNA damaging agents . Based on our current discovery that soxS and soxR are essential for the survival of a small sub-population only under inhibition of translation ( Figure 3B ) , we suggest that these genes support cell survival by detoxifying ROS [21] . ( ii ) deoC encodes deoxyribose-phosphate aldolase that is involved in the catabolism of deoxyribonucleosides in E . coli [35] . It was reported [36] that strain E . coli deoC− , in which a deoC of S . mutans was expressed , could grow on glucose minimal medium supplemented with deoxynucleotides . This makes it seem likely that the major sub-population , which undergoes a mazEF-mediated cell death process , may releases deoxynucleotides into the medium . The rest of the population , still alive , could survive by using those deoxynucleotides as a carbon and energy source . Another possibility is that deoC may contribute to the survival of a small sub-population by being involved in ROS detoxification . Like soxS and soxR , which are known to be involved in ROS detoxification [23] , deoC is essential to cell survival only upon the inhibition of translation ( Figure 3B ) which triggers ROS-dependent mazEF-mediated cell death [22] . ( iii ) rsuA encodes an enzyme which catalyzes pseudouridylation at position 516 in the 16S rRNA [37] , [38] , and ( iv ) yajQ encodes a protein of unknown function . We cannot yet determine how rsuA and yajQ can contribute to the above mentioned survival of a small sub-population . However , as suggested for deoC , we can speculate that these genes may also be involved in ROS detoxification . Once again , we base our suggestion on our finding that these genes are involved in cell survival only in ROS-dependent mazEF-mediated cell death pathway [22] , triggered by the inhibition of translation ( Figure 3B ) , and not in ROS-independent mazEF-mediated cell death pathway [22] , triggered by DNA damage ( Figure 3C ) . The possible roles of the identified “Survival Proteins” are summarized in Table 1 . Here we have shown , for the first time , that MazF induced downstream pathways required for both death and life , confirming our hypothesis [7] , [8] , [21] that MazF is a regulator of cell death rather than the cell executioner . This dual effect of MazF on two such opposite processes , cell death and cell survival , may provide an evolutionary rational to mazEF-mediated cell death . We suggest that when exposed to stressful conditions , while most of the bacterial cell population undergoes programmed cell death , an active process keeps a small fraction of the population alive . When the growth conditions become less stressful , these survivors probably become the nucleus of a new population . We have previously reported [18] , [19] that mazEF-mediated cell death is a population phenomenon requiring a quorum-sensing factor called EDF . That mazEF-mediated cell death is indeed a population phenomenon is strongly supported by the results of our work here showing that MazF induction contributed both to the death of most of the population and to the survival of a small sub-population . It should be noted that an analogous phenomenon , in which an active process of cell death of a sub-population enables the survival of the rest of the population , was found in Bacillus subtilis [39] , [40] . Based on our present results , we have developed our model [8] for mazEF-mediated cell death process ( Figure 4 ) . As we have shown previously [11] , [14]–[17] , inhibiting mazEF expression by various stressful conditions leads to the reduction in the cellular amount of the labile antitoxin MazE . Thereby , the stable toxin MazF can act freely as an endoribonuclease . As we have reported here , the unrestricted action of MazF leads to the inhibition of the synthesis of many proteins , particularly those larger than ∼20 kDa ( Figure 2 ) . However , some proteins , particularly those smaller than ∼20 kDa , can still be selectively synthesized ( Figure 2 ) . At least six of those proteins , which are selectively synthesized after MazF activation , are necessary for implementing the death of most of the cell population ( Figure 3 ) . Moreover , it seems that more than one death pathway can be activated by MazF . The specific pathway chosen appears to be a function of the particular stressful condition , like DNA damage or the inhibition of protein synthesis ( Figure 3 ) . We believe that the cell is led towards its own death by the combination of the inhibition of the general synthesis of proteins , necessary for life , and the parallel synthesis of proteins necessary for the death process . Furthermore , while at least six of the selectively synthesized proteins are required for the death of most of the cell population , at least three other small proteins , also selectively synthesized after MazF activation , are required for the survival of a small sub-population ( Figure 3 ) . It seems likely that the survival of that small sub-population would be supported by the dead cells , that would then release nutrients and other factors , like signal molecules , essential for survival .
We used E . coli strains MC4100 relA1 , MC4100 relA+ , and MC4100 relA+ ΔmazEF , which we have described previously [11] , [14] , [15] , [17] . In addition , using the procedure of Datsenko and Wanner [41] , we constructed the following derivatives of MC4100 relA+: MC4100 relA+ ΔahpC , MC4100 relA+ ΔclpP , MC4100 relA+ Δcrr , MC4100 relA+ ΔdeoC , MC4100 relA+ Δefp , MC4100 relA+ ΔelaC , MC4100 relA+ ΔnfnB , MC4100 relA+ ΔrsuA , MC4100 relA+ ΔslyD , MC4100 relA+ ΔsoxS , MC4100 relA+ ΔsoxR , MC4100 relA+ ΔyajQ , MC4100 relA+ ΔyfbU , MC4100 relA+ ΔyfiD , and MC4100 relA+ ΔygcR . Plasmid pSA1 is a derivative of pQE30 ( Qiagen , Hilden , Germany ) bearing lacIq and also bears mazF under the control of the T5 promoter and the lac operator . For viability assays , cells were grown in M9 minimal medium containing 1% glucose and a mixture of amino acids ( except for tyrosine and cysteine ) , each at 100 µg/ml . The cells were plated on rich Luria-Bertani ( LB ) agar plates as described previously [14] , [17] . For labeling experiments , cells were grown in M9 minimal medium containing 0 . 2% glucose and a mixture of amino acids ( except for methionine , tyrosine , tryptophan , and cysteine ) , each at 20 µg/ml . Strain MC4100 relA1 was transformed with pSA1 bearing mazF . The culture was grown in M9 medium without methionine ) with the addition of 100 µg/ml ampicillin , at 37°C . When the culture was in mid-logarithmic phase ( OD600 0 . 5 ) , it was divided in half , and each half was diluted 1∶200 . Cold methionine at 0 . 125 µg/ml was added to both sub-cultures . One sub-culture was kept as a control; to the other sub-culture 5 µM isopropyl β-D-thiogalactopyranoside ( IPTG ) was added to induce MazF synthesis . Immediately after induction by IPTG , both sub-cultures were labeled with [35S]methionine ( 13 . 75 µCi/ml ) and incubated at 37°C , without shaking . At various time intervals , samples were withdrawn and the reactions were stopped by the addition of trichloroacetic acid ( TCA ) to a final concentration of 5% , after which the reaction tubes were placed in ice . The samples were filtered through 0 . 45 µM filters using a vacuum pump . A BETAmatic I/II scintillation counter ( KONTRON ) was used to determine the radioactivity in the TCA-insoluble material . E . coli MC4100 relA1 , harboring plasmid pSA1 , was grown to mid-logarithmic phase ( OD600 0 . 5 ) as described above . Then , the culture was divided into two and 1 mM IPTG was added to one half of the culture . Both sub-cultures were incubated at 37°C , without shaking , for 15 min . [35S]methionine ( 110 µCi/ml ) was added to each sub-culture which were then further incubated at 37°C , without shaking , for 5 min . The labeling reaction was terminated by placing the samples in liquid nitrogen . The samples were centrifuged at 14000 rpm , for 10 min . The pellets were washed in 50 mM tris ( hydroxymethyl ) aminomethane ( Tris ) pH 7 . 5 and then resuspended in lysis buffer ( 0 . 5 mg/ml lysozyme , 10 mM Tris pH 8 , 1 mM ethylene diamine tetraacetic acid ( EDTA ) , 20 µg/ml DNase , 50 µg/ml RNase ) and 10% sodium dodecyl sulfate ( SDS ) . Lysates were incubated at 90°C for 5 min . These prepared lysates were loaded either onto a 10% SDS polyacrylamide gel [42] or onto a 16% N-Tris ( hydroxymethyl ) methylglycine ( Tricine ) -SDS polyacrylamide gel [43] . In addition , samples prepared for 1D-gel analysis were centrifuged ( 8000 rpm at 4°C for 5 min ) and then washed twice with cold Tris-EDTA and Phenylmethylsulfonyl Fluoride ( TE-PMSF ) ( 10 mM Tris pH 7 . 5 , 1 mM EDTA , 1 . 4 mM PMSF ) . The washed cells were resuspended in 0 . 5 ml of TE-PMSF and disrupted by sonication . Cell debris and protein aggregates were removed by centrifugation at 14000 rpm at 4°C for 30 min . The protein concentrations of the remaining supernatants were determined using the Bradford method with the BioRad Protein Assay kit ( Hercules , CA , USA ) [44] . These protein containing supernatants were lyophilized and further prepared for 2D-gel analysis as described previously [45] . Both the 2D-gel analysis and the determination of the level of increment in protein synthesis were done by the use of Delta2D software ( DECODON GmbH , Greifswald , Germany ) . E . coli strain MC4100 relA1 was transformed with pSA1 bearing mazF . The culture was grown in M9 medium without methionine with the addition of 100 µg/ml ampicillin at 37°C . When the culture was in mid-logarithmic phase ( OD600 0 . 5 ) , it was labeled with [35S]methionine ( 220 µCi/ml ) . The labeled culture was incubated at 37°C , without shaking , for 5 min . Then , both cold methionine ( 2 mg/ml ) and 1 mM IPTG were added . The culture was further incubated at 37°C , without shaking . Over a period of 16 min , samples were withdrawn from the culture every 4 min and placed in liquid nitrogen . The samples were centrifuged at 14000 rpm for 10 min . The pellets were washed in 50 mM Tris pH 7 . 5 and then resuspended in lysis buffer ( 0 . 5 mg/ml lysozyme , 10 mM Tris pH 8 , 1 mM EDTA , 20 µg/ml DNase , 50 µg/ml RNase ) and 10% SDS . Lysates were incubated at 90°C for 5 min . These prepared lysates were loaded onto 10% SDS polyacrylamide gel [42] . To identify proteins synthesized after the induction of MazF , we used an autoradiogram of a 2D-gel analysis of a labeled , MazF-induced culture . We chose spots that corresponded to proteins whose level of synthesis was either not changed or even increased after MazF induction ( Figure 3A ) . Those selected proteins were extracted from a parallel 2D-gel of an unlabeled , MazF-induced , culture; the proteins were identified by mass-spectrometry ( MALDI-MS ) as described previously [45] . E . coli MC4100 relA+ and its derivatives were grown in M9 minimal medium at 37°C . After 12–16 hours of growth , they were diluted 1∶100 in M9 minimal medium and grown again at 37°C . When the cultures reached OD600 0 . 6 , 0 . 5 ml aliquots were taken from the cultures , put into Eppendorf tubes , and incubated , without shaking , at 37°C . After 10 min of incubation , mazEF dependent death was induced by the addition to each sample of either 2 mg/ml spectinomycin or 1 mg/ml nalidixic acid . After an additional 10 min of incubation , without shaking , at 37°C , the samples were centrifuged for at 14000 rpm for 5 min . After centrifugation , the supernatants were removed and the pellets were resuspended in 0 . 5 ml of pre-warmed saline . The samples were serially diluted in pre-warmed LB and plated on pre-warmed LB plates and incubated at 37°C . The percentage of survival was determined by dividing the number of colonies obtained from the “treated” sample by the number of colonies obtained from the “untreated” sample .
|
The enteric bacterium E . coli , as most other bacteria , carries a pair of genes on its chromosome; one of them specifies a toxin and the other one an antitoxin . Previously , we have shown that that the mazEF toxin–antitoxin system in E . coli is responsible for bacterial cell death under stressful conditions . Clearly , a system that causes any given cell to die is not advantageous to that particular cell . On the other hand , the death of an individual cell may be advantageous for the bacterial population as a whole . Here , for the first time , we report that MazF activates a complex network of proteins . Moreover , we also show , for the first time , that MazF affects two opposite processes: cell death and cell survival . We suggest that this dual effect may provide an evolutionary rational for mazEF-mediated cell death . When exposed to stressful conditions , most of the cell population undergoes programmed cell death; however , there appears to be an active process that keeps a small fraction of the population alive . When growth conditions become less stressful , it is probably this small sub-population of survivors that becomes the basis of a new cell population .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/gene",
"function",
"molecular",
"biology/translational",
"regulation",
"microbiology",
"microbiology/microbial",
"physiology",
"and",
"metabolism",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2009
|
Escherichia coli MazF Leads to the Simultaneous Selective Synthesis of Both “Death Proteins” and “Survival Proteins”
|
It is well documented that the density of Plasmodium in its vertebrate host modulates the physiological response induced; this in turn regulates parasite survival and transmission . It is less clear that parasite density in the mosquito regulates survival and transmission of this important pathogen . Numerous studies have described conversion rates of Plasmodium from one life stage to the next within the mosquito , yet few have considered that these rates might vary with parasite density . Here we establish infections with defined numbers of the rodent malaria parasite Plasmodium berghei to examine how parasite density at each stage of development ( gametocytes; ookinetes; oocysts and sporozoites ) influences development to the ensuing stage in Anopheles stephensi , and thus the delivery of infectious sporozoites to the vertebrate host . We show that every developmental transition exhibits strong density dependence , with numbers of the ensuing stages saturating at high density . We further show that when fed ookinetes at very low densities , oocyst development is facilitated by increasing ookinete number ( i . e . , the efficiency of ookinete–oocyst transformation follows a sigmoid relationship ) . We discuss how observations on this model system generate important hypotheses for the understanding of malaria biology , and how these might guide the rational analysis of interventions against the transmission of the malaria parasites of humans by their diverse vector species .
The availability of the genomes of man , the mosquito and the malarial parasite has enabled penetrating new studies on the molecular organization of Plasmodium in its two hosts . Yet , we do not fully comprehend how parasite population densities may affect transmissibility . Without this knowledge our understanding of the impact of host responses , or of external intervention , upon the transmission of the parasite through endemic populations will remain incomplete . Within the mosquito , Plasmodium transforms from macrogametocyte to ookinete , oocyst , and finally to sporozoite . The many studies reporting marked fluctuations in parasite numbers during this development have been elegantly summarized by Vaughan [1] . In susceptible mosquitoes it is the oocyst that frequently represents the nadir of parasite numbers in the life cycle ( Figure 1 ) . Within the oocyst , parasite numbers reportedly increase by two or three orders of magnitude [2 , 3] before the daughter sporozoites make their inefficient passage to the salivary glands . Few are inoculated into the vertebrate host when the female mosquito takes a subsequent bloodmeal [4–8] . In a number of well characterized parasite–mosquito combinations , ookinetes completely fail to cross the midgut epithelium or sporozoites fail to invade the salivary glands [9 , 10] . An important question that has eluded enquiry , however , is whether the ookinete-oocyst bottleneck and the other developmental transitions through the mosquito are density-dependent . ( In this context , a transition is density-dependent when the rate at which such process occurs is determined by the parasite density of the previous stage . ) Contributory factors for this omission may include the facts that most previous studies have compared parasite numbers in just two life stages , looked at a single infection intensity , or investigated densities within narrow ranges [11–13] . In this paper we have attempted to address this question by measuring the relationships between wide-ranging densities of successive life stages ( macrogametocytes , ookinetes , oocysts and salivary gland sporozoites ) achievable in the laboratory model Plasmodium berghei–Anopheles stephensi , and by statistically fitting functional forms to these relationships . Determining the form of these relationships may increase our understanding both of the processes regulating the transmission of this parasite by the mosquito , and ( by extrapolation ) of the potential impact of intervention measures . Although the biological bases for the parameters thus estimated can be inferred , we make no attempt to verify them experimentally . We recognize that they may include , among others: interspecific competition between the parasite and its vertebrate and invertebrate hosts ( the former being largely , though not exclusively , confined to the mosquito bloodmeal , and the latter mediated , for instance , by immune attack from the mosquito ) ; intraspecific competition between parasites; and the possible “altruistic” apoptotic death of the parasite [14] . The incorporation of the functional forms that we report in this paper into a mathematical model describing the transition of Plasmodium within the mosquito , and its linkage to malaria models that take into account parasite density in the human host [15] , will be presented elsewhere .
The fitting procedures investigated the frequency distribution among mosquitoes of the outcome variable ( parasite numbers of the ensuing stage ) ; fitted an appropriate distribution ( usually overdispersed ) ; explored the relationship between the degree of overdispersion and mean parasite density; and fitted models to both the means ( to allow comparisons to be made with the literature ) and to the individual parasite counts ( to allow maximum use of the data available ) for the relationships between two consecutive parasite stages . Density dependence can be positive ( facilitation ) or negative ( limitation ) , characterized by the per capita parasite yield increasing or decreasing , respectively , with parasite density . Initial facilitation may be followed by subsequent limitation , producing a sigmoid relationship . Absence of density dependence , where the rate of success is constant with parasite density is characterized by proportionality . To encompass all these possible behaviors , we fitted the following generalized formula , where x represents the input and y the output parasite density . This function can describe a linear ( α > 0; β = 1; γ = 0 ) ; a saturating ( α > 0; β = 1; γ > 0 ) ; or a sigmoid ( α > 0; β > 1; γ > 0 ) relationship , with each function being nested into the following one . The attributes of each of these functional forms are described in Table 1 . Data for this transition originated from two researchers on four separate occasions . The behavior of this transition was independent of the worker conducting the infections ( results not shown ) . Analysing the datasets jointly revealed that the overall distribution of the number of ookinetes per mosquito was strongly overdispersed , confirmed by a variance over mean ratio ( VMR ) of 1108 ( where a VMR of 1 suggests a Poisson , random distribution ) . A negative binomial distribution ( NBD ) was fitted to the overall frequency of mosquitoes harboring a given number of ookinetes; this revealed an arithmetic mean infection of 643 ookinetes per mosquito and an overdispersion parameter estimate of 0 . 45 ( 95% confidence interval ( CI ) : 0 . 02–3 . 59 ) ( Figure 2A ) . The chi-square goodness of fit test indicated good agreement between the observed and expected distributions ( χ2 = 13 . 2 , degrees of freedom ( df ) = 12 , p = 0 . 36 ) . When a separate NBD was fitted for each macrogametocyte density and its parameters estimated , the most overdispersed distributions were found among the lowest ookinete densities , and the degree of overdispersion decreased with increasing density ( Figure 2B ) as found in other parasite–vector systems [16] . Due to the overdispersion found in the data , the relationship between the number of macrogametocytes offered in the bloodmeal , and the resulting mean ookinete density , was investigated using the geometric mean of Williams ( WM ) [17] as a measure of central tendency for ookinete density . Excluding one outlying datapoint ( where exceptional parasite death occurred ) the best-fitting expression proved to be a saturating , hyperbolic function ( with α = 0 . 008 ( 0 . 006–0 . 009 ) and γ = 2 . 36 × 10−6 ( 1 . 09 × 10−6–4 . 27 × 10−6 ) ) , indicating that the efficiency of this transition declines monotonically with increasing macrogametocyte density ( Figure 2C ) . Initially ∼125 macrogametocytes are required to produce one ookinete and ookinete numbers saturate at a value ( given by α/γ ) of ∼3 , 500 per mosquito . A hyperbolic relationship between the number of ingested macrogametocytes and the resulting numbers of ookinetes produced , was again indicated when plotting the ookinete counts for individual mosquitoes ( with α = 0 . 013 ( 0 . 012–0 . 014 ) and γ = 3 . 57 × 10−6 ( 2 . 56 × 10−6–4 . 67 × 10−6 ) ; Figure 2D ) . This again suggests that the probability of a macrogametocyte becoming an ookinete declines with increasing input . This analysis predicts that initially ∼77 macrogametocytes make one ookinete , and this relationship again saturates at a density of ∼3 , 500 ookinetes per mosquito . Data for this transition came from three researchers and nine experiments . We recognize that the following analysis may be influenced by the atypical relationship that might exist between the pre-formed ookinetes in the bloodmeal and the midgut of the insect . As in the previous transition , oocyst numbers were overdispersed ( VMR = 63 ) . An NBD was fitted to the combined datasets , and to the oocyst frequency distributions for each ookinete input density . However in this case , the NBD was not the most appropriate to describe the frequency of mosquitoes with a given oocyst density ( χ2 = 158 . 8 , df = 29 , p < 0 . 001 ) ( unlike [18] ) ; this may be because our ookinete feed technique ( unlike the gametocyte feeds used in [18] ) , rarely fails to infect mosquitoes . As in the previous transition , the degree of overdispersion was maximal at low oocyst mean densities , and decreased with increasing density . The fitted relationship between input ookinete density and the resulting WM oocyst number ( Figure 3A ) indicated a sigmoid relationship , suggesting the operation of initial facilitation and subsequent limitation ( with α = 6 . 87 × 10−5 ( 5 . 28 × 10−5–8 . 85 × 10−5 ) ; β = 1 . 98 ( 1 . 89–2 . 11 ) and γ = 2 . 64 × 10−6 ( 1 . 88 × 10−6–3 . 81 × 10−6 ) ) . Initially , 120 ookinetes produce one oocyst; at the point of maximum yield ∼60 ookinetes produce one oocyst; and as the number of ookinetes increases further , the transformation becomes constrained above a mean density of ∼26 oocysts per mosquito . As with the relationship identified for the means , the best fitting model for the number of oocysts per individual mosquito as a function of ookinete density ( Figure 3B ) , was sigmoid ( with parameters α = 5 . 84 × 10−4 ( 5 . 55 × 10−4 , 6 . 12 × 10−4 ) ; β = 1 . 74 ( 1 . 73 , 1 . 76 ) ; γ = 7 . 44 × 10−6 ( 6 . 89 × 10−6 , 8 . 11 × 10−6 ) ) . Initially ∼72 ookinetes are required to produce one oocyst; at the point of maximum yield this requirement decreases to ∼35 ookinetes per oocyst; and the relationship saturates at a maximum of ∼80 oocysts per mosquito as the number of ookinetes increases . When testing inter-experimenter variation it was clear that even in the same laboratory , oocyst production from ookinete membrane-feeds varied markedly between different researchers . Nonetheless each experimenter's data invariably indicated that , despite the differing overall efficiency of oocyst production , every relationship was sigmoid ( Figure 3B , Table 2 ) . Recognizing that the transformation of the ookinete into an oocyst is dependent not only on the ookinete locating and invading the midgut wall , but also on the ability of the intracellular ookinete to survive attack by the mosquito's innate immune mechanisms , we have , in replicate studies , examined directly the ability of green fluorescent protein ( GFP ) -expressing ookinetes to invade the gut cells . Following the ingestion of 3 , 600 GFP-tagged ookinetes , a mean of 516 ( or 14% ) were detected 24 hours later in the midgut epithelium . This number of ookinetes would be expected to produce 80 oocysts ( Figure 3B ) and thus at this infection intensity just 16% ( 80/516 ) of the ookinetes in the midgut epithelium are predicted to be detected as oocysts 9 days later . Data for this transition came from three independent experiments by one researcher . The number of sporozoites per mosquito showed strong overdispersion ( VMR = 1 , 393 ) . The distributions resulting from each oocyst density were separately examined for an NBD , and again , overdispersion decreased with increasing sporozoite density . However , as for the frequency distribution of oocysts , the NBD did not fit the distribution of the number of salivary gland sporozoites per mosquito ( χ2 = 41 . 3 , df = 19 , p < 0 . 002 ) as well as it did for the ookinete distribution . The relationship between the output WM number of salivary gland sporozoites per mosquito and the input WM oocyst density per mosquito ( Figure 4A ) was most parsimoniously fitted by a linear model ( with α = 14 . 61 ( 11 . 58–17 . 63 ) ) based on the likelihood ratio statistic ( LRS ) analyses , although this model was only marginally better than the hyperbolic fit . ( The latter was suggested as the better model according to the Akaike Information Criterion ( AIC ) and this discrepancy may indicate insufficient power to distinguish between the two models . ) This relationship predicts that in this study on average , each oocyst produces between just 12 and 18 sporozoites that successfully invade the salivary glands . This relationship became significantly nonlinear when analyzing the number of salivary gland sporozoites per individual mosquito as a function of the mean oocyst density ( Figure 4B ) . In this case , the best fitting model was hyperbolic ( with α = 62 . 21 ( 53 . 98–72 . 20 ) and γ = 0 . 018 ( 0 . 013–0 . 024 ) ) , indicating that initially 54–72 salivary gland sporozoites are produced per oocyst , and thereafter the efficiency of conversion declines with rising oocyst density , reaching a plateau of ∼3 , 500 salivary gland sporozoites per mosquito irrespective of rising oocyst input .
We recognize that one model system cannot accurately reflect the diverse biology of the hundreds of natural malaria–vector combinations found worldwide; nonetheless we also recognize that studies on the biology of Plasmodium spp . , and their interactions with Anopheles spp . , have been advanced considerably by the analysis of malaria parasites of rodents . Exploiting the rare opportunity to study , in a controlled environment , cloned populations of P . berghei in an inbred line of An . stephensi at widely differing parasite densities has permitted us to raise questions in the model that are orders of magnitude more difficult to study in the parasites of man . Under these specific conditions , we show that parasite's developmental transitions within the mosquito are density-dependent . These conclusions are consistent with limited ( mainly laboratory based ) studies on the malaria parasites of humans ( see below ) . They help generate key hypotheses that now require to be tested on other parasite–vector combinations . We recognize that future laboratory studies on , for example , P . falciparum–An . gambiae , would have to conduct experiments with lab adapted strains of both parasite and vector; not only are these experiments more costly ( in addition to requiring extensive safety management ) , but also the lab strains themselves lack the diversity of natural populations . We recognise therefore that the key studies will be those on the parasites and vectors in their numerous and different endemic areas . If validated in the human malarias , the hypotheses generated in this study may have important implications for the design of anti-malarial intervention programs . To our knowledge , this is the first quantitative investigation of the impact of parasite density upon all transitions between Plasmodium stages within the mosquito . Our data have permitted analyses not only of measures of central tendency in groups of mosquitoes ( the most frequent type of ( aggregate ) analysis reported in the literature ) , but also of parasite counts in individual mosquitoes , and their distribution . Gametocyte numbers in the bloodmeal are variously described as being normally distributed [1 , 19] or overdispersed [20 , 21]; ookinete numbers in the ingested bloodmeal as being normally distributed [1] or overdispersed [19 , 22]; and oocysts and salivary gland sporozoites as being markedly overdispersed [6 , 18 , 19 , 22 , 23] . Recognising that the NBD can originate , for instance , when each host /vector is infected according to a Poisson process whose mean is gamma-distributed ( i . e . , there is marked heterogeneity in mosquito susceptibility [18] ) , we were interested to find that even using highly “inbred” organisms , ookinete , oocyst , and gland sporozoite numbers per mosquito exhibit strong overdispersion ( see also [24 , 25] ) . As previously observed [16 , 18] , the severity of overdispersion was itself density-dependent , decreasing with mean parasite density both when the relationship between overdispersion and parasite load was analyzed separately ( having fitted distributions to each parasite density previously ) , and when the degree of overdispersion was allowed to vary with parasite density whilst jointly fitting models to individual parasite counts . In designing these experiments we were cognizant of early studies which recognized that infectivity of individual gametocyte carriers differs widely over the course of an infection [26–28] . Whereas P . berghei ookinete production in vitro faithfully reflected gametocyte density in the blood , oocyst formation in the mosquito host was severely compromised after day 5 of the blood infection [27] . Whereas some prior studies have concluded that gametocyte-oocyst intensities are linearly related [11 , 12] , others noted a nonlinear relationship in P . falciparum within , but not between experiments [29] . Ponnudurai et al . [29] and Pichon et al . [30] reported the operation of density-dependent suppression of oocyst development in P . falciparum; we have similarly found that the gametocyte-oocyst transformation is density-dependent . Notwithstanding the operation of nonlinearities , the efficiency of this conversion in our experimental P . berghei–An . stephensi system tends to be lower than that estimated in the natural combinations P . gallinaceum–Aedes aegypti [11] and P . falciparum–An . gambiae [12] ( results not shown ) . The numerous blood-borne variables responsible for modulating gametocyte-oocyst development in vivo transcend species , and many suppress the early conversion of gametocytes to ookinetes [31–40] . Mosquito factors regulating infection are less well understood but include physiological , immunologic and biotic variables [41–44] , the expression of which varies with both mosquito and parasite species [45 , 46] , as well as with genotypes [31 , 47] . Interestingly , it has been suggested that Plasmodium may have immunosuppressive effects upon the vector [48] . Consistent with the above , recent analyses on P . vivax concluded that it is vertebrate factors that impact largely upon fertilization , whereas mosquito factors determine ookinete losses [49] . Unlike these previous studies which investigated the relationship between gametocytes and oocysts ( and found or not some evidence of nonlinearity ) , our work has aimed at teasing out where exactly nonlinearities may be occurring , and therefore we have examined the transition from gametocytes to ookinetes separately from that of ookinetes to oocysts . In P . falciparum the efficiency of macrogametocyte to ookinete conversion in vivo reportedly varies widely ( from 0 . 025% to 42% ) [50] . In P . berghei , we find that the efficiency of conversion for the transition of macrogametocytes to ookinetes is maximal at the lowest parasite densities ( i . e . , the hyperbolic models were found to best fit this relationship ) . At the lowest macrogametocyte densities investigated ( i . e . , ∼7 , 000 per bloodmeal ) the efficiency of ookinete production in vivo is ∼1 . 3%; thereafter the efficiency falls progressively with increasing density ( 0 . 6% for 360 , 000 macrogametocyes/bloodmeal ) . This fall suggests competition for limited resources , or density-dependent stimulation of a parasite-killing response in the mosquito [44 , 51 , 52] . Interestingly , when examining ookinete to oocyst development we found a sigmoid relationship , indicating that at low ookinete densities , and with this experimental design , the transition was positively density-dependent ( i . e . , the per capita probability of successful oocyst establishment was very low at the lowest ookinete densities but increased initially with increasing ookinete density in the bloodmeal ) . Similarly , a sigmoid zygote-oocyst relationship was reported by Rosenberg et al . when P . gallinaceum female zygotes produced in vitro were membrane-fed to Ae . aegypti [53] . The maximal parasite yield ( 100% ) was achieved for zygote densities of ∼4/mosquito , but at ∼40 , 000/mosquito the per zygote efficiency had decreased to 0 . 3% . A possible biological explanation for this initial facilitation may lie in the difficulty that individual ookinetes may have to disrupt the mosquito's peritrophic matrix or midgut epithelial cells , whereas at higher ookinete densities , those which succeed in penetrating these structures may make it easier for other ookinetes to do so , facilitating their passage and ultimately their establishment under the basal lamina of the midgut epithelium . These results are entirely consistent with an earlier study by Munderloh and Kurtti [54] , who observed that low numbers of P . berghei ookinetes do not reliably produce oocyst infections . Assuming a bloodmeal volume of 2 . 13 μl [19] , our data suggest that to ensure infection with at least one oocyst , approximately 40–140 ookinetes/mosquito are required ( Table 2 ) , whereas they estimated that ∼1 , 100 ( purified ) ookinetes would be necessary . We attribute these differences in number to recent improvements in the parasite culture . Studies on P . falciparum in vivo [19] similarly indicate an apparent “threshold” ookinete density of 30/mosquito in An . gambiae . Sigmoid relationships have a “turning” parasite density at which the maximum probability of transformation is achieved and beyond which negative density dependence operates . The turning-point ookinete density ( assuming β = 2 for simplicity ) is 355 for the model fitted to mean P . berghei oocyst density ( Figure 3A ) , and 212 for the model fitted to all individual data combined ( black curve in Figure 3B; see Table 2 for researcher-specific turning points ) . Other examples of initial facilitation followed by subsequent limitation within a vector have been found among filarial parasites [16] . Earlier studies [50 , 55–57] remarked on the high cost of ookinete-oocyst transformation , and suggest it to be a critical block . Published work suggests that at this stage of development , mosquitoes can be significantly ( P . falciparum / An . albimanus [19]; P . yoelii / An . albimanus [22 , 58] ) or totally refractory ( P . berghei / Ae . Aegypti [55] ) . However , in compatible parasite–vector combinations , e . g . , P . falciparum / An . freeborni [19] and P . yoelii / An . stephensi [22] , transformation can be efficient and parasite losses at the late ookinete stage rare . It will be interesting to explore whether the specificity and/or magnitude of the mosquito's immune responses are sensitive to ookinete density . It is known that immune responses are qualitatively different in different parasite–vector combinations [59] , and their cost to vector fitness and survival is not insignificant [48 , 60–62] . The transition from oocyst to salivary gland sporozoite is usually inefficient and can be totally inhibited in some Plasmodium/mosquito combinations [3]; whether sporozoites released from the oocyst are removed from the hemocele by hemocytes , or lysed by immune peptides is still unknown [63] . Salivary gland burden is markedly increased if oocyst-infected mosquitoes take a second , uninfected bloodmeal , typically 4 days after infection [64] . This may overcome inter-oocyst competition for nutrients , but it has been suggested that it synchronizes sporozoite maturation [29 , 65] . Whilst some authors suggest that mosquito survivorship is not adversely impacted by oocyst density [10 , 66] , others found that mosquito mortality increased with oocyst burden [67 , 68] . Our experimental design was not influenced by these variables; the experiments were all subjected to the same ( single ) blood-feed regimen , and all examined mosquitoes were alive immediately prior to dissection . The mathematical relationship between abdominal oocysts and salivary gland sporozoites is complicated by the variable and unknown sporozoite production within oocysts [64 , 69 , 70] . Previous studies have reported that the number of sporozoites produced per oocyst is not density-dependent [22] , and studies on P . vivax [13 , 66] reported linear relationships between oocyst and salivary gland sporozoite prevalences . Others have described only a weak ( Pearson ) correlation between oocyst , and sporozoite number in the glands in both P . falciparum , and P . vivax [10] . Geometric mean oocyst loads of 2 . 6 ( 1–197 ) and 2 . 2 ( 1–26 ) in , respectively , naturally infected An . gambiae and An . funestus have been correlated with mean loads of 962 and 812 salivary gland sporozoites . Assuming a production of ∼10 , 000 sporozoites per oocyst [2] , this would translate in only 4% of sporozoites invading the salivary glands [23] . Our data suggest that mean oocyst numbers are linearly related to mean salivary gland sporozoite load ( Figure 4A , r = 0 . 9 ) , but that a saturating ( hyperbolic ) relationship best describes the sporozoite numbers per individual mosquito ( Figure 4B ) . At most naturally occurring oocyst burdens ( 1–5 per mosquito ) , our results suggest that the number of sporozoites in the glands is most likely to be directly proportional to oocyst load , as found with P . falciparum in An . gambiae , however the latter combination exhibits a much higher efficiency ( i . e . , ∼400–700 P . falciparum gland sporozoites/oocyst [23 , 50] versus 54–72 P . berghei gland sporozoites/oocyst ) . However at the high oocyst numbers ( >50 ) achievable in P . berghei / An . stephensi , it is evident that salivary gland numbers are rate-limited ( Figure 4B ) . Previous authors have discussed , for P . berghei , P . yoelii , and P . falciparum [2 , 5–9 , 71–74] , the importance of sporozoites being located in the salivary gland ducts at the time of feeding in relation to the probability of a sporozoite being inoculated into the skin of the host . Figure 5 illustrates the enormous variation reported in sporozoite inocula in the bites of An . stephensi with wide-ranging salivary gland burdens of sporozoites of the rodent malarias . Whilst it is widely conjectured that sporozoites of different Plasmodium spp . differ dramatically in their infectivity to their vertebrate hosts , published and unpublished data kindly made available to us ( Table 3 ) suggest that for all species studied , inocula of just 10 sporozoites can be infectious . It is therefore relevant to investigate the question “At what salivary gland sporozoite density , will the number of sporozoites in the bite fall below 10 ? ” Whilst the data suggest gland burdens as high as 30 , 000 can result in inocula below this “threshold” , it is clear that gland infections of just a few hundred sporozoites ( that could be derived from 1–2 oocysts in the P . falciparum / An . gambiae combination ) have clear infection potential . It is thus obvious that the prevalence and not the intensity of oocysts or salivary gland sporozoite infections will be the key practical parameter when considering the potential infectivity of individual mosquitoes to the vertebrate host . Our overarching goal is to develop mathematical models of the population biology of malaria within the mosquito that ultimately relate the prevalence and intensity of gametocytaemia in the vertebrate host with the entomological inoculation rate and the force of infection , and to link these frameworks with novel models of malaria in the human host [15] . Thus far , these studies have found little relationship between the infectiousness of human populations to vectors and the resulting transmission intensity from vectors back to humans . The very poor correlation between salivary gland burden and sporozoite inoculum at the next bite , suggests that reductions in oocyst number may not correlate well with the potential impact of intervention upon transmission . In contrast , it is abundantly clear that any effective transmission-blocking strategy will have to reduce both oocyst intensity and prevalence ( their inter-relationship being nonlinear [18] ) . The biological consequences of reductions in prevalence cannot be contested —uninfected mosquitoes cannot transmit . The varying reduction in prevalence that would be induced , by a 90% reduction in oocyst intensity , at different initial oocyst densities is illustrated in Figure 6 . If studies into transmission-blocking strategies were to discuss their efficacy in terms that unequivocally reduce the number of infectious bites ( and therefore the force of infection ) , as argued by those applying anti-vector policies , they might facilitate the wider understanding and acceptance of the obvious impact of such interventions in endemic communities . We will report elsewhere how our data will permit us to forward hypotheses as to how interventions targeted at different mosquito stages , e . g . , gametes , ookinetes , oocysts or sporozoites , might be expected to reduce the prevalence of infectious mosquitoes in a theoretical population .
To eliminate the impact of host and parasite genetic variability , parasite clones , and one inbred mosquito line were used . All but one experiment reported here used clone 234 of the rodent malarial parasite P . berghei strain ANKA . The parasite was maintained by serial passage , but no more than eight sequential mechanical blood passages took place before passage through mosquitoes [27 , 33] . This regimen maintains gametocyte infectivity to the mosquito , a critical property for this study . Gametocyte density and male: female ratios were determined in Giemsa stained smears , and infections were always done on days 3–5 when a low but rising gametocytaemia prevailed [27 , 33] . Male:female ratios invariably fell within the normal range for low-passage P . berghei infections , i . e . , 1 . 64 ± 0 . 93SD ( Dearsly , unpublished ) . All details of direct , or membrane feeds and parasite enumeration are as described previously [75] . Those experiments in which there was significant insect mortality were excluded , on the understanding that this phenomenon may itself be related to infection intensity [76] . In an effort to facilitate parasite identification , one set of experiments used the GFP-expressing transgenic clone ( PbGFPCON ) [77] derived from the HP line of P . berghei . Experiments counting ookinetes in the bloodmeal used the indirect fluorescent antibody test ( IFAT ) to reveal parasites expressing P28/Pbs21 [78] . In all calculations it has been assumed that the bloodmeal volume in A . stephensi is 2 . 13 μl [19] . All gametocytes were raised in Theiler's Original ( TO ) mice and transmitted to A . stephensi strain Sd 500 , maintained at 19 °C and 80% RH , and fed on 5% fructose/0 . 05% para-amino-benzoic acid as described previously [75] . The impact of numerous extraneous factors ( e . g . , host serum; mosquito midgut milieu ) upon gametocyte infectivity complicates the design and reproducibility of experiments in P . berghei [34 , 27] . Pools of gametocytes were therefore washed free of variable serum factors and re-suspended in a single pool of serum known to support P . berghei infectivity . Gametocytes , re-suspended at known parasite densities , were fed to replicate aliquots of A . stephensi from the same rearing brood . Fifteen hours later , GFP-expressing , or P28-positive retorts and ookinetes were counted by IFAT in dissected bloodmeals . Ten days after blood feeding , oocysts were counted by phase contrast microscopy of freshly dissected midguts . Sporozoites were similarly counted on days 21–25 by hemocytometry of dissected salivary glands . Unless otherwise stated , all experiments were repeated in triplicate and all mosquitoes examined were alive immediately prior to dissection . We have not attempted to study oocyst infections below a mean of 10 as the proportion of uninfected mosquitoes rises rapidly making the necessary group sizes unmanageably large [18] . To study the impact of ookinete density upon oocyst formation we exploited our ability to culture P . berghei ookinetes in vitro . Following culture ookinetes were resuspended at known densities in fresh heparinised mouse blood and fed by membrane feeder [75] . Under these conditions it is anticipated that some ookinetes will invade the midgut epithelium earlier than when mosquitoes are infected by gametocytes . Even with the advent of GFP-tagged parasite lines it is not possible to count midgut oocysts in vivo without compromising the subsequent development of the sporozoites in the mosquito , due to direct UV irradiation , or stress/damage to the insect . Thus , batches in excess of 200 mosquitoes were each infected with different ookinete numbers such that different oocyst burdens could be determined by dissecting groups of at least 50 mosquitoes [79] . The remainder of each mosquito batch ( >50 ) was incubated a further 8–15 days before counting the salivary gland sporozoites . With this design , both input and output parasite densities are random variables . Efforts to enumerate sporozoites in the saliva secreted into a bloodmeal , using PCR , were not successful , and have been superseded by the elegant studies of others [5 , 6 , 72] . However , for the sake of completeness , we give an account of our procedures in Text S1 . Frequency distributions . Frequency distribution histograms were prepared for each of the output life stages of interest , i . e . , the number of ookinetes , the number of oocysts , and the number of salivary gland sporozoites per mosquito . The VMR was calculated for each distribution , which indicated the presence of overdispersion ( the ratio was significantly greater than 1 as ascertained by chi-square tests [80] ) . An NBD ( whose parameters are the arithmetic mean , and the exponent , k , an inverse measure of the degree of overdispersion ) was fitted to the data using maximum likelihood ( ML ) . The goodness of fit of the negative binomial distribution to the observed frequency distribution was assessed using chi-square tests [80] . For each life stage , values for k were thus estimated for the combined data , and for each input parasite density separately , and plotted against the corresponding arithmetic mean number of parasites to assess whether the overdispersion parameter was independent or related to the means of the distributions . The functional form of the relationship between k and the mean was determined by fitting various models ( detailed in Text S2 ) using ML , which were compared using the LRS [81] for nested models and the AIC [82] for non-nested models . The LRS and AIC results can be found in Table S1 , and the resulting parameter values in Table S2 . Models fitted to mean parasite densities . Functional relationships were fitted to mean parasite densities as a function of the ( pre-defined ) previous life stage density , using the WM [17] as a measure of central tendency for the outcome variable ( given the overdispersed nature of the data as assessed above ) . Asymmetric CIs were estimated for the WMs [83] . When the explanatory variable was also the mean of a random variable ( rather than a pre-determined parasite density fed to mosquitoes ) , CIs were also calculated for the variable plotted on the horizontal axis . Heterogeneity in the data between experiments and between workers was examined to determine whether a single model should be fitted to the entire dataset , to individual experiments , or to experimenter subsets . The form of the relationship between two subsequent life stages , using the mean values , was determined by fitting Equation 1 by ML using quasi-Newton algorithms . We maximized a sample size-weighted log-likelihood for a normally distributed variable after applying a suitable ( square root ) transformation of the means . The weighting procedure allowed us to take into account the number of mosquitoes contributing to each mean value in addition to the total number of mean values [84] . In this case , y was the WM of the outcome variable and x was the parasite density of the previous ( input ) life stage . Parameters α , β , and γ determine the shape of the functional form as described in Table 1 , and the resulting models were compared using the LRS [81] as the models were nested ( linear versus hyperbolic and versus sigmoid; hyperbolic versus sigmoid ) . ( Details of the LRS results are provided in Table S3 . ) Asymptotic 95% CIs [85] were estimated for each parameter in the final model . Models fitted to individual parasite densities . In order to make better use of the data available , models were also fitted to individual parasite densities rather than the mean values , using parasite densities from individual mosquitoes against the pre-defined ( or mean ) parasite densities of the previous life stage . The form of the relationship between two subsequent life stages using the individual values was determined by fitting the functional form given in Equation 1 , where x , α , β , and γ , were as described above , and y was the parasite count of the subsequent stage as observed in individual mosquitoes . Parameters were estimated , using quasi-Newton algorithms , by maximising a negative binomial log-likelihood that allowed the overdispersion parameter to be a function of the mean [86] . Again , models were compared using LRS , and asymptotic 95% CIs [85] were estimated for the parameters in the final model , including overdispersion parameters . Details of model comparison can be found in Table S4 . In all cases p < 0 . 05 was considered to indicate a significant departure from the null hypothesis in question . Limitations of the statistical analyses are discussed in Text S3 .
|
Malaria , one of the world's most devastating parasitic diseases , is caused by protozoan parasites of the genus Plasmodium and is transmitted between mammalian hosts by Anopheles mosquitoes . Within the mosquito , the parasite undergoes four sequential developmental transformations as it passes from the bloodmeal through the mosquito's midgut epithelium to the salivary glands , from where the parasite is inoculated when the mosquito bites the vertebrate host . This study demonstrates , in a laboratory model , that parasite input density at every developmental stage in the mosquito regulates output to the ensuing form . Statistical models were fitted to experimental data to identify and describe the most appropriate functional relationships . In all cases , the relationships between two consecutive parasite stages can saturate at high parasite densities , suggesting that at high parasite densities parasite numbers may have to be reduced substantially to effect an appreciable decrease in parasite transmission . These results may help establish a rational basis for new studies on species of medical importance and further our understanding of how interventions designed to reduce parasite survival within the mosquito might be expected to impact upon transmission .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"mosquito",
"infectious",
"diseases",
"malaria",
"dynamics"
] |
2007
|
Progression of Plasmodium berghei through Anopheles stephensi Is Density-Dependent
|
Pulmonary cavities , the hallmark of tuberculosis ( TB ) , are characterized by high mycobacterial load and perpetuate the spread of M . tuberculosis . The mechanism of matrix destruction resulting in cavitation is not well defined . Neutrophils are emerging as key mediators of TB immunopathology and their influx are associated with poor outcomes . We investigated neutrophil-dependent mechanisms involved in TB-associated matrix destruction using a cellular model , a cohort of 108 patients , and in separate patient lung biopsies . Neutrophil-derived NF-kB-dependent matrix metalloproteinase-8 ( MMP-8 ) secretion was up-regulated in TB and caused matrix destruction both in vitro and in respiratory samples of TB patients . Collagen destruction induced by TB infection was abolished by doxycycline , a licensed MMP inhibitor . Neutrophil extracellular traps ( NETs ) contain MMP-8 and are increased in samples from TB patients . Neutrophils lined the circumference of human pulmonary TB cavities and sputum MMP-8 concentrations reflected TB radiological and clinical disease severity . AMPK , a central regulator of catabolism , drove neutrophil MMP-8 secretion and neutrophils from AMPK-deficient patients secrete lower MMP-8 concentrations . AMPK-expressing neutrophils are present in human TB lung biopsies with phospho-AMPK detected in nuclei . These data demonstrate that neutrophil-derived MMP-8 has a key role in the immunopathology of TB and is a potential target for host-directed therapy in this infectious disease .
The lung cavity is a hallmark of pulmonary tuberculosis , a globally important disease of man . The cavity has high bacillary burden and is associated with spread of infection . Polymorphonuclear leukocytes or neutrophils are abundant in areas of TB lung cavities [1] . Excessive neutrophil recruitment associates with pathology in animal models [2 , 3] and in man [4] but the mechanism of how neutrophils drive pathology in human TB is not defined . Zinc-containing matrix metalloproteinases ( MMPs ) have key roles in the inflammatory immunopathology in a wide range of diseases including cancer and arthritis [5 , 6] . On the basis of diverse evidence , it has been shown that a matrix-degrading phenotype develops in TB in which MMP activity is relatively unopposed by the specific tissue inhibitors of metalloproteinases ( TIMPs ) [7] . MMPs are crucial in granuloma formation in the zebrafish model of TB [8] and may drive different stages of lung pathology . Collagenases , a subgroup of the MMPs , are key in TB pathology since collagen is the main structural protein of the lung , the primary site of infection . Patients with pulmonary TB have increased collagenases which correlate significantly with radiological markers of tissue destruction [9 , 10] . Neutrophils secrete MMP-8 , a potent collagenase , and increased neutrophil-derived MMPs associate with disease severity in CNS-TB [11 , 12] , implicating neutrophils in the immunopathology of human TB . The concept of metabolism regulating host immunity is only recently emerging [13] . Adenosine monophosphate-activated protein kinase ( AMPK ) , a serine/threonine kinase is a central regulator of metabolic responses acting as an activator of cellular catabolism [13] . In addition , AMPK is known to have a role in immune responses determining the effector versus memory fate of CD8 T-cells [14] . Inhibition of glucose uptake and AMPK inhibition impedes T cell chemotaxis [15] . Dissecting the mechanism of how metabolism regulates immunity may be key to understanding immunity in chronic infections such as TB . We hypothesize that neutrophils drive tissue destruction in human pulmonary TB . Animal models of infection such as murine strains which develop pulmonary necrosis and cavities are useful in dissecting areas of the immune response in TB [16–18] . Murine models also demonstrated the critical importance of IFN-γ and TNF-α in the host defence against TB [19–21] . However , there are inherent differences between murine and human neutrophils with divergences in cytokine secretion [22] , peptides such as defensins [23] , and intracellular signalling pathways [24] . Therefore , in this study , we investigate the role of the neutrophil in MMP-dependent tissue destruction in human pulmonary TB , a disease that affects man as the primary host , and examine the signaling pathways regulating this process . First , we investigate the effect of neutrophil-derived collagenase MMP-8 in a human cellular model and examine MMP-8 expression and collagenolytic activity in patients . Neutrophils secrete MMP-8 on direct M . tb infection and in M . tb-infected monocyte-dependent networks . Neutrophil MMP-8 is expressed in TB patients’ biopsy specimens , with the secretion of MMP-8 dependent on NF-kB . We found in a cohort of 108 TB patients and controls that increased MMP-8 is closely associated with neutrophil markers and correlates with radiological and clinical disease severity . Sputum MMP-8 from TB patients is functionally active , causing matrix destruction , and patients with pulmonary cavities on chest radiographs have higher MMP-8 concentrations in their respiratory secretions . Gelatin degradation in the respiratory samples is raised , but is not dependent on neutrophil gelatinase MMP-9 . We demonstrate that AMPK regulates MMP-8 dependent tissue destruction , both at the level of protein secretion and gene expression , using data from a cellular model of infection and by investigating biopsy samples from TB patients and immune responses in AMPK deficient patients . Taken together , our data demonstrate that neutrophils cause tissue destruction in TB by an MMP-8-dependent process , regulated by the pro-catabolic AMPK pathway .
First , we investigated MMP-8 secretion from primary human neutrophils stimulated by live , virulent M . tb . Neutrophil MMP-8 secretion increased over time and in a dose-dependent manner in response to higher M . tb multiplicity of infection ( MOI ) ( Fig 1A and 1B ) . TIMP-1 and -2 are the MMP inhibitors secreted by neutrophils [25 , 26] . TIMP-1 was not secreted in response to stimulation by M . tb ( S1A Fig ) . TIMP-2 concentrations increased significantly but to a 20-fold lower concentration than MMP-8 ( S1B Fig ) . We demonstrated that neutrophil MMP-8 secretion was blocked by the NF-kB p65 subunit inhibitor Helenalin ( IC50 10–50 μM ) , in a dose-dependent manner and the effect was maximal at 100 μM ( S1C Fig , P<0 . 001 ) . The dose-dependent suppression of neutrophil MMP-8 was replicated with additional specific NF-kB inhibitors caffeic acid phenethyl ester ( CAPE ) [27] ( S1D Fig ) and SN50 [28] ( S1E Fig ) . Neutrophil viability was greater than 95% for all conditions by FACS staining with Annexin V and live/dead dye ( S2 Fig ) . To determine the cellular source of MMP-8 in patients with TB , we analyzed lung biopsies from patients who were diagnosed with pulmonary TB . Polymorphonuclear neutrophils were observed along the entire circumference of the inner wall of cavities on H & E staining ( Fig 1C ) . Neutrophil accumulation was confirmed by specific positive staining for neutrophil elastase ( Fig 1D ) . Neutrophils in the same location stained positive for MMP-8 ( Fig 1E and 1F ) . MMP-8 expression was also found in the central area of necrosis of granulomas ( S3 Fig ) , suggesting that MMP-8 may be associated with the process of necrosis . To determine if other proteases from neutrophils were similarly up-regulated , we analyzed MMP-9 ( neutrophil gelatinase ) secretion from M . tb infected neutrophils . M . tb caused a dose-dependent increase of MMP-9 secretion ( Fig 1G ) . Furthermore , MMP-9 staining of patient lung biopsy specimens also showed presence of MMP-9 in neutrophils ( Fig 1H ) . In addition to neutrophils , monocytes are among the early cells to be recruited in M . tb infection [29] and substantial cross-talk may occur between neutrophils and monocytes [30] . Using conditioned media from monocytes infected by M . tb ( CoMTB ) to model intercellular stimulation of neutrophils , we found significant up-regulation of MMP-8 secretion similar to M . tb infection ( Fig 2A ) . We assessed the functional consequences of MMP-8 activity on degradation of Type I collagen , the main extracellular matrix fibril providing structural support in human lung parenchyma [31] . Both M . tb-infected and CoMTB-stimulated neutrophils degrade DQ collagen as assessed by a quantitative fluorescence assay ( Fig 2B ) . Confocal microscopy demonstrated collagen degradation at the neutrophil-collagen interface both in M . tb-infected and CoMTB-stimulated neutrophils ( Fig 2C and 2D ) . There was dose-dependent inhibition of collagenase activity to baseline when neutrophil supernatants were treated with doxycycline , an MMP inhibitor licensed in the USA for use in periodontal disease [32] . The effect was maximal after treatment with 100 μM doxycyline ( Fig 2E , P<0 . 001 ) . Next we showed that neutrophils generate NETs when stimulated with M . tb in vitro ( S4A Fig ) , and NETs were digested by DNAse ( S4B and S4C Fig ) . Neutrophil extracellular traps ( NETs ) are scaffolds containing DNA , histones and antimicrobial granule proteins . We demonstrated for the first time that MMP-8 co-localizes with NETs ( Fig 3A and 3B ) . Next , we evaluated NETs in induced sputum from TB patients and healthy controls from a clinical study [33] ( S1 Table ) . Sputum from TB patients had increased NET concentrations of 1548 mg/ml ( ± standard error 256 mg/ml ) compared to controls at 372 mg/ml ( ± S . E . 150mg/ml ) ( Fig 3C , P<0 . 001 ) . Citrulline H3 , an established marker of NETs [34 , 35] was detected in induced sputum of TB patients and not in healthy controls ( Fig 3D ) . This was not due to dead or dying cells since these neutrophils do not contain citrulline H3 ( S4D Fig ) . MMP-8 is substantially elevated in the induced sputum of TB patients compared to other MMPs [33] . To determine if MMP-8 was neutrophil derived , we analyzed two established markers of neutrophil activation , myeloperoxidase ( MPO ) and neutrophil gelatinase associated lipocalin ( NGAL ) [36 , 37] in induced sputum . In a cohort of 51 TB patients and 57 healthy controls randomly selected from our previously reported study of 137 patients [33] , MPO and NGAL concentrations were increased in induced sputum of TB patients compared to controls ( Fig 4A and 4B ) . Both sputum MPO and NGAL concentrations correlated strongly with MMP-8 ( r = 0 . 83 , P<0 . 0001 and r = 0 . 68 , P<0 . 0001 respectively ) ( Fig 4C and 4D ) , indicating that MMP-8 in induced sputum of TB patients is likely to be principally derived from neutrophils . Next , we demonstrated that induced sputum from TB patients had increased collagenase activity compared to healthy controls using the DQ collagen degradation assay ( Fig 4E , P = 0 . 02 ) , confirmed on confocal microscopy ( Fig 4F ) . Sputum MMP-8 concentrations strongly correlated with collagenase activity ( r = 0 . 7 , P = 0 . 0004 ) ( Fig 4G ) and MMP-8 neutralization decreased collagenase activity in respiratory secretions of TB patients ( P = 0 . 01 ) ( Fig 4H ) . When the cohort was stratified according to the presence or absence of lung cavities , patients with pulmonary cavitation secreted a median of 5-fold higher MMP-8 concentration than those without cavities . ( P<0 . 028 , Fig 4I ) . In addition , MMP-8 sputum concentrations positively correlated with the TB score ( r = 0 . 56 for n = 108; P<0 . 0001 ) , a clinical marker of disease severity . The other major neutrophil-derived MMP , MMP-9 , had a much weaker although statistically significant correlation with TB score ( r = 0 . 3453 for n = 108; P = 0 . 003 ) . Analyzing CXR consolidation score as a radiological marker of tissue destruction demonstrated a similar strong MMP-8 correlation ( r = 0 . 52 for n = 74; P<0 . 0001 ) and a weaker MMP-9 correlation with pathology ( r = 0 . 31 for n = 74; P = 0 . 0077 ) . To determine if MMP-9 contributes to matrix destruction in TB patients , we assessed the gelatinase activity of the respiratory secretions . Induced sputum from TB patients showed an increased gelatinase activity ( P<0 . 0001; S5A Fig ) . However , MMP-9 neutralization with an inhibitory antibody at 10 μg/ml which completely suppresses gelatinase activity from MMP-9 [38] , did not decrease gelatinase activity in the respiratory secretions ( S5B Fig ) . To investigate the key regulatory pathways of neutrophil MMP-8 secretion , we performed a screening human phosphokinase array and observed that the AMP-activated protein kinase ( AMPK ) pathway was consistently activated in M . tb-infected neutrophils , especially AMPKα2 ( T172 ) ( Fig 5A ) . This activation was confirmed by immunoblotting ( Fig 5B ) . Components of the MAP-kinase , STAT pathways , p53 and Src family of kinases were also activated consistent with previous data [39–41] ( S6A , S6B and S6C Fig ) . AMPK is considered a master regulator of cellular energy homeostasis , existing as a heterotrimeric complex comprising catalytic α-subunits and regulatory β- and γ-subunits [13] . Its activation sets off a cascade of catabolic pathways including glycolysis and ketogenesis which can lead to the wasting which is characteristic in TB patients . We demonstrated that AMPKα was activated by phosphorylation in neutrophils directly infected with M . tb ( Fig 5C ) . The specific AMPK inhibitor Compound C ( Comp C ) blocked neutrophil MMP-8 secretion in a dose-dependent manner towards baseline levels ( Fig 5D ) and also suppressed gene expression of neutrophil MMP-8 ( Fig 5E ) , confirming AMPK is functionally active in regulating neutrophil MMP-8 secretion . Since the AMPK pathway may be downstream of the Akt/PI3-kinase pathway [42 , 43] and the Akt/PI3-kinase pathway may drive tissue destruction in TB [44] , we investigated whether this path regulates human neutrophil MMP-8 secretion . Neutrophil Akt was phosphorylated in response to CoMTB ( S7A Fig ) but MMP-8 secretion was not suppressed by either the Akt-inhibitor ( Akt-i ) or the broad-spectrum PI3-kinase inhibitor LY 294002 ( S7B and S7C Fig ) . We also examined whether the mTOR/p70S6 kinase regulated neutrophil MMP-8 secretion as this is downstream of AMPK [45] . p70S6 kinase was phosphorylated in neutrophils stimulated by CoMTB ( S7D Fig ) but the mTOR inhibitor rapamycin did not inhibit neutrophil MMP-8 secretion ( S7E Fig ) , indicating that neutrophil MMP-8 secretion is independent of this pathway . Finally , we studied AMPK in vivo . In human TB lung specimens , AMPKα was phosphorylated within the nuclei of neutrophils in TB cavities ( Fig 6A and 6B ) , indicating a state of energy depletion [46] . AMPK regulation of neutrophil MMP-8 was further investigated using neutrophils from a group of patients with defects in AMPK activation . Their clinical phenotype is typically similar to those found in glycogen storage disorders [47] . Genotypically , these patients have an AMPKγ2 mutation [48] . AMPKα mutations have not been described in man . In these patients , AMPKα was phosphorylated in both unstimulated and stimulated neutrophils but not in healthy donors , demonstrating increased basal phosphorylation of AMPKα ( Fig 6C ) . Such increased basal AMPK activity reduces the sensitivity of the protein kinase to AMP , resulting in functional AMPK deficiency [47] . MMP-8 secretion from AMPK-deficient neutrophils stimulated by CoMTB was significantly less than MMP-8 secreted by AMPK-replete neutrophils ( Fig 6D , P<0 . 01 ) , implicating AMPK in the regulation of neutrophil MMP-8 secretion in man .
Neutrophils are emerging as key mediators of TB-associated inflammation . They drive the unique TB transcript signatures in man [39] and predominate in respiratory secretions of patients with pulmonary TB [1 , 49] . We found sputum from TB patients had increased MMP-8 concentrations , neutrophil myeloperoxidase ( MPO ) and neutrophil gelatinase associated lipocalin ( NGAL ) compared to controls . MMP-8 was strongly associated with markers of neutrophil activation , MPO and NGAL , indicating that sputum MMP-8 is likely to be neutrophil-derived . Neutrophils expressing MMP-8 were found in the inner walls of tuberculous cavities and may further erode the lung matrix , extending previous findings that neutrophils are the predominant phagocytic cells in the respiratory secretions of TB patients [1] . We also found higher MMP-8 in TB patients with cavities on their chest radiographs than those without cavitation . These findings are consistent with and extend our previous data in smaller groups of patients that demonstrated a trend to increased MMP-8 compared to patients with respiratory symptoms [9 , 10] . Furthermore , MMP-8 is significantly elevated in plasma samples of patients with TB compared to respiratory symptomatics [50] . These observations underscore the importance of collagenases , such as MMP-8 and MMP-1 , which are the only enzymes capable of degrading the collagen triple helix at neutral pH . The consistent elevation of MMP-8 across the different patient cohorts implicates neutrophils as key players in tissue destruction in TB . In animal models of TB , neutrophil influx is associated with poorer outcomes with higher bacterial burden , earlier mortality and tissue inflammation [51–53] . However , the mechanisms linking neutrophil excess and poor outcomes are unclear . In our human study , M . tb drove neutrophil MMP-8 secretion , causing destruction of collagen , the main structural protein in human lung , both in vitro and in TB patients . We showed neutrophil MMP-8 closely correlated with sputum collagenase activity as well as clinical CXR score and TB severity score , implicating neutrophils in driving pathology in man by their collagenolytic activity . MMP-8 was also associated with NETs in M . tb infection and NET components such as citrulline H3 , which were not present in dead neutrophils , were increased in the respiratory secretions of TB patients . This may further contribute to immunopathology since NETs are recognized to induce cell death [54 , 55] . It is likely that different MMPs predominate in different stages of disease in TB immunopathology . There is good evidence that neutrophils are present not only during the acute phases of TB infection with macrophages , but are also a dominant cell type at the site of established infection of the cavity together with lymphocytes [1 , 51 , 52] . Such neutrophils contain high concentration of pre-synthesized MMP-8 [56] , and so can drive the later stages of TB which leads to lung cavitation , morbidity and death MMP-8 inhibition may be a target to abrogate excessive host tissue destruction . MMP-8 is a critical mediator of lung parenchymal damage other lung diseases , such as COPD [57] and ventilator-induced lung injury , and MMP-8 inhibition improves outcomes in a murine model of lung injury [58] . We showed higher MMP-8 in the respiratory secretions of patients with cavities than those without and MMP-8 neutralization decreased the matrix destruction in the sputum of TB patients . Although neutrophil gelatinase MMP-9 is secreted in TB and expressed in TB patients , neutralizing MMP-9 did not reduce gelatinase activity in TB patients . Neutrophil MMP-8 secretion in TB was inhibited by NF-kB inhibitors helenalin , CAPE and SN50 without altering neutrophil viability . Furthermore , doxycycline reduced neutrophil collagenase activity to baseline and MMP-8 neutralizing antibody decreased collagen destruction ex vivo in TB patients . Such immunomodulatory agents have potential to reduce tissue destruction in TB . For the first time in TB , we have shown that there is interaction between the metabolic AMPK signaling pathway in the regulation of neutrophil MMP-8 secretion and innate-immune mediated tissue destruction . AMPK activates catabolic pathways such as fatty acid oxidation and glycolysis to generate ATP , while switching off energy-consuming processes including protein and fatty-acid biosynthesis and cell-cycle progression . Two studies have shown that the development of lung injury in murine models is dependent on the pro-catabolic AMPK pathway [59 , 60] with AMPK activation decreasing lung injury . This contrasted with our findings where AMPK inhibition decreased neutrophil MMP-8 secretion , and maybe due to AMPK having divergent effects on different cells . The AMPK pathway was up-regulated in our cellular model and drives neutrophil MMP-8 secretion and gene expression , which was inhibited by the AMPK inhibitor Compound C . This finding was repeated in a small cohort of extremely rare patients with functional AMPK deficiency from a γ2-subunit mutation , where M . tb-driven neutrophil MMP-8 was decreased compared to healthy volunteers . While this does not definitively prove that AMPK regulates MMP-8 secretion as metabolic differences in neutrophils may cause divergent secretion , it implicates AMPK in driving neutrophil-mediated pathology by MMPs in TB . This finding is in keeping with a recent study where AMPKα2 deficient mice had decreased MMP-2 and were found to be resistant to developing abdominal aneurysms , a process which was demonstrated to be MMP-dependent [61] . Recent data demonstrate that AMPK has the ability to shuttle through the nucleus and contains both cytosolic components and nuclear components [46] with the ability to control transcription [62] . We showed AMPKα was phosphorylated in M . tb-infected neutrophils and adjacent to human TB lung cavities , with phospho-AMPKα located in the nuclei , signifying a state of energy depletion [46] . Together , our data from our human cellular model and in patients demonstrate that neutrophils drive MMP-8-dependent tissue destruction in TB , providing an insight as to how excessive neutrophil infiltration exerts a detrimental effect on the host . This process is controlled by the metabolic regulator AMPK as demonstrated in vitro and in AMPK deficient patients . This study highlights a previously unappreciated connection between metabolic paths that directly interact with innate immune responses causing immunopathology in human TB . Interventions specifically targeting the intersection of metabolic and innate immune responses to decrease tissue destruction may improve outcomes in TB and other inflammatory disorders .
Mouse anti-human beta-actin , rapamycin and doxycycline hyclate were from Sigma . Helenalin , SN50 and Compound C were from Merck Biochemicals . Caffeic acid phenethyl ester ( CAPE ) was from Tocris ( R&D Biosystems ) . Mouse anti-human MMP-8 , mouse anti-human MMP-9 , rabbit anti-human GAPDH , rabbit anti-human histone 2B , rabbit anti-human phospho-p70S6k ( T229 ) , rabbit anti-Mycobacterium tuberculosis , rabbit anti-human neutrophil elastase , rabbit anti-human phospho-AMPK alpha 1 and 2 ( T172 ) , sheep anti-human histone 2B , rabbit anti-human histone H3 ( citrulline 2 + 8 + 17 ) , donkey anti-Sheep IgG DyLight 488 , goat anti-mouse DyLight 549 , goat anti-rabbit IgG Cy5 were from Abcam . Rabbit anti-human phospho-Akt , total-Akt , phospho-AMPKα1/2 ( T172 ) , total AMPKα , and goat anti-rabbit HRP linked were from Cell Signalling Technology . Goat anti-mouse IgG ( H+L ) was from Jackson Immunoresearch laboratories . Goat anti-human MMP-8 was from R&D Biosystems and mouse anti-human MMP-9 was from Millipore . Rabbit anti-human MMP-8 was from Novus Biologicals . Mouse anti-human neutrophil elastase was from Dako . Mouse anti-human MMP-9 was from Millipore . M . tuberculosis H37Rv was cultured in supplemented Middlebrook 7H9 medium ( BD Biosciences ) . For infection experiments , mycobacteria were used at mid-logarithmic growth at an optical density of 0 . 60 ( Biowave cell density meter; WPA ) . Whole blood were drawn in preservative-free heparin and mixed with equal volumes of 3% dextran saline to remove erythrocytes . Neutrophils were isolated from the resulting cell suspension using Ficoll-Paque density centrifugation and three rounds of hypotonic lysis . Neutrophil purity was over 95% by FACS and viability >99% by trypan blue assay . In some experiments , neutrophils were pre-incubated with specific inhibitors/agents as indicated for 30 minutes unless otherwise stated . In all experiments involving live M . tuberculosis H37Rv , tissue culture medium was sterile filtered through 0 . 2 -μm Anopore membranes ( Millipore ) before removing from the CL3 laboratory . All experiments were performed using 4 hour incubations unless otherwise stated . Primary human blood monocytes were prepared from donor leukocyte cones from healthy donors ( National Blood Transfusion Service , UK ) . After density gradient centrifugation ( Ficoll Paque ) followed by adhesion purification , monocyte purity was over 95% by FACS analysis . Monocytes were infected with M . tuberculosis at a multiplicity of infection ( MOI ) of 1 . After incubation at 37°C for 24 h , conditioned medium was harvested and was termed CoMTB . Media from uninfected monocytes was termed CoMCont . TIMP-1 and 2 concentrations were measured using the Duoset ELISA Development System ( R&D Systems ) and detected a minimum of 31 . 2 pg/ml for both . The human MPO Quantikine ELISA kit ( R&D Systems ) was performed according to manufacturer’s instructions and the lower limit for MPO detection was 0 . 1 ng/ml . NGAL was measured using the human NGAL ELISA kit ( Bioporto Diagnostics ) which had minimum detection limit of 1 . 6 pg/ml . MMP-8 and -9 concentrations were analyzed by Fluorokine multianalyte profiling kit according to the manufacturer’s protocol ( R&D Systems ) on the Luminex platform ( Bio-Rad ) . The minimum level of detection for MMP-8 and -9 was 110 pg/ml and 65 pg/ml respectively . Cytokine concentrations were analyzed using a human 30-plex panel ( Invitrogen ) . The Proteome Profiler Human Phospho-kinase array kit ( R&D Systems ) which detects 45 phosphorylated proteins was performed according to the manufacturer’s protocol and developed with the ECL system ( Amersham Biosciences ) . Thirty minutes after neutrophils were stimulated with CoMTB , the cells were pelleted and lysed in lysis buffer . Equal amounts of total protein were loaded on to each array . Densitometric analysis of arrays was performed using Scion Image version Beta . 4 . 0 . 2 . Type I collagen and gelatin degradation was assessed using the EnzChek Gelatinase/Collagenase Assay kits ( Molecular Probes ) . Samples were activated with 2 mM of 4-amino-phenyl mercuric acetate ( APMA ) for 1 hour at 37°C . 80μL of reaction buffer or inhibitor ( doxycyline hyclate , Goat anti-human MMP-8 or Mouse anti-human MMP-9 ) were added with 20μL of either DQ collagen or gelatin ( Invitrogen ) at a final concentration of 25μg/ml . Activated samples were subsequently added , and activity detected at specified times using a fluorometer ( FLUOstar Galaxy ) . Human neutrophils were infected with M . tb at an MOI of 10 and 20 nM PMA was used as a positive control . 5 U/ml of micrococcal nuclease ( Fermentas ) was added in each well for 10 minutes at 37°C , after which EDTA was used to halt the reaction . Supernatants were collected , sterile filtered and stored at 4°C . NETs were quantified using QuantiT PicoGreen ( Invitrogen ) according to manufacturer’s instructions . Pelleted neutrophils infected with M . tb or stimulated with CoMTB were mixed with SDS lysis buffer . Samples were run on the NuPAGE 4–12% Bis-Tris gels with SDS Running buffer ( Invitrogen ) . Protein was transferred onto a nitrocellulose membrane ( GE Healthcare ) . Primary antibody was diluted in 5% BSA/0 . 1% Tween and incubated overnight at 4°C with agitation . Secondary antibody was added diluted in blocking buffer . Luminescence was demonstrated with ECL Substrate Reagent ( Amersham Science ) according to manufacturer’s instructions and exposing the membrane to Hyperfilm ECL . Densitometric analysis was performed using Image J 1 . 43U ( NIH , USA ) . Total RNA was extracted from 2 x 10 6 neutrophils using the RNeasy Mini Kit ( Qiagen ) . Quantitative real-time RT-PCR was performed using the OneStep RT- PCR master mix ( Qiagen ) according to the manufacturer’s instruction on a Stratagene Mx3000P platform using 5–10 μg per sample . MMP-8 primer and probe mixes were obtained from Applied Biosystems . GAPDH ( Forward primer 5’- CGCTTCGCTCTCTGCTCCT-3’ , reverse primer 5’- CGACCAAATCCGTTGACTCC-3’ , probe 5’-HEX-CGTCGCCAGCCGAGCCACAT-TAMRA-3’ ) was analyzed in parallel . To accurately determine the quantitative change in RNA , standard curves were prepared from plasmids subjected to real-time PCR as above . MMP-8 data were normalized to GAPDH detected in the same sample . Cell viability was assessed by staining neutrophils with Annexin V-FITC apoptosis detection kit ( eBioscience , Affymetrix , California , USA ) and live/dead fixable dead cell stain kit ( Invitrogen ) . Neutrophils were stimulated with 200 ng/ml staurosporine to induce apoptosis and this was used as a positive control for all experiments . Annexin V was detected on the FL-1 channel and live/dead dye on FL-3 . A total of 50 , 000 events were gated and analysed on BD FACSCalibur flow cytometer using CellQuest . Data was analysed using FlowJo 7 . 6 . 5 ( Tree Star ) . Permanox chamber slides ( Nunc Labtech ) were coated with 0 . 1 mg/ml fibrinogen with or without 25 μg/ml of DQ collagen for 30 minutes . For experiments involving NETs , 10 Ü/ml DNase ( Fermentas ) was added for 20 minutes at room temperature at the end of the experiment . Samples were then fixed with 4% paraformaldehyde for 30 minutes and permeabilized with 0 . 5% saponin for 10 minutes . Cells were washed before blocking with 10% human AB serum with 2 . 5% BSA and 0 . 05% saponin . Primary antibodies were added overnight . Chamber slides were washed prior to the addition of secondary antibodies . The chambers were subsequently removed from the slide , and Fluoroshield Mouting medium with DAPI ( Abcam ) was added . Images were captured using Leica confocal microscope ( Leica TCS SP5 ) and processed using Leica LAS AF Lite 2 . 6 . 0 ( Leica Microsystems , Germany ) and Image J 1 . 43U ( NIH , USA ) . Five non immunosuppressed patients with biopsy proven pulmonary M . tb infection were analysed . The positive controls colon tumours ( AMPK ) from 10 patients and inflammed appendix ( MMP-8 ) . Negative controls were performed using the appropriate isotype control antibodies . Sections were immunostained for MMP-8 and phospho-AMPK alpha 1 and 2 ( T172 ) ; neutrophil elastase with epitope retrieval performed by enzyme digestion using Bond Enzyme Pretreatment Kit . All antibodies were incubated for 15 minutes at room temperature . All immunohistochemistry was performed using the Leica Bond-III automated platform and associated ancillary reagents ( all Leica Biosystems ) . The antibodies were detected using the Bond Polymer Refine Detection System and Bond DAB Enhancer according the manufacturer’s instructions . Data were analyzed using GraphPad Prism ( version 5 . 04 , GraphPad Software ) . Data are expressed as mean ± s . d . unless stated otherwise . All experiments are performed in biological triplicates on at least 2 separate occasions . Multiple intervention experiments are compared with one-way ANOVA followed by Tukey’s post-test correction , while continuous variables between 2 sets of data are assessed using two-tailed Mann-Whitney-U test . Spearman’s rank correlation tests are used for correlation analyses . P values of less than 0 . 05 are taken as statistically significant .
|
Neutrophil infiltration is characteristic of immune-induced pathology in tuberculosis but mechanisms whereby neutrophils cause tissue destruction are not fully understood . In this study , we show that neutrophils secrete the collagenase MMP-8 in response to direct infection with Mycobacterium tuberculosis and via cellular networks . MMP-8 is up-regulated in respiratory samples from TB patients , driving matrix destruction associated with neutrophil activation and reflects disease severity . Neutrophils are present adjacent to the wall of TB cavities in human histology specimens . The metabolic pathway AMP-activated protein kinase ( AMPK ) regulates neutrophil MMP-8 secretion with data supported by studies in human neutrophils from AMPK-deficient patients . Host-directed therapy against neutrophil MMP-8 may reduce innate-immune mediated tissue damage in TB .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Neutrophil-Derived MMP-8 Drives AMPK-Dependent Matrix Destruction in Human Pulmonary Tuberculosis
|
The TEAD family of transcription factors ( TEAD1-4 ) bind the MCAT element in the regulatory elements of both growth promoting and myogenic differentiation genes . Defining TEAD transcription factor function in myogenesis has proved elusive due to overlapping expression of family members and their functional redundancy . We show that silencing of either Tead1 , Tead2 or Tead4 did not effect primary myoblast ( PM ) differentiation , but that their simultaneous knockdown strongly impaired differentiation . In contrast , Tead1 or Tead4 silencing impaired C2C12 differentiation showing their different contributions in PMs and C2C12 cells . Chromatin immunoprecipitation identified enhancers associated with myogenic genes bound by combinations of Tead4 , Myod1 or Myog . Tead4 regulated distinct gene sets in C2C12 cells and PMs involving both activation of the myogenic program and repression of growth and signaling pathways . ChIP-seq from mature mouse muscle fibres in vivo identified a set of highly transcribed muscle cell-identity genes and sites bound by Tead1 and Tead4 . Although inactivation of Tead4 in mature muscle fibres caused no obvious phenotype under normal conditions , notexin-induced muscle regeneration was delayed in Tead4 mutants suggesting an important role in myogenic differentiation in vivo . By combining knockdown in cell models in vitro with Tead4 inactivation in muscle in vivo , we provide the first comprehensive description of the specific and redundant roles of Tead factors in myogenic differentiation .
The TEAD ( 1–4 ) transcription factors [1 , 2] bind to a consensus MCAT ( 5’-CATTCCA/T-3’ ) element , originally identified as the SV40 enhancer GT-II motif [3] [4–6] , through the evolutionarily conserved TEA/ATTS DNA binding domain [7 , 8] . Mammalian TEADs are widely expressed with prominent Tead1 and Tead4 expression in skeletal muscle , lung , heart and nervous system . Tead factors act as mediators of the Hippo signalling pathway interacting with the Yap and Wwtr1 ( Taz ) transcriptional co-activators to regulate proliferation , oncogenesis , stem cell maintenance and differentiation and control of organ size [9–14] . Teads also play an important role in skeletal , cardiac , and smooth muscle differentiation and physiology [15–18] . Tead4 is expressed in developing skeletal muscle in mouse embryos [2] and at later stages both Tead1 and Tead4 are co-expressed and co-localise to somites [19] . Blais et al . showed that Myod1 and Myog directly bind the Tead4 promoter and activate its expression during C2C12 cell differentiation in vitro [20] . Subsequently , we showed that stable shRNA-mediated Tead4 knockdown led to formation of shortened C2C12 myotubes [21] . ChIP-chip experiments in C2C12 cells overexpressing Flag-HA-Tead4 revealed that Tead4 occupied 867 promoters including Myog , and Cav3 . RNA-seq identified a set of genes down-regulated upon Tead4 knockdown amongst which are muscle structural and regulatory proteins . While our data described a role for Tead4 in activating muscle genes during differentiation , we also suggested that Tead4 may repress Ctgf and Ccnd1 expression contributing to cell cycle exit . However , we did not specifically address the role of other Teads in these cells . Here , we show by siRNA silencing that Tead factors are essential for primary myoblast ( PM ) differentiation , but that Tead1 , Tead2 or Tead4 play partially redundant roles . In contrast to C2C12 cells , silencing of Tead1 or Tead4 both impaired differentiation indicating a differential requirement for these factors in PMs and C2C12 cells . ChIP-seq of Tead1 and Tead4 identified their binding sites C2C12 cells and in mature muscle fibres . RNA-seq identified distinct but overlapping sets of genes deregulated by Tead silencing in C2C12 cells and PMs . Furthermore , somatic inactivation in muscle in vivo revealed an important role for Tead4 in muscle regneration . We therefore provide the first comprehensive study of the specific and redundant regulatory roles of Tead factors in myogenic differentiation .
We previously showed that Tead4 plays an essential role in C2C12 cell differentiation [21] . To extend the study of Tead4 function , we isolated primary myoblasts from 3–4 week-old C57BL/6 mice and differentiated them in vitro for 6 days . Quantitative RT-PCR analysis showed that Tead4 mRNA expression was strongly induced at days 3 and 6 during differentiation ( Fig 1A ) . Similarly , expression of Tead1 was also strongly induced , whereas Tead2 expression did not show strong variation and Tead3 was not significantly expressed in myoblasts . Immunoblots confirmed that Tead4 protein was strongly induced upon differentiation , whereas Tead1 protein was present in undifferentiated and differentiated cells ( S1A Fig ) . To address the function of Tead4 , control and Tead4 siRNAs were transfected 24 hours before the initiation of differentiation . Compared to control siRNA , siTead4 led to a potent reduction in Tead4 expression , but did not affect Myog and Myod1 levels ( Fig 1B ) . Staining of the transfected cells for myosin heavy chain ( hereafter Myh ) expression showed that Tead4 silencing did not impair myoblast differentiation as the Tead4-silenced cells formed long multinucleate myotubes ( Fig 1C and 1D ) . As Tead1 was also present in differentiating PMs , we investigated potential redundancy amongst the Teads . We used siRNAs to silence Tead1 , Tead2 or Tead4 and examined how this affected the expression of the other family members . Tead1 expression was strongly down-regulated by siTead1 , less so by siTead2 , but not by siTead4 ( Fig 1E and S1A Fig ) . Similarly , expression of Tead4 mRNA was not affected by silencing of Tead1 or Tead2 , but Tead4 protein levels were increased in undifferentiated cells when Tead1 was silenced . Expression of Tead2 was reduced in the undifferentiated state upon silencing of either Tead1 or Tead2 , but its expression during differentiation was minimally affected . Thus , expression of each of the Teads was rather independent of the others . As seen above , siTead4 silencing had little effect on differentiation ( Fig 1F ) . Similarly , Tead1 or Tead2 silencing had little effect on differentiation ( Fig 1F and 1G ) . Examination of gene expression showed nevertheless that Myh1 and Myh2 were reduced by each knockdown ( Fig 1H ) . In contrast , expression of Cav3 , a well-defined Tead4 target gene in C2C12 cells , and Myh7 was strongly and selectively diminished in siTead4 cells indicating a specific requirement for Tead4 at these genes that cannot be compensated by expression of the other Teads . We performed combinatorial silencing of Tead1 and Tead4 or of all three expressed Teads . The expression of the corresponding Tead mRNAs was reduced in all cases ( Fig 2A ) . In contrast to individual Tead1 and Tead4 knockdowns , their combinatorial knockdown had a potent effect on differentiation , with many cells expressing Myh , but no fusion and the prevalence of shorter myotubes ( Fig 2B and 2C ) . Moreover , a larger number of cells failed to initiate Myh expression . Similar observations were made using the Tead1 , Tead2 and Tead4 siRNA combination ( Fig 2B ) . Strongly reduced Myh1 , Myh2 , Myh7 and Tnni1 expression was seen in the siTead1/Tead2/Tead4 cells ( Fig 2D ) . Together , these observations showed that Teads play essential , but partially redundant functions in differentiating PMs . The redundant roles for Tead factors in differentiating PMs contrast with the critical role for Tead4 in C2C12 cells . To investigate this more closely , we performed single and combinatorial siRNA Tead knockdowns in C2C12 cells . As reported [21] , Tead4 was strongly induced during C2C12 cell differentiation ( Fig 3A ) and its induction was not diminished by Tead1 silencing , but was somewhat reduced by Tead2 silencing . Tead2 was also induced albeit less strongly than Tead1 or Tead4 , but its activation was strongly diminished by Tead1 or Tead4 silencing . Tead1 mRNA was induced during differentiation , but this was strongly reduced by siTead4 . In all situations , transfection with siRNAs against individual Teads or combinations of Teads had the potent and expected effects on their own expression . Tead1 or Tead4 silencing led to reduced myoblast fusion with the absence of longer and thicker fibres in favour of shorter less developed fibres ( Fig 3B and 3C ) . A similar , but less pronounced , effect was seen upon Tead2 silencing . Combinatorial Tead1/Tead4 silencing led to more dramatic effects with fewer and shorter fibres , while upon silencing of all three Teads few elongated myotubes were observed ( Fig 3B and 3C ) . These results revealed that normal expression of each Tead was essential for full differentiation and generation of long and thick fibres , and that Tead1 and Tead4 both strongly contributed to differentiation . Western blot analyses showed increased Tead4 protein levels in differentiated cell extracts ( S1B Fig ) . Tead1 on the other hand was decreased at day 6 in agreement with previous observations [22] . While Tead4 was increased in siTead1 cells , Tead1 was reduced in siTead4 cells . This highlights a difference with PMs where Tead1 was strongly induced even in the absence of Tead4 ( S1A Fig ) , whereas in C2C12 cells Tead4 is required for maximal Tead1 expression . Immunostaining showed Tead1 nuclear localisation in non-differentiated C2C12 cells , whereas Tead4 was present in both the nucleus and cytoplasm ( S1C Fig ) . At day 6 , Tead1 remained nuclear in cells that did not undergo differentiation , but was absent from differentiated myotube nuclei . In contrast , Tead4 expression was not detected in cells that did not undergo differentiation , but showed strong nuclear staining in myotubes . Strikingly , a comparison with PMs showed that Tead1 was strongly expressed in the nuclei of both myoblasts and myotubes , while Tead4 was both cytoplasmic and nuclear in myoblasts , but nuclear in myotubes ( S1D Fig ) . This observation could account for the differential requirement for Tead1 and Tead4 in PMs and C2C12 cells . In PMs , both proteins were nuclear in myotubes and can thus partially compensate for each other , whereas in C2C12 myotubes , Tead1 was down-regulated by siTead4 and absent from the nucleus and thus unable to compensate for loss of Tead4 . To understand how Tead1 and Tead4 regulate gene expression in C2C12 cells , we used ChIP-seq to profile their genomic occupancy . Chromatin was prepared before differentiation and after 6 days of differentiation and ChIP was performed with antibodies selective for either Tead4 or Tead1 . In undifferentiated cells , Tead4 occupied 2940 sites located mainly distant from the transcription start sites ( TSS ) ( Fig 4A and S1 Dataset ) . In differentiated cells , more than 8100 sites were occupied , the majority of which were again located distant from the TSS ( Fig 4B and S1 Dataset ) . Occupied sites in both undifferentiated and differentiated cells showed strong enrichment of the MCAT motif ( Fig 4C and 4D ) . Other motifs co-occurred with the MCAT motif at higher than expected frequency . In undifferentiated cells , enrichment in motifs for the AP1 ( Fos and Jun ) family was observed along with Runx1 that cooperates with AP1 and Myod1 to drive myoblast proliferation during muscle regeneration [23] ( Fig 4C ) . In differentiated cells , AP1 and Runx motifs were enriched , but additional motifs became prominent such as Ctcf , and Tcf3 and the E-Box ( Fig 4D ) . Comparison of sites in undifferentiated and differentiated cells indicated that occupancy of 1495 sites was lost during differentiation , whereas more than 6700 sites were gained and 1434 sites were occupied under both conditions ( Fig 4E ) . For example , Tead4 constitutively occupied sites upstream of the Ctgf and Ccnd1 genes , whereas occupancy of sites at the Acta1 locus is seen only during differentiation ( S2A Fig ) . Sites specifically occupied in undifferentiated and differentiated cells showed enrichment in AP1 motifs in the undifferentiated state , but enrichment of Myod1 , Myog , Tcf3 and Tcf12 that cooperates with Myod1 to promote myogenic differentiation [24] in differentiated cells ( Fig 4F ) . To assess co-localisation between Tead4 and transcription factors whose motifs were enriched at its binding sites , we compared the Tead4 profiles with those of Jun and Srf in C2C12 cells and Runx1 in primary myoblasts [25] [26] . Around 35% of sites bound by Tead4 in undifferentiated and differentiated cells overlapped with those bound by Jun ( S2B Fig ) . In contrast , there was little overlap with the sites preferentially bound in the differentiated state , where the frequency of co-occurring AP1 motifs was also reduced . Similarly , a strong binding overlap between Tead4 , Runx1 and Srf was seen involving sites occupied in undifferentiated and differentiated cells . A similar analysis of Tead1 identified 1400 bound sites in undifferentiated cells , enriched in MCAT , AP1 and Runx motifs ( S3A and S3B Fig ) . Nevertheless , in agreement with its absence from the differentiated cell nucleus , Tead1 occupancy was strongly reduced in the differentiated state with only 274 detected sites ( S3C Fig ) . Even at sites occupied in both conditions , Tead1 occupancy was reduced in the differentiated state ( S3D Fig ) . Thus , transition to the differentiated state involved a switch from Tead1 and Tead4 occupancy to predominantly Tead4 occupancy . In the undifferentiated state , Tead1 and Tead4 co-occupy more than 900 sites ( S3E Fig ) . A set of sites showed preferential occupancy by Tead1 , but only few sites showed exclusive Tead1 occupancy . Thus despite the fact that these two proteins bind identical sequences and that Tead1 occupancy was globally lower than Tead4 , a set of sites was preferentially occupied by Tead1 . As Tead4 regulated Tead1 expression during differentiation , we examined Tead4 occupancy at the Tead1 locus . Two constitutive Tead1/Tead4 occupied sites were observed , one upstream of the promoter of the longest isoform and a second upstream of an alternative promoter for a shorter isoform ( S4A Fig ) . During differentiation Tead1 occupancy diminished , but Tead4 occupancy was maintained suggesting that Tead4 directly regulates Tead1 . Integration of Tead1/4 ChIP-seq data with public data on histone modifications in undifferentiated and differentiated C2C12 cells revealed the presence of H3K27ac , a mark of active promoters and enhancers , at the Tead1 promoter in undifferentiated cells overlapping with the Tead1/4 occupied sites . Interestingly , upon differentiation , H3K27ac increased at the Tead1/4 occupied sites and new regions marked by H3K27ac appeared upstream of and overlapping with the alternative promoter . Moreover , integration with public ChIP-seq data indicated Myod1 and Myog [27] binding at the H3K27ac-enriched regions . These observations suggest that Tead4 cooperates with Myog and Myod1 to activate Tead1 expression during differentiation via constitutive and inducible enhancer elements , in agreement with Tead1 down-regulation in siTead4 knockdown C2C12 cells ( Fig 3A ) . At the Tead4 locus , Tead4 occupied a H3K27ac-marked site immediately upstream of its own promoter whose occupancy increased upon differentiation ( S4B Fig ) . In contrast , almost no Tead1 occupancy was seen . Upon differentiation , regions in the Tead4 gene body acquired H3K27ac , several which coincided with binding of Myod1 and Myog . This suggests that Tead4 positively regulates its own expression together with these factors that bind to differentiation-induced enhancer elements downstream of the Tead4 TSS . The above data show that at the Tead1 and Tead4 loci , enhancers binding Myog became activated during differentiation perhaps driving their expression . We tested this by performing siMyog silencing in C2C12 cells and in PMs . In both cell types , siMyog strongly inhibited differentiation ( S5A Fig ) . In C2C12 cells , Tead4 expression was reduced upon Myog silencing , while that of Tead1 was unaffected , and Ccnd1 expression was increased ( S5B Fig ) . Expression of Mef2c bound by both Myog and Tead4 ( S5C Fig ) was also strongly repressed . Hence , Myog is required for Tead4 , but not Tead1 , expression and differentiation . In a global analysis , we clustered Tead4-occupied sites in undifferentiated and differentiated cells with H3K4me3 a mark of active promoters , H3K4me1 , a mark of active and poised enhancers as well as H3K27ac . As few Tead4 sites localised at the TSS , only a limited overlap ( 280 of 2940 ) with H3K4me3 was observed ( Fig 5A and 5B ) . In contrast , 1698 Tead4 sites in undifferentiated cells showed strong association with H3K4me1 and/or H3K27ac defining a set of sites at active and poised enhancer elements . A similar situation was seen in differentiated cells where almost half were marked by H3K4me1 and H3K27ac and up to 1500 sites associated with H3K4me3 ( Fig 5A and 5B ) . Tead4 therefore occupied a set of functional regulatory elements in both undifferentiated and differentiated cells . A similar situation was seen for Tead1 in undifferentiated cells ( S6A and S6B Fig ) . Due to their low number , we did not analyse Tead1 sites in differentiated cells . To define the regulatory potential of Tead4 , we identified the genes closest to the Tead4-occupied sites associated with active chromatin marks . In undifferentiated cells , 1262 genes enriched in the ontology terms cell structure and motility , developmental processes , oncogenesis and cell cycle control were annotated ( Fig 5C ) . Interestingly , KEGG pathway analysis revealed that Tead4 ( and Tead1 , S6C Fig ) occupied sites associated with critical components of the Tgfβ ( Smad2 , 3 6 and 7 as well as Tgfb2 ) and Wnt-signalling ( Fzd1 , Fzd5 , Tcf7l2 ) pathways ( Fig 5C ) . In addition , several genes of the Hippo pathway such as Amotl1 , Amotl2 and Lats2 were also associated with Tead1/4 occupied sites . In differentiated cells , more than 2000 genes enriched in terms associated with developmental processes , muscle differentiation and contraction were annotated including the important regulatory genes Myod1 , Myog and Mef2a as well as numerous structural genes of the muscle fibre . At many sites , Tead4 binding and H3K27ac was either enriched or acquired de novo at these genes during differentiation ( S6D Fig ) . We analysed global co-localisation of Tead4 with Myod1- and Myog-occupied sites . In differentiated cells , more than 2000 Tead4 sites were co-occupied by all three factors ( S7A Fig and S2 Dataset ) . As Tead1 occupied sites essentially only in non-differentiated cells , a comparison with Myod1 and Myog-occupied sites in differentiated cells revealed only a limited overlap of around 50 sites ( S7B Fig ) . The Tead4-Myod1-Myog-occupied sites showed enrichment not only in the recognition motifs for these factors , but also for Tcf3 , Tcf12 , Runx , and Klf5 , whereas the AP1 family sites were less represented than in the overall Tead4 profile ( S7C Fig ) . We compared Tead1/4 occupancy with that of Mef2a , another myogenic factor for which a public data set is available in undifferentiated C2C12 cells [28] and identified a set of sites co-occupied with Tead1 and Tead4 ( S7E and S7F Fig ) . This analysis identified Tead4 sites closely associated with Myod1/Myog . Nevertheless , as shown above at the Tead4 and Mef2c loci , Tead4 may cooperate with Myod1/Myog to activate these genes despite more distant localization of the binding sites . We therefore defined genes associated with Tead4-occupied sites and compared them with genes associated with Myog/Myod1 sites to identify those potentially regulated by these factors despite binding more distantly spaced promoter and/or enhancer elements . A large majority of Tead4 associated genes was associated with Myog/Myod1 whose potential target genes also showed a strong overlap ( S7D Fig ) We used RNA-seq to investigate gene expression in differentiating PMs and C2C12 cells and how simultaneous Tead1 and Tead4 silencing affected these regulatory programs to impair differentiation . SiRNAs were transfected and RNA prepared after 24 hours ( day 0 ) before cells were moved to differentiation media and RNAs prepared 3 and 6 days later ( Fig 6A ) . Changes in expression in siTead1/4 compared to the siControl were quantified to identify genes showing a greater than Log2 fold change of 1 with adjusted p value <0 , 05 . In control C2C12 cells , 3137 genes were induced at day 3 and day 6 with respect to day 0 , while in PMs 3626 genes were induced of which 1845 genes were commonly induced in both cell types ( S8A Fig ) . Commonly regulated genes were associated with muscle differentiation . Similarly , 2375 genes were repressed during C2C12 cell differentiation and 2799 genes repressed in PMs with 1495 common to both cell types ( S8B Fig ) . Commonly repressed genes were associated with cell cycle , consistent with proliferation arrest during differentiation . Thus , similar but not identical , gene expression programs were activated and repressed during the differentiation of these two cell types . Analysis of the 5512 genes regulated during differentiation of siControl C2C12 cells identified genes with different expression profiles that could be summarised in 6 clusters ( S9A Fig ) . Genes in clusters 1 , 3 and 4 were down-regulated with different kinetics , while those in clusters 2 and 5 were up-regulated with different kinetics , and those in cluster 6 were transiently induced at day 3 . Further analyses showed that myogenic genes were amongst the most significantly up-regulated at days 3 and 6 , whereas cell cycle progression genes were strongly repressed ( S9B Fig ) . In addition , adipogenesis genes were also induced along with a metabolic switch involving increased expression of genes involved in oxidative phosphorylation . Following siTead1/4 silencing , up and down-regulated genes were seen at day 0 . The number of de-regulated genes increased at day 3 and diminished by day 6 ( Fig 6A ) . In total , 249 genes were up-regulated by siTead1/4 silencing between day 0–6 , while 549 were repressed ( Fig 6A and 6B and S3 Dataset ) . Analysis up-regulated genes at days 3 and 6 indicated strong enrichment in cell cycle , Notch , Wnt and Tgfβ signalling and epithelial to mesenchymal transition ( Fig 6C ) . In contrast , genes involved in myogenesis and oxidative phosphorylation were repressed . Hence , Tead1/4 contribute to activation of the myogenic differentiation program , but they also directly or indirectly repress growth promoting pathways leading to defective cell cycle arrest . To identify genes directly regulated by Tead4 , we determined those within 50 kb of Tead4 binding sites enriched specifically in differentiated cells . Around 4300 Tead4 occupied sites were associated with 4100 potential target genes showing expression in the RNA-seq data ( Fig 6D ) . A set of 172 down-regulated genes enriched in muscle differentiation functions was associated with Tead4 binding sites in differentiated cells ( Fig 6D ) . Similarly , a set of 107 up-regulated genes was associated with sites preferentially bound in the differentiated state . These genes were enriched in the cell cycle , Notch and Wnt signalling identified by the GSEA analyses . Hence , binding of Tead4 to these repressed genes during differentiation provided further evidence that Tead4 contributed to their repression . A similar analysis of differentiating PMs clustered gene expression ( S10A Fig ) in 6 kinetic classes and showed that differentiation was characterised by activation of genes involved in myogenesis , oxidative phosphorylation and adipogenesis , while cell cycle genes were repressed ( S10B Fig ) . Following siTead1/4 silencing , deregulated genes were observed at day 0 , increased at day 3 and then diminished at day 6 ( Fig 7A ) . In total , 563 genes were up-regulated between day 0–6 , while 377 were repressed ( Fig 7A and 7B and S4 Dataset ) . Down-regulated genes were associated with myogenesis , and oxidative phosphorylation , the hallmarks of differentiation , whereas up-regulated genes were enriched in angiogenesis and as seen in C2C12 cells in Wnt signalling ( Fig 7C ) . In PMs , up-regulation of cell cycle genes was observed , but the values were less significant reflecting the reduced proliferative capacity of PMs compared to C2C12 cells . Thus , Tead factors were essential to activate genes involved in PM differentiation , but also to repress Wnt signalling and signalling pathways like Tgfβ inhibiting PM differentiation . We compared genes de-regulated by siTead1/4 silencing in PMs and C2C12 cells . As the kinetics of their activation of repression may differ , we compared non-redundant lists of all genes deregulated between days 0–6 in each cell type . A set of 119 genes strongly enriched in muscle differentiation functions were commonly down-regulated ( S11A Fig ) . Strikingly however , a large set of 430 genes , again strongly enriched in muscle differentiation functions , was specifically down-regulated in C2C12 cells ( S5 Dataset ) . A smaller set of 258 genes was specifically down-regulated in PMs , but showed low enrichment in more diverse functions . Only 65 genes involved in signalling and proliferation were commonly up-regulated . However , a large set of almost 500 genes was specifically up-regulated in PMs ( S11B Fig ) . Remarkably , these genes showed enrichment in nervous system development and other neurogenesis functions ( S5 Dataset ) . Tead1/4 knockdown appeared to modify PM cell identity leading to the expression of neurogenesis genes , not normally expressed during PM differentiation . These results showed that Tead1/4 silencing had distinct effects on gene expression in PMs and C2C12 cells . We next addressed genome occupancy by Tead4 in mouse muscle in vivo . We developed a protocol to prepare chromatin from dissected hind-limb muscle ( see Materials and methods ) and performed ChIP-seq for Tead4 , H3K27ac and RNA Polymerase II ( Pol II ) . We analysed the Pol II ChIP-seq to determine whether the signal obtained reflected mainly Pol II occupancy in muscle or in contaminating non-muscle cells . More than 38000 Pol II peaks were identified most of which localised at the TSS ( Fig 8A ) . Transcribed genes can show high levels of promoter paused Pol II and low levels in the gene body or low pausing , but abundant elongating Pol II [29] . The second class often corresponds to tissue identity genes controlled by so-called “super enhancers” [30 , 31] . Analyses of the Pol II ChIP-seq data identified around 1000 genes with high levels of transcribing Pol II ( class A in Fig 8B ) , a second class ( B ) also with high Pol II in the transcribed regions and larger groups of genes ( C and D ) with high Pol II at the promoter , but lower levels in the gene body . Class A genes also showed high levels of H3K27ac throughout the gene body typical of what has been described at cell identify genes ( Fig 8B ) . Class A genes associated with high Pol II and H3K27ac were enriched in terms associated with muscle fibre ( Fig 8C ) . For example , the locus comprising Myh2 , 1 , 4 , 8 and 13 showed high Pol II density specifically over the Myh4 gene with much lower densities over the Myh1 and Myh2 genes , but no transcription of other myosin genes at this locus ( Fig 8D ) . These loci also showed extensive H3K27ac surrounding and throughout the gene body . The Pol II and H3K27ac ChIP-seq therefore identified a set of highly transcribed muscle identity genes confirming the signal comes predominantly from muscle cells . Of the 2220 identified Tead4 sites , 686 were associated with active promoters marked by Pol II and H3K27ac and enriched in muscle specific functions ( Fig 9A and S12A Fig ) . Genes associated with Tead4-bound sites showed enrichment in terms associated with muscle structural proteins ( Fig 9B ) . Aligning the muscle Tead4 ChIP-seq to the coordinates of the differentiated C2C12 cell peaks revealed 1558 sites with significant signal ( Fig 9C ) . Genes associated with these shared sites were enriched in muscle structural proteins . In the converse comparison using the top 2200 Tead4-bound sites in muscle as reference , 341 common peaks were identified ( Fig 9D ) . These comparisons revealed Tead4 sites in muscle that were not called amongst the 2200 high confidence sites , but although showing lower occupancy in muscle were shared with differentiated C2C12 cells . Tead4 therefore bound a distinct repertoire of sites in C2C12 cells and muscle and sites with high occupancy in muscle did not necessarily show high occupancy in C2C12 cells and vice-versa . We next performed ChIP-seq from muscle of mice in which Tead4 was specifically inactivated in fibres using the Hsa::Cre-ERT2 driver [32] . Mice with floxed Tead4 alleles were crossed to generate Hsa::Cre-ERT2::Tead4lox/lox animals . These mice were injected at 6–7 weeks with tamoxifen for 4 consecutive days and 3 weeks after injection Tead4 expression was strongly reduced in the tibialis anterior and gastrocnemius muscles showing efficient recombination ( Fig 9E ) . We performed Pol II , H3K27ac and Tead4 ChIP-seq from these Tead4musc-/- animals . Aside a small number of sites with signal in Tead4musc-/- animals , Tead4 binding was lost ( Fig 9F ) indicating that observed signal came almost exclusively from sites bound in muscle . Comparison of Pol II and H3K27ac profiles in wild-type and Tead4musc-/- muscle ( Fig 9G ) showed only minor changes in low intensity signals and hence that Tead4 loss did not affect global Pol II or H3K27ac distribution . For example , at the Acta1 locus Tead4 occupancy was lost in mutant muscle , whereas no change in Pol II and H3K27ac profiles was observed ( S13 Fig ) . Similarly , Tead4 occupied 3 sites at the Amolt2 locus in both C2C12 cells and wild-type muscle including an upstream enhancer site marked by H3K27ac ( S14 Fig ) . The shared sites co-localised with those occupied by Myod1 and Myog in C2C12 cells . In mutant muscle , Tead4 occupancy was lost , but no change for Pol II and H3K27ac was seen . In agreement with the unaltered Pol II and H3K27ac profiles , little change in expression of potential Tead4 target genes was seen in mutant muscle , only minor reductions in Myh2 and Myl2 expression were observed ( S15A and S15B Fig ) . Moreover , Tead4musc-/- animals did not show any marked phenotype in terms of muscle fibre size , muscle mass and grip strength ( S15C–S15E Fig ) . One potential explanation is redundancy with Tead1 . To investigate this possibility , we performed Tead1 ChIP-seq in muscle . Of the 358 sites identified , 188 were shared with Tead4 ( S12B Fig ) . Genes associated with Tead1-occupied sites were however enriched in muscle function . For example , prominent Tead1 and Tead4 occupancy was observed at the Acta1 and Amotl2 loci ( S13 and S14 Figs ) . Thus , redundancy with Tead1 may in part account for the lack of phenotype seen upon Tead4 inactivation . Alternatively , differentiating C2C12 cells and PMs represent a very different physiological state from mature differentiated fibres . A more comparable situation is muscle fibre regeneration . We therefore investigated the role of Tead4 in muscle fibre regeneration in vivo . We employed a protocol similar to that previously used to demonstrate the role of SRF in regeneration using the Hsa::Cre-ERT2 driver ( [33] and see Fig 10A ) . Muscle degeneration in Hsa::Cre-ERT2::Tead4lox/lox and Hsa::Cre-ERT2::Tead4+/+ animals was induced by notexin injection and Tead4 was inactivated in regenerating fibres by regular subsequent Tam injection ( Fig 10A ) . At 15 days after notexin injection , tibialis anterior mass was significantly lower in the Tead4musc-/- compared to the Tead4+/+ animals ( Fig 10B ) . Similarly , fibre cross-section area was significantly altered with more small fibres and less large fibres in Tead4musc-/- ( Fig 10C ) . Expression of several Tead4 target genes such as Myh1 and Myh2 , Ankrd2 , Lats2 and Amotl2 were all significantly reduced ( Fig 10D ) . Moreover , Tead1 and Myog expression were also reduced , whereas Ccnd1 expression was increased . Thus , the gene expression changes induced by Tead4 inactivation during notexin-induced regeneration were similar but not identical to those seen following siTead4 in PMs . At 30 days after notexin injection , tibialis anterior mass remained somewhat reduced in the mutant animals , but fibre cross-section area was comparable to that in control animals ( Fig 10E and 10F ) . Thus , Tead4 inactivation delayed the normal regeneration process .
Here we show the essential role of Tead factors in PM differentiation . While silencing of each individual Tead had little discernible effect at the cellular level , Tead4 silencing specifically affected expression of its direct targets Myh7 and Cav3 . Nevertheless , combinatorial Tead1/4 or Tead1/2/4 silencing strongly impaired PM differentiation with fewer cells initiating Myh expression and shorter myotubes . Functional redundancy may be explained by the persistent expression and nuclear localisation of Tead1 during differentiation of siTead4 PMs and vice-versa . In contrast , siTead4 silencing impaired C2C12 cell differentiation with formation of shorter myotubes . Individual siTead1 or siTead2 silencing also impaired differentiation , revealing differences in Tead contributions in PM and C2C12 cells . In C2C12 cells , Tead4 silencing diminished Tead1 and Tead2 expression . Indeed , Tead4 occupied Tead1 regulatory sequences to directly regulate its expression . In addition , while Tead1 and Tead4 were nuclear in differentiated PMs , Tead1 was absent from the differentiated C2C12 cell nuclei and therefore could not compensate Tead4 silencing . C2C12 cell differentiation is however impaired by siTead1 showing that it contributed to early events in this process . Differential contribution of Teads in the two cell types can thus be explained by differences in their regulation and intra-cellular localisation . Immunostaining detected Tead1 uniquely in the nucleus of non-differentiated C2C12 cells , whereas Tead4 expression was lower and distributed in both nucleus and cytoplasm . However , ChIP-seq showed higher genomic occupancy of Tead4 than Tead1 suggesting its preferentially recruitment to the non-differentiated cell genome . While it is possible that the ChIP-efficiency of the Tead4 antibody is higher than the Tead1 antibody , a set of sites showed preferential occupancy by Tead1 rather suggesting the overall lower binding of Tead1 cannot simply be explained by lower ChIP efficiency . Indeed , it has previously been reported that the Vgll2 cofactor induced during C2C12 cell differentiation inhibits Tead1 , but not Tead4 DNA binding [22] . Hence , it is possible that during differentiation Vgll2 acts to selectively inhibit Tead1 genomic binding leading either to its export from the nucleus and/or its reduced expression . In our previous study [21] , we performed ChIP in cells constitutively overexpressing tagged Tead4 . Despite constitutive Tead4 overexpression , we identified sites occupied only during differentiation consistent with their acquisition of H3K4me3 or H3K27ac . Others , exemplified by a site upstream of the Myog locus ( see S6D Fig ) , were occupied by exogenous , but not endogenous Tead4 in proliferating C2C12 cells . Thus , while Tead4 occupies many sites in undifferentiated C2C12 cells , there exists a subset of sites occupied only during differentiation irrespective of Tead4 expression levels , whereas others can be occupied in the undifferentiated state upon increased Tead4 expression . Tead4 genome occupancy is therefore not only regulated by its expression level , but also by changes in chromatin state during differentiation . Integration of Tead4 ChIP-seq data with that of chromatin modifications showed strong association of Tead4 occupied sites with active H3K27ac-marked regulatory elements in both undifferentiated and differentiated cells . Moreover , many sites showed Tead4/Myog/Myod1 co-occupancy . These observations reinforce the idea that Tead4 in particular and Teads in general may cooperate with Myod1 and Myog to regulate gene expression during differentiation . Myod1 orchestrates the activation of a compendium of muscle enhancer elements [25] [34] . The DNA sequences at these enhancers were enriched for the AP1 and Runx families , but not for the MCAT motif . Tead4-occupied sites in non-differentiated cells were enriched in AP1 and Runx motifs suggesting these factors collaborate to drive proliferation . Recently , sites occupied by Tead factors driving motility in cancer cells were identified and showed not only enrichment in AP1 motifs , but also many of the motifs that we found enriched at Tead4 sites in C2C12 cells [35] . In contrast , Tead4 sites preferentially occupied in differentiated cells showed little overlap with Jun , but better co-localisation with Runx and were enriched for E-boxes for Myod1/Myog consistent with their observed co-localisation . Nevertheless , although many Tead4 occupied sites are co-occupied by Myod1/Myog , these sites constitute only a smaller subset of a larger collection of Myod1/Myog sites explaining why the MCAT motif was not detected in the analyses of Blum et al . , [25] . RNA-seq showed that Tead1/4 drive distinct but overlapping gene expression programs in the two cell types . This partly reflects the different gene expression programs of PMs and C2C12 cells , but also suggests that Tead4 may occupy a different , but overlapping set of sites in these two cell types . Interestingly , a large number of muscle function genes are specifically down-regulated in C2C12 cells by Tead1/4 silencing . This may reflect the additional contribution of Mef2c that was down-regulated in C2C12 cells , but not PMs . More striking is the specific up-regulation of neuron-expressed genes in PMs suggesting that Tead1/4 silencing leads to altered cell identity . Nevertheless , many genes critical for fibre formation , contraction and neuromuscular junction are down-regulated by siTead1/4 in both cell types . Diminished expression of these genes contributes to the impaired differentiation observed . We previously suggested that in addition to activating muscle differentiation genes , Tead4 may also repress genes such as Ccnd1 and Ctgf that drive cell proliferation and whose expression is reduced upon differentiation [21] . While Tead4 binds genes like Ccnd1 before and during differentiation , we observed here sites preferentially bound in differentiated cells and associated with cell cycle such as E2f8 or Chek2 , or Wnt , Tgfβ and Notch signalling genes . These genes are normally repressed during differentiation , but were up-regulated after siTead1/4 silencing . Proper regulation of Wnt , Notch and Tgfβ signalling is essential for normal myogenic differentiation [36] [37] [38] and their mis-regulation impairs myogenesis and can lead to fibrosis [39] . For example , siTead1/4 silencing up-regulated its target genes Notch3 , and to a lesser extent Notch1 , and Dll1 ligand , accompanied by up-regulation of the Notch mediators Hey1 and Hey2 shown to inhibit myogenesis [40] . On the other hand , siTead1/4 silencing up-regulated its target gene Nkd1 , an antagonist of Wnt signalling [41] [42] that is normally required to promote differentiation . Thus , Tead4 plays a dual role during differentiation , not only activating the myogenic program , but also repressing cell cycle and signalling genes . Some discrepancies remain with our previous observations using shTead4 silencing in C2C12 cells . For example , shTead4 silencing strongly inhibited Myog expression , while this was not seen upon siTead4 silencing . This may reflect a fundamental difference in the two approaches . In the shRNA experiments , C2C12 myoblasts were infected , selected and Tead4 expression was silenced for up to 10 days before differentiation was initiated . As Tead4 occupies more than 2800 binding sites in proliferating myoblasts , it is likely that diminished Tead4 levels for several days prior to differentiation can affect activation of genes that are rapidly induced after differentiation . Extensive Tead4 genome occupancy in proliferating C2C12 cells may therefore play a critical role in establishing the proper chromatin state permissive for activation of genes during differentiation . Performing ChIP-seq directly from mature muscle fibres identified a set of a highly transcribed and H3K27ac-marked muscle cell identity genes and Tead4 and Tead1 bound sites . Tead1 and Tead4 occupied an overlapping set of sites that partially overlapped with those in C2C12 cells . Shared sites were strongly enriched at genes encoding muscle structural proteins and also at a smaller set of genes encoding signalling and cell cycle proteins . In particular , Tead1 and Tead4 occupied sites at genes of the Hippo signalling pathway like Lats2 and Amotl2 in C2C12 cells and in mature fibres . Similarly , in post-mitotic muscle , Tead4 occupied sites at the Ccnd1 and Ctgf loci that normally contribute to its proliferative function . This may reflect the known role of the Tead4-Yap1 axis in regulating muscle fibre size [13] . Despite the observed genomic occupancy , Tead4 inactivation in mature fibres had no marked effect on Pol II and H3K27ac distribution , led to only minor effects on target gene expression , and resulted in no evident phenotype . In contrast , Tead4 inactivation led to delayed muscle regeneration after notexin treatment . Significant reductions in muscle mass and fibre size were seen after 15 days , but by 30 days these parameters were comparable to those seen in control animals . In addition , Tead4 contributed to activation of muscle structural genes during regeneration-induced differentiation . The absence of phenotype in mature fibres is in agreement with a previous report where Tead4 was inactivated in post-implantation embryos [43] . Similarly , while Tead4 loss delayed regeneration , this process was not completely impaired , explaining the absence of a notable muscle phenotype seen in the study of Yagi et al . , although they did not specifically assay regeneration in their Tead4 mutant animals [43] . The results obtained with differentiating PMs in vitro and in muscle in vivo are all in accordance with strong redundancy between Tead1 and Tead4 in the myogenic process that minimises the effects seen upon loss of Tead4 alone . Nevertheless , our study defines for the first time the critical roles of these factors in myogenic differentiation .
Mice were kept in accordance with the institutional guidelines regarding the care and use of laboratory animals and in accordance with National Animal Care Guidelines ( European Commission directive 86/609/CEE; French decree no . 87–848 ) . All procedures were approved by the French national ethics committee . Intra-peritoneal injection of Tamoxifen ( 100μl of 1mg/ml ) for four consecutive days was performed on 6–7 week-old animals . After 3 weeks , animals were sacrificed , the tibialis anterior and gastrocnemius muscles were dissected and deletion of Tead4 was verified by PCR genotyping and RNA was prepared . For regeneration , notexin was injected in the tibialis anterior of mice previously treated with Tam . Four subsequent Tam injections were performed following notexin treatment to inactivate Tead4 in the newly forming fibres . Hind-limb grip strength was measured using a Bioseb Grip Strength Meter . Three consecutive readings were performed for each mouse within the same session and the mean value was recorded as the maximal grip strength for each mouse . Body weight was recorded using an electronic balance after sacrificing the mice . The tibialis anterior muscle was then dissected and its mass was measured . The tibialis anterior mass is represented as % of body weight . For fibre cross-section area measurements , transverse cryosections ( 8μm ) of mouse tibialis anterior muscle were stained with hematoxylin and eosin . Slides were scanned using NanoZoomer-XR Digital slide scanner ( Hamamatsu Photonics K . K . ) . Cross-section area was analyzed using the RoiManager plugin of Fiji image analysis software . The conditional Tead4 mutant allele was generated using a targeting construct where Tead4 exons 2 and 3 were flanked by two loxP sites . A neomycin resistance cassette ( PGK-Neo ) flanked by two Frt sites was inserted immediately downstream of the 5’ loxP site . The targeting vector based on pKOII contained a diphtheria toxin A ( DTA ) counter selection cassette to enrich for homologous recombination events . Homology arms were subcloned from cosmids MPMGcPO454Q2 and MPMGc121P0454Q01 ( The German Resource Center for Genome Research ) using the restriction enzymes NaeI and EcoRV ( 3 . 8 kb , short arm ) and EcoRV and KpnI ( 11 . 2kb , long arm ) . The Tead4 targeting construct was electroporated into V6 . 5 F1 hybrid embryonic stem ( ES ) cells [44] after linearization with NotI and subjected to G418 selection . Homologous recombination events in individual ES cell clones were detected by Southern blot analysis of XbaI digested DNA using a probe located outside of the homology arms and by PCR analysis using the following primers: ( 5’-loxP-FW ) AGTGCATGAGGCAAGAGGC , ( 5’-loxP-RV ) GCTCCTGGGACCATAGTTA; ( 3’-loxP-FW ) CAGGCCTCTCTCTGAGGTGA , ( 3’-loxP-RV ) ACTATGAGAGCCTCACAGGC . A positive clone was microinjected into C57BL/6 ( B6 ) blastocysts before transplantation into pseudopregnant foster mothers . Chimeric mice were mated to Flp-expressing transgenic mice to remove the neomycin resistance cassette by Flp-mediated recombination leaving behind a single Flp site and two loxP sites flanking exon 2 and 3 . These mice were the bred with previously described Hsa::CreERT2 mice [32] . C2C12 cells were grown in 20% foetal calf serum ( FCS ) containing DMEM medium and were differentiated for most experiments up to six days in 2% horse serum ( HS ) containing DMEM medium . Adult mouse primary myoblasts were isolated from C57BL/6 wild type 3–4 week-old mice and plated on matrigel-coated dishes . The primary myoblasts were grown in IMDM GLUTAMAX-I medium with 20% FCS and were differentiated in the same medium with 2% HS . The siRNA transfection experiments were performed as per the Lipofectamine RNAiMAX manufacturer’s protocol and cells were harvested at indicated time points of differentiation after the siRNA transfection . ON-TARGET-plus SMARTpool siRNAs for Tead1 , Tead2 and Tead4 knockdown and non-targeting siRNA were purchased from Dharmacon Inc . ( Chicago , Il . , USA ) . The siRNA experiments were performed at least in triplicates . Phase contrast images were taken at 4x magnification using the EVOS digital microscope . A list of all antibodies and primers used can be found in S6 Dataset . Whole cell extracts were prepared by the standard freeze-thaw technique using LSDB 500 buffer ( 500 mM KCl , 25 mM Tris at pH 7 . 9 , 10% glycerol , 0 . 05% NP-40 , 1 mM DTT , and protease inhibitor cocktail ) and Immunoblotting was performed by standard procedure . 1x105 cells were seeded on coverslips in 35mm dishes with matrigel for primary myoblasts and without matrigel for C2C12 cells and were transfected with siRNA 4 hours after seeding . Cells were refreshed 6 to 8 hours after the siRNA treatment and fixed on day 6 of differentiation with 4% formaldehyde for 10 mins . Cells were washed with PBS and permeabilized with 0 . 5% triton for 10 mins , washed twice with PBS-tween 0 . 2% and blocked with 5% BSA for 30 minutes . Cells were incubated with primary antibody overnight at 4°C followed by three PBS-tween 0 . 2% washes . Secondary antibody incubation was done for 30 minutes at room temperature . Cells were washed thrice with PBS-tween 0 , 2% and stained with DAPI . Coverslips were mounted on superfrost glass slides using Vectashield . Slides were visualised using an inverted fluorescence microscope at 10x magnification in all experiments . To quantify the fusion in double and triple knockdown experiments , we calculated the fusion index as the percentage of number of nuclei within the Myh-positive cells above total number of nuclei counted in a field . Nuclei in fields from three replicate experiments were counted and analysed by a two-tailed t-test . Note that Myh positive cells with only 3 nuclei were taken for the counting of the nuclei . Total RNA was extracted using the GenElute Mammalian Total RNA Miniprep Kit from Sigma . cDNA was prepared with using SuperScript II Reverse Transcriptase ( RT ) using the kit protocol and quantitative PCR was carried out with the SYBR Green I ( Qiagen ) and monitored using the Roche Lightcycler 480 . Primer sequences were designed using Primer3plus software and beta-actin was used as normalization control . Messenger-RNA-seq was performed essentially as described [45 , 46] . Sequence reads mapped to reference genome mm9/NCBI37 using Tophat [47] . Data normalization and quantification of gene expression was performed using the DESeq 2 Bioconductor package [48] . Significantly deregulated genes were selected using a log2 fold change >1 and <1 and adjusted p-value <0 , 05 . Gene ontology analyses were performed using the DAVID functional annotation clustering tool available at the website- https://david . ncifcrf . gov/ . For GSEA analyses , we used the mean of the log2 fold changes of the biological replicates as metric for the H Hallmark gene sets of the BROAD javaGSEA tool with 1 , 000 permutations and the canonical pathway ( cp ) subcollection of the C2 curated BROAD molecular signature gene-set collection . Chromatin immunoprecipitation from C2C12 cells was performed by standard procedures as previously described [45] [49] . For ChIP from mouse muscle , muscles harvested from hind limbs of three adult Hsa::Cre-ERT2::Tead4lox/lox mice with or without Tamoxifen injection were either snap frozen or immediately used for ChIP . The tissue was minced and quickly homogenised in cold hypotonic buffer with protease inhibitors using an Ultraturax homogeniser . The homogenised tissue lysate was fixed with 1% formaldehyde in fresh hypotonic buffer for 10 mins shaking at room temperature . Fixation was stopped by adding glycine to 0 . 15M concentration . Lysate was centrifuged at 3000rpm 5 mins at 4°C and the pellet was resuspended in fresh hypotonic buffer ( 25 mM HEPES , pH 7 . 8 , 1 . 5 mM MgCl2 , 10 mM KCl , and 0 . 1% NP-40 ) , supplemented with Protease Inhibitor Cocktail ( Roche , Basel , Switzerland ) . Lysate was filtered to eliminate debris and nuclei were collected using cell strainer of 70 μm pore size . The filtrate was centrifuged for 5 mins at 3000rpm to obtain a nuclear pellet that was resuspended and incubated 10 min at 4°C in sonication buffer ( EDTA 10mM , Tris-HCl , pH 8 . 0 , 50mM , SDS 1% with protease inhibitor cocktail and PMSF ) and then sonicated using Covaris sonicator for 20 to 25 mins . Lysate was then centrifuged for 15 mins at 11000g at 4°C to obtain the chromatin supernatant fraction that was the used for ChIP . ChIP-seq libraries were prepared and sequenced on an Illumina Hi-seq2500 as single-end 50-base reads . After sequencing , peak detection was performed using the MACS software [50] http://liulab . dfci . harvard . edu/MACS/ ) . Global clustering , meta-analyses and quantitative comparisons were performed using seqMINER and R ( http://www . r-project . org/ ) . Peaks were annotated with Homer ( http://homer . salk . edu/homer/ngs/annotation . html ) using a window of ±50 kb ( or as nearest gene ) relative to the transcription start site of RefSeq transcripts . De novo motif discovery was performed on the 200 base pairs surrounding the top 600 Tead1 and Tead4 peaks using MEME-ChIP . Motif enrichment analyses were performed using in house algorithms as described [49] . The public data for H3K27ac and H3K4me3 data were taken from the GEO accession GSE25308 [51] , Jun GSE37525 , Srf , GSM915168 , Runx , GSM1354734 . Myod1 and Myog ChIP-seq raw data were from GSE44824 [27] and re-analyzed in parallel to the Tead4 and Tead1 ChIP-seq data . The data in this paper have been assigned the accession number GSE82193 in the GEO database .
|
Aspects of muscle differentiation can be reproduced using the C2C12 cell line or primary myoblasts both of which can be differentiated to form myotubes in vitro . While the functions of recognised myogenic proteins such as Myogenin , Myod1 and MEF-family transcription factors in this process have been extensively studied , the role of the Tead factors has received only limited attention . Tead factors have well defined roles as mediators of Hippo signalling in promoting cell growth and oncogenic transformation , but are also involved in myogenic differentiation involving cell cycle arrest and activation of the myogenic gene expression program . Using integrative genomics and knockdowns in cell based models , we show that Tead factors are essential for differentiation of C2C12 cells and primary myoblasts , but make different contributions activating a distinct set of myogenesis genes in each cell type . We also developped effective chromatin immunoprecipitation from mature mouse muscle fibres allowing identification of highly transcribed muscle identify genes and identification of Tead1 and Tead4 occupied sites . Somatic inactivation in vivo revealed an important role for Tead4 in muscle fibre regeneration . The integration of genomics and loss of function in cell models in vitro and muscle in vivo provide the first comprehensive description Tead factors in myogenic differentiation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"gene",
"regulation",
"cell",
"differentiation",
"muscle",
"regeneration",
"developmental",
"biology",
"organism",
"development",
"stem",
"cells",
"morphogenesis",
"small",
"interfering",
"rnas",
"muscle",
"differentiation",
"animal",
"cells",
"gene",
"expression",
"myoblasts",
"genetic",
"loci",
"biochemistry",
"rna",
"regeneration",
"nucleic",
"acids",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"non-coding",
"rna"
] |
2017
|
TEAD transcription factors are required for normal primary myoblast differentiation in vitro and muscle regeneration in vivo
|
In Drosophila , multiple lines of evidence converge in suggesting that beneficial substitutions to the genome may be common . All suffer from confounding factors , however , such that the interpretation of the evidence—in particular , conclusions about the rate and strength of beneficial substitutions—remains tentative . Here , we use genome-wide polymorphism data in D . simulans and sequenced genomes of its close relatives to construct a readily interpretable characterization of the effects of positive selection: the shape of average neutral diversity around amino acid substitutions . As expected under recurrent selective sweeps , we find a trough in diversity levels around amino acid but not around synonymous substitutions , a distinctive pattern that is not expected under alternative models . This characterization is richer than previous approaches , which relied on limited summaries of the data ( e . g . , the slope of a scatter plot ) , and relates to underlying selection parameters in a straightforward way , allowing us to make more reliable inferences about the prevalence and strength of adaptation . Specifically , we develop a coalescent-based model for the shape of the entire curve and use it to infer adaptive parameters by maximum likelihood . Our inference suggests that ∼13% of amino acid substitutions cause selective sweeps . Interestingly , it reveals two classes of beneficial fixations: a minority ( approximately 3% ) that appears to have had large selective effects and accounts for most of the reduction in diversity , and the remaining 10% , which seem to have had very weak selective effects . These estimates therefore help to reconcile the apparent conflict among previously published estimates of the strength of selection . More generally , our findings provide unequivocal evidence for strongly beneficial substitutions in Drosophila and illustrate how the rapidly accumulating genome-wide data can be leveraged to address enduring questions about the genetic basis of adaptation .
A central challenge of evolutionary biology is to elucidate the nature of adaptive changes to the genome: do they comprise a negligible or substantial fraction of differences among species ? When they occur , are they driven by strong positive selection or are they fine-tunings of minor consequence to fitness ? In Drosophila , perhaps the most studied taxon in these respects , there are conflicting accounts regarding the intensity of selection driving adaptations [1]–[4] but accumulating lines of evidence suggest that adaptation may be prevalent [5]–[7] . The evidence is based primarily on two kinds of signatures that beneficial substitutions leave in their wake . The first is an excess of divergence at functional sites compared to that expected under neutrality , detected using the approach introduced by McDonald and Kreitman [8]–[11] . Numerous studies based on extensions of this approach indicate that approximately one in two amino acid and one in five non-coding differences between Drosophila species may be adaptive [7] , [11]–[14] . These findings remain tentative , however , because other factors , and notably plausible demographic scenarios , could cause a substantial overestimation of the fraction of beneficial substitutions [7] , [8] , [15]–[17] . Moreover , McDonald-Kreitman based approaches can provide only very limited information about the strength of positive selection . The second footprint of adaptation is in diversity patterns . When a rare or new allele is favored and fixes in the population , it drags closely linked neutral alleles to loss or fixation . This “selective sweep” leads to a transient reduction in levels of neutral diversity around a beneficial substitution , where the size of the affected region decreases with the recombination rate and increases with the intensity of positive selection [18]–[20] . In accordance with a model of recurrent selective sweeps , levels of synonymous diversity across the genomes of a number of Drosophila species increase with rates of crossing over [21]–[23] and decrease with increasing numbers of amino acid substitutions [2] , [3] . Making reliable inferences about adaptation based on these relationships has been challenging , with two decades of effort focused on distinguishing the effects of positive selection from those of background ( i . e . , purifying ) selection and from possible mutagenic effects of recombination [5] , [24]–[29] . By necessity , previous studies relied on limited summaries of the data , thereby losing much of the information carried by the spatial signature of beneficial fixations . In particular , measurements of diversity , recombination , and functional divergence were taken in arbitrarily chosen window sizes , making it harder to distinguish the effects of adaptation from other evolutionary forces [29] , [30] , and likely biasing estimates of adaptive parameters of interest ( e . g . , the rate and intensity of selection ) [7] . As an illustration , based on the relationship between diversity levels and amino acid divergence seen in 100 kb windows , Macpherson et al . [3] inferred few beneficial amino acid substitutions with a large selective coefficient of ∼1%; in contrast , focusing on the same relationship in individual genes , Andolfatto [2] inferred many beneficial amino substitutions with a selective coefficient of ∼10−3%; the two studies differed in other regards , but the disparate conclusions may reflect in part the choice of window size [7] . In summary , despite accumulating evidence that adaptation may be widespread in Drosophila , we still lack characterizations that capture genome-wide signatures that are specific to adaptive evolution and do not rely on an a priori choice of scale .
Here , we take advantage of genome-wide variation data from Drosophila in order to produce a readily interpretable characterization of the effects of positive selection that overcomes a number of limitations . To do so , we consider the average level of neutral diversity as a function of distance from amino acid substitutions . Our reasoning is as follows: Beneficial amino acids that fixed in the recent evolutionary past ( ∼Ne generations [20] ) should create a trough in diversity levels around them , whereas amino acid substitutions that were selectively neutral or occurred farther in the past should have little effect on diversity patterns . If we consider the effects of all amino acid substitutions in the genome jointly , and a non-negligible fraction of amino acid fixations were favored – as McDonald-Kreitman based estimates suggest – then we should expect a trough in the average level of neutral diversity around amino acid substitutions . The depth of this trough is expected to increase with the fraction of beneficial amino acid substitutions , and its width will reflect the intensity of selection driving these substitutions . In contrast to previous approaches , this characterization does not depend on an a priori choice of window size , and captures much more of the footprint of adaptive substitutions . To generate this plot , we use autosomal amino acid substitutions on the lineage leading from the common ancestor of Drosophila simulans and D . melanogaster to D . simulans , relying on the genomes of D . erecta and D . yakuba as outgroups [31] . As a measure of neutral diversity , we consider the number of synonymous polymorphisms divided by the overall number of codons at a given distance from an amino acid substitution . The polymorphism levels in D . simulans are measured using a recent dataset of six inbred lines [5] , down-sampled to have a uniform sample size of 4 lines at ∼50% of the codons in the genome . Ideally , we would like to plot diversity levels as a function of genetic distance from amino acid substitutions , since the expected reduction in diversity depends on genetic rather than physical distance from the selected loci . Since there are no high-resolution estimates of recombination rates in D . simulans , we use physical distance instead , but consider only regions for which the homologous regions in D . melanogaster have an estimated recombination rate above 0 . 75cM/Mb . The collated plot in Figure 1A ( red ) thus obtained is averaged over n = 26 , 834 amino acid substitutions . Because the plot is constructed by conditioning on a substitution at the center , diversity patterns could be distorted even in the absence of adaptive evolution . Namely , if mutation rates vary across the genome then they might , on average , be elevated near substitutions . Considering the average synonymous divergence between D . melanogaster and D . yakuba as a proxy for the mutation rate confirms this expectation , as it reveals a small increase near substitutions ( Figure 1B ) . To correct for this elevation in rates , we divide the average level of diversity around amino acid substitutions at a given distance by the average divergence ( Figure 1C ) . Moreover , as a control , we compare the patterns around amino acid substitutions with plots that were constructed analogously but around synonymous substitutions instead ( Figure 1A–1C: black ) [28] . As predicted by a model of recurrent selective sweeps , we find a clear reduction in diversity levels around amino acid substitutions relative to the synonymous control . This reduction is statistically significant within a window of ∼15kb around amino acid substitutions ( at the 1% level , as assessed by bootstrapping; see Text S1 ) . Farther from substitutions , where sweeps are unlikely to have an effect on diversity , the curves for synonymous and amino acid substitutions are indistinguishable . This pattern is robust to the effects of synonymous codon usage bias ( Figure 4 in Text S1 ) , as well as to changes in the recombination rate threshold ( Figure 5 in Text S1 ) , and to the choice of outgroup used to correct for the mutation rate ( not shown ) . In addition , we see similar patterns when we examine the substitutions that occur on any one of the autosomal chromosome arms ( Figure 6 in Text S1 ) . This pattern is a distinctive signature of adaptive evolution . Demographic processes would not lead to systematically decreased diversity around amino acid substitutions . In turn , for background selection to generate the observed trough centered on amino acid substitutions , its effects in regions of the genome with moderate to high recombination rates would have to be strong enough to lead to both a substantial reduction in diversity and to the fixation of many weakly deleterious amino acid mutations . Modeling indicates that , given plausible parameters for Drosophila , this is highly unlikely [32] . Our analyses also reveal that amino acid substitutions are clustered near one another ( Figure 2A: red ) . This clustering is greater and more localized than the clustering of synonymous substitutions around amino acid substitutions ( Figure 2A: black ) , implying that it is caused by more than the spatial distribution of exons in the genome and an elevated mutation rate near amino acid substitutions . The difference between the clustering of amino acid and synonymous substitutions further suggests that variation in constraint and possibly in adaptability among and within genes contribute to the pattern for amino acid substitutions ( [33]; also see Text S1 ) . Aside from being an interesting finding in itself , this clustering could influence the observed reduction in diversity . If two amino acid substitutions occur in close proximity and one led to a recent selective sweep , the reduction in diversity that it caused will also be observed around the other substitution . This effect will reduce diversity around both non-synonymous and synonymous substitutions , but it will have a larger effect around amino acid substitutions because the density of amino acid substitutions nearby is on average greater ( Figure 2A ) . Indeed , the level of synonymous diversity decreases strongly with the density of amino acid substitutions surrounding a substitution ( Figure 2B; Figure 8 in Text S1; Spearman's ρ = −0 . 93 for amino acid substitutions and ρ = −0 . 88 for synonymous substitutions; p<10−15 for both ) , consistent with previous studies [2] , [3] . We also find , however , that the average level of synonymous diversity around amino acid substitutions is consistently lower than that around synonymous substitutions when the two are matched for the density of amino acid substitutions in their vicinity ( Figure 2B; Figure 8 in Text S1; signs test p<10−4 ) . In other words , there is a substantial relative reduction in diversity around amino acid substitutions that is not explained by the amplifying effects of clustering . In addition to providing compelling evidence for the prevalence of beneficial amino acid substitutions , the collated plot carries information about selection parameters , as the shape of the trough in diversity is indicative of the rate of adaptive protein evolution and of the distribution of selective effects of fixations . To learn about these parameters , we develop a coalescent-based model for average diversity levels as a function of distance from an amino acid substitution , accounting for their clustering ( see Text S1 ) . Using this model , we infer adaptive parameters by jointly maximizing the composite-likelihood of diversity patterns as a function of different distances from the focal substitution ( i . e . , the likelihood of points along the entire curve ) , thus mining a richer summary of the data than previous approaches . When we assume that a fraction α of beneficial substitutions were driven by a selection coefficient s and the rest were neutral , we estimate that ∼5% of the substitutions were beneficial with a relatively strong selection coefficient of ∼0 . 4% ( Table 5 in Text S1 ) . Using a Gamma distribution for the selection coefficients , α increases to ∼6 . 5% and the average selection coefficient remains similarly high; despite the additional parameter , the likelihood is barely higher ( Table 5 in Text S1 ) . These estimates are relatively insensitive to assumptions about other parameters ( with the exception of the assumptions about recombination rates , as discussed below ) ; in particular , simulations suggest that the estimated strength of selection is robust to demographic assumptions ( see Text S1 for details ) . A visual comparison suggests a reasonable fit of these models to the data ( Figure 3A ) . However , the inference based on models with one selection coefficient , or even a Gamma distribution of coefficients , might be dominated by the broad features of the plot , such that any narrower trough caused by beneficial substitutions with weaker selection coefficients could be overlooked . A closer look around the focal substitutions supports this notion , revealing a small trough inside the main trough , on the scale of several hundred bps , which is not captured by either of the two models ( Figure 3B ) . We therefore consider another model , with two beneficial selection coefficients . Using it , we estimate that ∼13% of the substitutions were beneficial , ∼3% with a large selective advantage of ∼0 . 5% and the rest with a much weaker effect , of approximately one hundredth of a percent ( Table 5 in Text S1 ) . A mixture model with two exponentials reveals a similar picture: ∼4% of substitutions are estimated to come from a distribution with a mean selective coefficient of ∼0 . 5% and 11% from a distribution with a mean of ∼4·10−5 ( Table 5 in Text S1 ) . Importantly , both models provide a substantially better fit to the data ( Table 5 in Text S1 ) and they capture the smaller as well as the larger troughs in diversity ( Figure 3A and 3B ) . In turn , estimates under a model with three beneficial selective coefficients are similar to those obtained in model with only two and offer no improvement to the fit ( Table 5 in Text S1 ) . Taken together , these findings indicate that selective sweeps are driven by two classes of beneficial fixations: a minority with large beneficial effects that account for most of the reduction in diversity and a majority with much weaker effects . Moreover , they help explain why previous inferences based on the signatures of sweeps in Drosophila yielded markedly different estimates ( ranging over three orders of magnitudes ) [1]–[4] . Our estimates of the fraction of beneficial amino acid substitutions ( ∼13% ) are on the same order of magnitude but lower than previous McDonald-Kreitman based estimates ( ∼50%; cf . [7] ) . Some of this difference might arise from violations of the assumptions on which the inferences rely; in particular , in our approach , that adaptive parameters have remained constant in the D . simulans lineage , or in McDonald-Krietman based inferences , that the efficacy of purifying selection has not changed markedly [8] , [16] , [34] . An intriguing alternative is that the two approaches are actually estimating parameters of somewhat different modes of adaptation . Our inference is based on the effects of beneficial substitutions that arise from new mutations and likely misses some contribution of adaptation from standing variation . Specifically , a subset of beneficial substitutions could stem from previously neutral or deleterious alleles that were segregating in the population before a change in the environment rendered them beneficial . If these alleles were young when the environment changed , they would still generate the signature of a selective sweep and contribute , at least partially , to our estimated fraction of beneficial substitutions . This is likely for alleles that were previously deleterious and at mutation-selection balance , but also possible for neutral alleles [35]–[37] . If , however , the segregating alleles were older when they became beneficial and at higher frequency in the population , they would lead to a negligible effect on diversity and would therefore not contribute to the signature on which our inference relies . These beneficial substitutions would nonetheless contribute to an excess of non-synonymous divergence compared to the neutral expectation , and should therefore be picked by the McDonald-Kreitman based inferences , leading to higher estimates of adaptive substitutions than obtained by our approach . Other modes of adaptation , such as polygenic selection , may also contribute differentially to the two inference methodologies [38] . We note that a current limitation of our inference is its reliance on rough estimates of the recombination rate , and its assumption of a constant rate per base . In the logistic approximation to the trajectory of a beneficial allele , the expected reduction in diversity as a function of distance from the beneficial substitution depends on s/r , where s is the selection coefficient and r is the genetic distance to the substitution ( Equation 2 in Text S1 ) . This implies , for example , that if our inference relies on a recombination rate consistently two-fold greater than the real rate , our estimated selection coefficient will be two-fold overestimated ( see Table 3 in Text S1 ) . We therefore consider our estimates of selection coefficients to be rough approximations . In addition , heterogeneity in the recombination rate , such as is known to exist in other taxa ( e . g . , [39] , [40] ) , could also affect our inferences . The heterogeneity would have to be of a highly specific nature in order to account for our finding of two markedly different scales of selection coefficients , but at the moment , we cannot rule out the possibility . For these reasons , it would be important to revisit the inference once we possess high-resolution genetic maps in D . simulans . In summary , our findings establish a distinctive , genome-wide signature of adaptation in D . simulans , suggesting that many amino acid substitutions are beneficial and are driven by two classes of selective effects . Enabled by a richer summary of diversity patterns that avoids an a priori choice of scale , these conclusions offer a coherent interpretation of the results of previous inferences . It will now be interesting to see whether similar findings emerge in other Drosophila species , which vary in their recombination rates , effective population sizes , and ecology .
We reconstructed the sequence of the ancestor of D . melanogaster and D . simulans in order to identify substitutions along the D . simulans lineage . For that purpose , we use a four species alignement from the 12 Drosophila genomes project [31] consisting of D . simulans , D . melanogaster , D . yakuba and D . erecta , and removed codons containing gaps in either of them . We then inferred the ancestral sequences using PAML , with the CODEML model and the ( ( D . mel , D . sim ) , ( D . yak , D . ere ) ) tree [41] . To measure polymorphism levels at coding regions of the D . simulans genome , we used resequencing data from six inbred lines of D . simulans and their alignment with D . melanogaster [5] . We applied quality control filters and randomly down-sampled the remaining codons to four , in order to maintain a uniform sample size in measuring polymorphism . In the end , we retained ∼50% of all protein-coding DNA . Unless otherwise noted , our analysis was performed on data from autosomal regions , for which the sex-averaged recombination rate in the homologous region of D . melanogaster was greater than 0 . 75cM/Mb ( using the genetic map as in [3] ) . See Section 1 in Text S1 for more details . We used synonymous polymorphisms to measure the average levels of diversity as a function of distance from amino acid and synonymous substitutions along the D . simulans lineage . To measure the average level of diversity at distance x , we divided the number of codons segregating for a synonymous polymorphism by the overall number of codons observed in the D . simulans polymorphism dataset at distance x from one of the amino acid ( or synonymous ) substitution . In order to control for variation in the neutral mutation rate around substitutions , we calculated the average synonymous divergence around both amino acid and synonymous substitutions . For that purpose , we identified synonymous substitutions between D . melanogaster and D . yakuba and measured the average level of divergence at distance x by dividing the number of codons exhibiting a synonymous substitution between D . melanogaster and D . yakuba by the overall number of codons observed in the alignment of these species at distance x from one of the amino acid ( or synonymous ) substitutions . For further details and the robustness analysis , see Sections 2–4 in Text S1 . The shape of the collated plot around amino acid substitutions carries information about the rate of adaptive protein evolution and the intensity of selection driving it , two parameters of long-standing interest . To learn about these parameters , we developed a model describing the expected neutral diversity levels around substitutions , which relies on Gillespie's pseudohitchhiking coalescent model [42] . We then used a composite likelihood approach [43] to estimate the parameters . For a description of the approach and assessments of its reliability , see Section 6 in Text S1 .
|
Characterizing the nature of beneficial changes to the genome is essential to our understanding of adaptation . To do so , researchers identify and analyze footprints that beneficial changes leave in patterns of genetic variation within and between species . In order to teach us about adaptive evolution , these footprints need to be specific to positive selection as well as rich enough to allow for reliable inferences . Here , we identify such a footprint: a pronounced trough in the average levels of genetic diversity surrounding amino acid substitutions throughout the D . simulans genome . Based on this pattern , we infer that approximately 13% of amino acid substitutions were beneficial , a minority of which ( 3% ) conferred a large selective advantage of nearly 0 . 5% and the majority of which ( 10% ) conferred a much smaller advantage of about 0 . 01% . These findings offer insights into the distribution of selection effects driving beneficial changes to the D . simulans genome and suggest how the widely varying estimates obtained in previous studies of Drosophila may be reconciled . Moreover , the approach that we introduce is readily applicable to other taxa and thus should help to gain important insights into how the rate and strength of adaptive evolution vary depending on life-history , population size , and ecology .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"evolutionary",
"biology",
"genetics",
"and",
"genomics"
] |
2011
|
Pervasive Adaptive Protein Evolution Apparent in Diversity Patterns around Amino Acid Substitutions in Drosophila simulans
|
There is a popular belief in neuroscience that we are primarily data limited , and that producing large , multimodal , and complex datasets will , with the help of advanced data analysis algorithms , lead to fundamental insights into the way the brain processes information . These datasets do not yet exist , and if they did we would have no way of evaluating whether or not the algorithmically-generated insights were sufficient or even correct . To address this , here we take a classical microprocessor as a model organism , and use our ability to perform arbitrary experiments on it to see if popular data analysis methods from neuroscience can elucidate the way it processes information . Microprocessors are among those artificial information processing systems that are both complex and that we understand at all levels , from the overall logical flow , via logical gates , to the dynamics of transistors . We show that the approaches reveal interesting structure in the data but do not meaningfully describe the hierarchy of information processing in the microprocessor . This suggests current analytic approaches in neuroscience may fall short of producing meaningful understanding of neural systems , regardless of the amount of data . Additionally , we argue for scientists using complex non-linear dynamical systems with known ground truth , such as the microprocessor as a validation platform for time-series and structure discovery methods .
The MOS 6502 ( and the virtually identical MOS 6507 ) were the processors in the Apple I , the Commodore 64 , and the Atari Video Game System ( VCS ) ( see [10] for a comprehensive review ) . The Visual6502 team reverse-engineered the 6507 from physical integrated circuits [11] by chemically removing the epoxy layer and imaging the silicon die with a light microscope . Much like with current connectomics work [12 , 13] , a combination of algorithmic and human-based approaches were used to label regions , identify circuit structures , and ultimately produce a transistor-accurate netlist ( a full connectome ) for this processor consisting of 3510 enhancement-mode transistors . Several other support chips , including the Television Interface Adaptor ( TIA ) were also reverse-engineered and a cycle-accurate simulator was written that can simulate the voltage on every wire and the state of every transistor . The reconstruction has sufficient fidelity to run a variety of classic video games , which we will detail below . The simulation generates roughly 1 . 5GB/sec of state information , allowing a real big-data analysis of the processor . The simplicity of early video games has led to their use as model systems for reinforcement learning [14] and computational complexity research [15] . The video game system ( “whole animal” ) has a well defined output in each of the three behavioral conditions ( games ) . It produces an input-dependent output that is dynamic , and , in the opinion of the authors , quite exciting . It can be seen as a more complex version of the Mus Silicium project [16] . It is also a concrete implementation of a thought experiment that has been mentioned on and off in the literature [17–20] . The richness of the dynamics and our knowledge about its inner workings makes it an attractive test case for approaches in neuroscience . Here we will examine three different “behaviors” , that is , three different games: Donkey Kong ( 1981 ) , Space Invaders ( 1978 ) , and Pitfall ( 1981 ) . Obviously these “behaviors” are qualitatively different from those of animals and may seem more complicated . However , even the simple behaviors that are studied in neuroscience still involve a plethora of components , typically including the allocation of attention , cognitive processing , and multiple modalities of inputs and outputs . As such , the breadth of ongoing computation in the processor may actually be simpler than those in the brain . The objective of clever experimental design in neuroscience often is to find behaviors that only engage one kind of computation in the brain . In the same way , all our experiments on the chip will be limited by us only using these games to probe it . As much as more neuroscience is interested in naturalistic behaviors [21] , here we analyze a naturalistic behavior of the chip . In the future it may be possible to excute simpler , custom code on the processor to tease apart aspects of computation , but we currently lack such capability in biological organisms . Much has been written about the differences between computation in silico and computation in vivo [22 , 23]—the stochasticity , redundancy , and robustness [24] present in biological systems seems dramatically different from that of a microprocessor . But there are many parallels we can draw between the two types of systems . Both systems consist of interconnections of a large number of simpler , stereotyped computing units . They operate on multiple timescales . They consist of somewhat specialized modules organized hierarchically . They can flexibly route information and retain memory over time . Despite many differences there are also many similarities . We do not wish to overstate this case—in many ways , the functional specialization present in a large mammalian brain far eclipses that present in the processor . Indeed , the processor’s scale and specialization share more in common with C . elegans than a mouse . Yet many of the differences should make analysing the chip easier than analyzing the brain . For example , it has a clearer architecture and far fewer modules . The human brain has hundreds of different types of neurons and a similar diversity of proteins at each individual synapse [25] , whereas our model microprocessor has only one type of transistor ( which has only three terminals ) . The processor is deterministic while neurons exhibit various sources of randomness . With just a couple thousand transistors it is also far smaller . And , above all , in the simulation it is fully accessible to any and all experimental manipulations that we might want to do on it . Importantly , the processor allows us to ask “do we really understand this system ? ” Most scientists have at least behavioral-level experience with these classical video game systems , and many in our community , including some electrophysiologists and computational neuroscientists , have formal training in computer science , electrical engineering , computer architecture , and software engineering . As such , we believe that most neuroscientists may have better intuitions about the workings of a processor than about the workings of the brain . What constitutes an understanding of a system ? Lazbnick’s original paper argued that understanding was achieved when one could “fix” a broken implementation . Understanding of a particular region or part of a system would occur when one could describe so accurately the inputs , the transformation , and the outputs that one brain region could be replaced with an entirely synthetic component . Indeed , some neuroengineers are following this path for sensory [26] and memory [27] systems . Alternatively , we could seek to understand a system at differing , complementary levels of analysis , as David Marr and Tomaso Poggio outlined in 1982 [28] . First , we can ask if we understand what the system does at the computational level: what is the problem it is seeking to solve via computation ? We can ask how the system performs this task algorithmically: what processes does it employ to manipulate internal representations ? Finally , we can seek to understand how the system implements the above algorithms at a physical level . What are the characteristics of the underlying implementation ( in the case of neurons , ion channels , synaptic conductances , neural connectivity , and so on ) that give rise to the execution of the algorithm ? Ultimately , we want to understand the brain at all these levels . In this paper , much as in systems neuroscience , we consider the quest to gain an understanding of how circuit elements give rise to computation . Computer architecture studies how small circuit elements , like registers and adders , give rise to a system capable of performing general-purpose computation . When it comes to the processor , we understand this level extremely well , as it is taught to most computer science undergraduates . Knowing what a satisfying answer to “how does a processor compute ? ” looks like makes it easy to evaluate how much we learn from an experiment or an analysis . We can draw from our understanding of computer architecture to firmly ground what a full understanding of a processor would look like ( Fig 1 ) . The processor is used to implement a computing machine . It implements a finite state machine which sequentially reads in an instruction from memory ( Fig 1a , green ) and then either modifies its internal state or interacts with the world . The internal state is stored in a collection of byte-wide registers ( Fig 1a , red ) . As an example , the processor might read an instruction from memory telling it to add the contents of register A to the contents of register B . It then decodes this instruction , enabling the arithmetic logic unit ( ALU , Fig 1a , blue ) to add those registers , storing the output . Optionally , the next instruction might save the result back out to RAM ( Fig 1a , yellow ) . It is this repeated cycle that gives rise to the complex series of behaviors we can observe in this system . Note that this description in many ways ignores the functions of the individual transistors , focusing instead on circuits modules like “registers” which are composed of many transistors , much as a systems neuroscientist might focus on a cytoarchitecturally-distinct area like hipppocampus as opposed to individual neurons . Each of the functions within the processor contains algorithms and a specific implementation . Within the arithmetic logic unit , there is a byte wide adder , which is in part made of binary adders ( Fig 1b ) , which are made out of AND/NAND gates , which are made of transistors . This is in a similar way as the brain consists of regions , circuits , microcircuits , neurons , and synapses . If we were to analyze a processor using techniques from systems neuroscience we would hope that it helps guide us towards the descriptions that we used above . In the rest of the paper we will apply neuroscience techniques to data from the processor . We will finally discuss how neuroscience can work towards techniques that will make real progress at moving us closer to a satisfying understanding of computation , in the chip , and in our brains .
The earliest investigations of neural systems were in-depth anatomical inquiries [29] . Fortunately , through large scale microscopy ( Fig 2a ) we have available the full 3d connectome of the system . In other words , we know how each transistor is connected to all the others . The reconstruction is so good , that we can now simulate this processor perfectly—indeed , were it not for the presence of the processor’s connectome , this paper would not have been possible . This process is aided by the fact that we know a transistor’s deterministic input-output function , whereas neurons are both stochastic and vastly more complex . Recently several graph analysis methods ranging from classic [30] to modern [31 , 32] approaches have been applied to neural connectomes . The approach in [31] was also applied to a region of this processor , attempting to identify both circuit motifs as well as transistor “types” ( analogous to cell types ) in the transistor wiring diagram . Fig 3 ( adapted from [31] ) shows the results of the analysis . We see that one identified transistor type contains the “clocked” transistors , which retain digital state . Two other types contain transistors with pins C1 or C2 connected to ground , mostly serving as inverters . An additional identified type controls the behavior of the three registers of interest ( X , Y , and S ) with respect to the SB data bus , either allowing them to latch or drive data from the bus . The repeat patterns of spatial connectivity are visible in Fig 3a , showing the man-made horizontal and vertical layout of the same types of transistors . While superficially impressive , based on the results of these algorithms we still can not get anywhere near an understanding of the way the processor really works . Indeed , we know that for this processor there is only one physical “type” of transistor , and that the structure we recover is a complex combination of local and global circuitry . In neuroscience , reconstructing all neurons and their connections perfectly is the dream of a large community studying connectomics [33 , 34] . Current connectomics approaches are limited in their accuracy and ability to definitively identify synapses [13] , Unfortunately , we do not yet have the techniques to also reconstruct the i/o function–neurotransmitter type , ion channel type , I/V curve of each synapse , etc . —of each neuron . But even if we did , just as in the case of the processor , we would face the problem of understanding the brain based on its connectome . As we do not have algorithms that go from anatomy to function at the moment that go considerably beyond cell-type clustering [31 , 35 , 36] it is far from obvious how a connectome would allow an understanding of the brain . Note we are not suggesting connectomics is useless , quite the contrary–in the case of the processor the connectome was the first crucial step in enabling reliable , whole-brain-scale simulation . But even with the whole-brain connectome , extracting hierarchical organization and understanding the nature of the underlying computation is incredibly difficult . Lesions studies allow us to study the causal effect of removing a part of the system . We thus chose a number of transistors and asked if they are necessary for each of the behaviors of the processor ( Fig 4 . In other words , we asked if removed each transistor , if the processor would then still boot the game . Indeed , we found a subset of transistors that makes one of the behaviors ( games ) impossible . We can thus conclude they are uniquely necessary for the game—perhaps there is a Donkey Kong transistor or a Space Invaders transistor . Even if we can lesion each individual transistor , we do not get much closer to an understanding of how the processor really works . This finding of course is grossly misleading . The transistors are not specific to any one behavior or game but rather implement simple functions , like full adders . The finding that some of them are important while others are not for a given game is only indirectly indicative of the transistor’s role and is unlikely to generalize to other games . Lazebnik [9] made similar observations about this approach in molecular biology , suggesting biologists would obtain a large number of identical radios and shoot them with metal particles at short range , attempting to identify which damaged components gave rise to which broken phenotype . This example nicely highlights the importance of isolating individual behaviors to understand the contribution of parts to the overall function . If we had been able to isolate a single function , maybe by having the processor produce the same math operation every single step , then the lesioning experiments could have produced more meaningful results . However , the same problem exists in neuroscience . It is extremely difficult or technically impossible to produce behaviors that only require a single aspect of the brain . Beyond behavioral choices , we have equivalent problems in neuroscience that make the interpretation of lesioning data complicated [37] . In many ways the chip can be lesioned in a cleaner way than the brain: we can individually abolish every single transistor ( this is only now becoming possible with neurons in simple systems [38 , 39] ) . Even without this problem , finding that a lesion in a given area abolishes a function is hard to interpret in terms of the role of the area for general computation . And this ignores the tremendous plasticity in neural systems which can allow regions to take over for damaged areas . In addition to the statistical problems that arise from multiple hypothesis testing , it is obvious that the “causal relationship” we are learning is incredibly superficial: a given transistor is obviously not specialized for Donkey Kong or Space Invaders . While in most organisms individual transistors are not vital , for many less-complex systems they are . Lesion individual interneurons in C . elegans or the H1 neuron in the fly can have marked behavioral impacts . And while lesioning larger pieces of circuitry , such as the entire TIA graphics chip , might allow for gross segregation of function , we take issue with this constituting “understanding” . Simply knowing functional localization , at any spatial scale , is only the most nacent step to the sorts of understanding we have outlined above . We may want to try to understand the processor by understanding the activity of each individual transistor . We study the “off-to-on” transition , or “spike” , produced by each individual transistor . Each transistor will be activated at multiple points in time . Indeed , these transitions look surprisingly similar to the spike trains of neurons ( Fig 5 ) . Following the standards in neuroscience we may then quantify the tuning selectivity of each transistor . For each of our transistors we can plot the spike rate as a function of the luminance of the most recently displayed pixel ( Fig 6 ) . For a small number of transistors we find a strong tuning to the luminance of the most recently displayed pixel , which we can classify into simple ( Fig 6a ) and ( Fig 6b ) complex curves . Interestingly , however , we know for each of the five displayed transistors that they are not directly related to the luminance of the pixel to be written , despite their strong tuning . The transistors relate in a highly nonlinear way to the ultimate brightness of the screen . As such their apparent tuning is not really insightful about their role . In our case , it probably is related to differences across game stages . In the brain a neuron can calculate something , or be upstream or downstream of the calculation and still show apparent tuning making the inference of a neurons role from observational data very difficult [40] . This shows how obtaining an understanding of the processor from tuning curves is difficult . Much of neuroscience is focused on understanding tuning properties of neurons , circuits , and brain areas [41–44] . Arguably this approach is more justified for the nervous system because brain areas are more strongly modular . However , this may well be an illusion and many studies that have looked carefully at brain areas have revealed a dazzling heterogeneity of responses [45–47] . Even if brain areas are grouped by function , examining the individual units within may not allow for conclusive insight into the nature of computation . Moving beyond correlating single units with behavior , we can examine the correlations present between individual transistors . We thus perform a spike-word analysis [48] by looking at “spike words” across 64 transistors in the processor . We find little to very weak correlation among most pairs of transistors ( Fig 7a ) . This weak correlation suggests modeling the transistors’ activities as independent , but as we see from shuffle analysis ( Fig 7b ) , this assumption fails disastrously at predicting correlations across many transistors . In neuroscience , it is known that pairwise correlations in neural systems can be incredibly weak , while still reflecting strong underlying coordinated activity . This is often assumed to lead to insights into the nature of interactions between neurons [48] . However , the processor has a very simple nature of interactions and yet produces remarkably similar spike word statistics . This again highlights how hard it is to derive functional insights from activity data using standard measures . The activity of the entire chip may be high dimensional , yet we know that the chip , just like the brain , has some functional modularity . As such , we may be able to understand aspects of its function by analyzing the average activity within localized regions , in a way analogous to the local field potentials or the BOLD signals from functional magnetic imaging that are used in neuroscience . We thus analyzed data in spatially localized areas ( Fig 8a ) . Interestingly , these average activities look quite a bit like real brain signals ( Fig 8b ) . Indeed , they show a rather similar frequency power relation of roughly power-law behavior . This is often seen as a strong sign of self-organized criticality [49] . Spectral analysis of the time-series reveals region-specific oscillations or “rhythms” that have been suggested to provide a clue to both local computation and overall inter-region communication . In the chip we know that while the oscillations may reflect underlying periodicity of activity , the specific frequencies and locations are epiphenomena . They arise as an artifact of the computation and tell us little about the underlying flow of information . And it is very hard to attribute ( self-organized ) criticality to the processor . In neuroscience there is a rich tradition of analyzing the rhythms in brain regions , the distribution of power across frequencies as a function of the task , and the relation of oscillatory activity across space and time . However , the example of the processor shows that the relation of such measures to underlying function can be extremely complicated . In fact , the authors of this paper would have expected far more peaked frequency distributions for the chip . Moreover , the distribution of frequencies in the brain is often seen as indicative about the underlying biophysics . In our case , there is only one element , the transistor , and not multiple neurotransmitters . And yet , we see a similarly rich distribution of power in the frequency domain . This shows that complex multi-frequency behavior can emerge from the combination of many simple elements . Analyzing the frequency spectra of artifacts thus leads us to be careful about the interpretation of those occurring in the brain . Modeling the processor as a bunch of coupled oscillators , as is common in neuroscience , would make little sense . Granger causality [50] has emerged as a method of assessing putative causal relationships between brain regions based on LFP data . Granger causality assesses the relationship between two timeseries X and Y by comparing the predictive power of two different time-series models to predict future values of Y . The first model uses only past values of Y , whereas the second uses the history of X and Y . The additon of X allows one to assess the putative “causality” ( really , the predictive power ) of X . To see if we can understand information transmission pathways in the chip based on such techniques , we perform conditional Granger causality analysis on the above-indicated LFP regions for all three behavioral tasks , and plot the resulting inferences of causal interactions ( Fig 9 ) . We find that the decoders affect the status bits . We also find that the registers are affected by the decoder , and that the accumulator is affected by the registers . We also find communication between the two parts of the decoder for Donkey Kong , and a lack of communication from the accumulator to the registers in Pitfall . Some of these findings are true , registers really affect the accumulator and decoders really affect the status bits . Other insights are less true , e . g . decoding is independent and the accumulator obviously affects the registers . While some high level insights may be possible , the insight into the actual function of the processor is limited . The analysis that we did is very similar to the situation in neuroscience . In neuroscience as well , the signals come from a number of local sources . Moreover , there are also lots of connections but we hope that the methods will inform us about the relevant ones . It is hard to interpret the results—what exactly does the Granger causality model tell us about . Granger causality tells us how activity in the past are predictive of activity in the future , and the link from there to causal interactions is tentative at best [51] and yet such methods are extensively used across large subfields of neuroscience . Even if such methods would reliably tell us about large scale influences , it is very hard to get from a coarse resolution network to the microscopic computations . In line with recent advances in whole-animal recordings [2 , 6–8] , we measure the activity across all 3510 transistors simultaneously for all three behavioral states ( Fig 10 ) and plot normalized activity for each transistor versus time . Much as in neural systems , some transistors are relatively quiet and some are quite active , with a clear behaviorally-specific periodicity visible in overall activity . While whole-brain recording may facilitate identification of putative areas involved in particular behaviors [52] , ultimately the spike-level activity at this scale is difficult to interpret . Thus scientists turn to dimensionality reduction techniques [2 , 53 , 54] , which seek to explain high-dimensional data in terms of a low-dimensional representation of state . We use non-negative matrix factorization [55] to identify constituent signal parts across all time-varying transistor activity . We are thus , for the first time in the paper , taking advantage of all transistors simultaneously . Non-negative matrix factorization assumes each recovered timeseries of transistor activity is a linear combination of a small number of underlying nonnegative time-varying signals ( dimensions ) . Analogous with [2] we plot the recovered dimensions as a function of time ( Fig 11a ) and the transistor activity profile of each component ( Fig 11b ) . We can also examine a map of transistor-component activity both statically ( Fig 11c ) and dynamically ( S1–S3 Videos available in online supplementary materials ) . Clearly there is a lot of structure in this spatiotemporal dataset . To derive insight into recovered dimensions , we can try and relate parts of the low-dimensional time series to known signals or variables we know are important ( Fig 12a ) . Indeed , we find that some components relate to both the onset and offset ( rise and fall ) of the clock signal ( Fig 12b and 12c ) . This is quite interesting as we know that the processor uses a two-phase clock . We also find that a component relates strongly to the processors read-write signal ( Fig 12d ) . Thus , we find that variables of interest are indeed encoded by the population activity in the processor . In neuroscience , it is also frequently found that components from dimensionality reduction relate to variables of interest [56 , 57] . This is usually then seen as an indication that the brain cares about these variables . However , clearly , the link to the read-write signal and the clock does not lead to an overly important insight into the way the processor actually processes information . Similar questions arise in neuroscience where scientists ask if signals , such as synchrony , are a central part of information processing or if they are an irrelevant byproduct [58] . We should be careful at evaluating how much we understand and how much we are aided by more data . Pondering the results of the processor analysis we can obtain some insights into the developments needed to better utilize dimensionality reduction towards an understanding . The narrow range of games that we considered and the narrow range of their internal states ( we just simulated booting ) , means that many aspects of computation will not be reflected by the activities and hence not in the dimensionality reduction results . Moreover , the fact that we used linear reduction only allows for linear dependencies and transistors , just like neurons , have important nonlinear dependencies . Lastly , there is clearly a hierarchy in function in the processor and we would need to do it justice using hierarchical analysis approaches . The results of dimensionality reduction should be meaningful for guiding new experiments , necessitating transfer across chips in the same way as neuroscience experiments should transfer across animals . Importantly , the chip can work as a test case while we develop such methods .
Here we have taken a reconstructed and simulated processor and treated the data “recorded” from it in the same way we have been trained to analyze brain data . We have used it as a test case to check the naïve use of various approaches used in neuroscience . We have found that the standard data analysis techniques produce results that are surprisingly similar to the results found about real brains . However , in the case of the processor we know its function and structure and our results stayed well short of what we would call a satisfying understanding . Obviously the brain is not a processor , and a tremendous amount of effort and time have been spent characterizing these differences over the past century [22 , 23 , 59] . Neural systems are analog and and biophysically complex , they operate at temporal scales vastly slower than this classical processor but with far greater parallelism than is available in state of the art processors . Typical neurons also have several orders of magnitude more inputs than a transistor . Moreover , the design process for the brain ( evolution ) is dramatically different from that of the processor ( the MOS6502 was designed by a small team of people over a few years ) . As such , we should be skeptical about generalizing from processors to the brain . However , we cannot write off the failure of the methods we used on the processor simply because processors are different from neural systems . After all , the brain also consists of a large number of modules that can equally switch their input and output properties . It also has prominent oscillations , which may act as clock signals as well [60] . Similarly , a small number of relevant connections can produce drivers that are more important than those of the bulk of the activity . Also , the localization of function that is often assumed to simplify models of the brain is only a very rough approximation . This is true even in an area like V1 where a great diversity of co-localized cells can be found [61] . Altogether , there seems to be little reason to assume that any of the methods we used should be more meaningful on brains than on the processor . To analyze our simulations we needed to convert the binary transistor state of the processor into spike trains so that we could apply methods from neuroscience to ( see Methods ) . While this may be artefactual , we want to remind the reader that in neuroscience the idea of an action potential is also only an approximate description of the effects of a cell’s activity . For example , there are known effects based on the extrasynaptic diffusion of neurotransmitters [62] and it is believed that active conductances in dendrites may be crucial to computation [63] . Our behavioral mechanisms are entirely passive as both the transistor based simulator is too slow to play the game for any reasonable duration and the hardware for game input/output has yet to be reconstructed . Even if we could “play” the game , the dimensionality of the input space would consist at best of a few digital switches and a simple joystick . One is reminded of the reaching tasks which dominate a large fraction of movement research . Tasks that isolate one kind of computation would be needed so that interference studies would be really interpretable . If we had a way of hypothesizing the right structure , then it would be reasonably easy to test . Indeed , there are a number of large scale theories of the brain [5 , 64 , 65] . However , the set of potential models of the brain is unbelievably large . Our data about the brain from all the experiments so far , is very limited and based on the techniques that we reviewed above . As such , it would be quite impressive if any of these high level models would actually match the human brain to a reasonable degree . Still , they provide beautiful inspiration for a lot of ongoing neuroscience research and are starting to exhibit some human-like behaviors [64] . If the brain is actually simple , then a human can guess a model , and through hypothesis generation and falsification we may eventually obtain that model . If the brain is not actually simple , then this approach may not ever converge . Simpler models might yield more insight—specifically seeking out an “adder” circuit might be possible , if we had a strong understanding of binary encoding and could tease apart the system to specifically control inputs and outputs of a subregion—examine it in slice , if you will . The analytic tools we have adopted are in many ways “classic” , and are taught to graduate students in neuroinformatics courses . Recent progress in methods for dimensionality reduction , subspace identification , time-series analysis , and tools for building rich probabilistic models may provide some additional insight , assuming the challenges of scale can be overcome . Culturally , applying these methods to real data , and rewarding those who innovate methodologically , may become more important . We can look at the rise of bioinformatics as an independent field with its own funding streams . Neuroscience needs strong neuroinformatics to make sense of the emerging datasets and known artificial systems can serve as a sanity check and a way of understanding failure modes . We also want to suggest that it may be an important intermediate step for neuroscience to develop methods that allow understanding a processor . Because they can be simulated in any computer and arbitrarily perturbed , they are a great testbed to ask how useful the methods are that we are using in neuroscience on a daily basis . Scientific fields often work well in situations where we can measure how well a project is doing . In the case of processors we know their function and we can know if our algorithms discover it . Unless our methods can deal with a simple processor , how could we expect it to work on our own brain ? Machine learning and statistics currently lack good high-dimensional datasets with complex underlying dynamics and known ground truth . While not a perfect match , the dynamics of a processor may provide a compelling intermediate step . Additionally , most neural datasets are still “small data”—hundreds of cells over tens of minutes . The processor enables the generation of arbitrary complexity and arbitrarially-long timeseries , enabling a focus on scalable algorithms . We must be careful to not over-fit , but neuroscience is rife with examples of adopting analytic tools from vary different domains ( linear system theory , stochastic process theory , kalman filtering ) to understand neural systems . In the case of the processor , we really understand how it works . We have a name for each of the modules on the chip and we know which area is covered by each of them ( Fig 13a ) . Moreover , for each of these modules we know how its outputs depend on its inputs and many students of electrical engineering would know multiple ways of implementing the same function . In the case of the brain , we also have a way of dividing it into regions ( Fig 13b , adopted from [66] ) . However , we only use anatomy to divide into modules and even among specialists there is a lot of disagreement about the division . Most importantly though , we do not generally know how the output relates to the inputs . As we reviewed in this paper , we may even want to be careful about the conclusions about the modules that neuroscience has drawn so far , after all , much of our insights come from small datasets , with analysis methods that make questionable assumptions . There are other computing systems that scientists are trying to reverse engineer . One particularly relevant one are artificial neural networks . A plethora of methods are being developed to ask how they work . This includes ways of letting the networks paint images [67] and ways of plotting the optimal stimuli for various areas [68] . While progress has been made on understanding the mechanisms and architecture for networks performing image classification , more complex systems are still completely opaque [69] . Thus a true understanding even for these comparatively simple , human-engineered systems remains elusive , and sometimes they can even surprise us by having truly surprising properties [70] . The brain is clearly far more complicated and our difficulty at understanding deep learning may suggest that the brain is hard to understand if it uses anything like gradient descent on a cost function . What kind of developments would make understanding the processor , and ultimately the brain , more tractable ? While we can offer no definitive conclusion , we see multiple ways in which we could have better understood the processor . If we had experiments that would more cleanly separate one computation then results would be more meaningful . For example , lesion studies would be far more meaningful if we could also simultaneously control the exact code the processor was executing at a given moment . Better theories could most obviously have helped; if we had known that the microprocessor has adders we could have searched for them . Lastly , better data analysis methods , e . g . those that can explicitly search for hierarchical structure or utilize information across multiple processors . Development in these areas seems particularly promising . The microprocessor may help us by being a sieve for ideas: good ideas for understanding the brain should also help us understand the processor . Ultimately , the problem is not that neuroscientists could not understand a microprocessor , the problem is that they would not understand it given the approaches they are currently taking .
All acquisition and development of the initial simulation was performed in James [11] . 200°F sulfuric acid was used to decap multiple 6502D ICs . Nikon LV150n and Nikon Optiphot 220 light microscopes were used to capture 72 tiled visible-light images of the die , resulting in 342 Mpix of data . Computational methods and human manual annotation used developed to reconstruct the metal , polysilicon , via , and interconnect layers . 3510 active enhancement-mode transistors were captured this way . The authors inferred 1018 depletion-mode transistors ( serving as pullups ) from the circuit topology as they were unable to capture the depletion mask layer . An optimized C++ simulator was constructed to enable simulation at the rate of 1000 processor clock cycles per wallclock second . We evaluated the four provided ROMs ( Donkey Kong , Space Invaders , Pitfall , and Asteroids ) ultimately choosing the first three as they reliably drove the TIA and subsequently produced image frames . 10 seconds of behavior were simulated for each game , resulting in over 250 frames per game . Whole-circuit simulation enables high-throughput targeted manipulation of the underlying circuit . We systematically perturb each transistor in the processor by forcing its input high , thus leaving it in an “on” state . We measure the impact of a lesion by whether or not the system advances far enough to draw the first frame of the game . Failure to produce the first frame constitutes as a loss of function . We identified 1560 transistors which resulted in loss of function across all games , 200 transistors which resulted in loss of function across two games , and 186 transistors which resulted in loss of function for a single game . We plot those single-behavior lesion transistors by game in Fig 4 . Using the acquired netlist , we implement the authors method from [31] on the region of the processor consisting of the X , Y , and S registers . A nonparametric distance-dependent stochastic block model is jointly fit to six connectivitiy matrices: G → C1 , G → C2 , C1 → C2 C2 → C1 , C1 → G , C2 → G , and via Markov-chain Monte Carlo , seeks the maximum a posteriori estmate for the observed connectivity . We chose to focus on transistor switching as this is the closest in spirit to discrete action potentials of the sort readily available to neuroscientific analysis . The alternative , performing analysis with the signals on internal wires , would be analogous to measuring transmembrane voltage . Rasters were plotted from 10 example transistors which showed sufficient variance in spiking rate . We compute luminance from the RGB output value of the simulator for each output pixel to the TIA . We then look at the transistor rasters and sum activity for 100 previous timesteps and call this the “mean rate” . For each transistor we then compute a tuning curve of mean rate versus luminance , normalized by the frequency of occurrence of that luminance value . Note that each game outputs only a small number of discrete colors and thus discrete luminance values . We used SI as it gave the most equal sampling of luminance space . We then evaluate the degree of fit to a unimodial Gaussian for each resulting tuning curve and classify tuning curves by eye into simple and complex responses , of which Fig 4 contains representative examples . For the SI behavior we took spiking activity from the first 100ms of SI and performed spike word analysis on a random subset of 64 transistors close to the mean firing rate of all 3510 . To derive “local field potentials” we spatially integrate transistor switching over a region with a Gaussian weighting of σ = 500μm and low-pass filter the result using a window with a width of 4 timesteps . We compute periodograms using Welch’s method with 256-sample long windows with no overlap and a Hanning window . We adopt methods for assessing conditional Granger causality as outlined in [71] . We take the LFP generated using methods in section and create 100 1ms-long trials for each behavioral experiment . We then compute the conditional Granger causality for model orders ranging from 1 to 31 . We compute BIC for all behaviors and select a model order of 20 as this is where BIC plateaus . The transistor switching state for the first 106 timestamps for each behavioral state is acquired , and binned in 100-timestep increments . The activity of each transistor is converted into a z-score by subtracting mean and normalizing to unit variance . We perform dimensionality reduction on the first 100 , 000 timesteps of the 3510-element transistor state vectors for each behavioral condition . We use non-negative matrix factorization , which attempts to find two matrices , W and H , whose product WH approximates the observed data matrix X . This is equivalent to minimizing the objective | | W H - X | | 2 2 . The Scikit-Learn [72] implementation initialized via nonnegative double singular value decomposition solved via coordinate descent , as is the default . We use a latent dimensionality of 6 as it was found by hand to provide the most interpretable results . When plotting , the intensity of each transistor in a latent dimension is indicated by the saturation and size of point . To interpret the latent structure we first compute the signed correlation between the latent dimension and each of the 25 known signals . We show particularly interpretable results .
|
Neuroscience is held back by the fact that it is hard to evaluate if a conclusion is correct; the complexity of the systems under study and their experimental inaccessability make the assessment of algorithmic and data analytic technqiues challenging at best . We thus argue for testing approaches using known artifacts , where the correct interpretation is known . Here we present a microprocessor platform as one such test case . We find that many approaches in neuroscience , when used naïvely , fall short of producing a meaningful understanding .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"nervous",
"system",
"computer",
"hardware",
"light",
"neuroscience",
"electromagnetic",
"radiation",
"luminance",
"brain",
"mapping",
"computational",
"neuroscience",
"visible",
"light",
"neuronal",
"tuning",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"connectomics",
"physics",
"microprocessors",
"cellular",
"neuroscience",
"neuroanatomy",
"cell",
"biology",
"anatomy",
"computer",
"architecture",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"computational",
"biology",
"behavioral",
"neuroscience"
] |
2017
|
Could a Neuroscientist Understand a Microprocessor?
|
The Pantanal is a hotspot for arbovirus studies in South America . Various medically important flaviviruses and alphaviruses have been reported in domestic and wild animals in the region . To expand the knowledge of local arbovirus circulation , a serosurvey for 14 Brazilian orthobunyaviruses was conducted with equines , sheep and free-ranging caimans . Sera were tested for specific viral antibodies using plaque-reduction neutralization test ( PRNT ) . Monotypic reactions were detected for Maguari , Xingu , Apeu , Guaroa , Murutucu , Oriboca , Oropouche and Nepuyo viruses . Despite the low titers for most of the orthobunyaviruses tested , the detection of monotypic reactions for eight orthobunyaviruses suggests the Pantanal as a region of great orthobunyavirus diversity . The present data , in conjunction with previous studies that detected a high diversity of other arboviruses , ratify the Pantanal as an important natural reservoir for sylvatic and medically important arboviruses in Brazil .
The Bunyaviridae family contains over 300 enveloped viruses divided into five genera , including Orthobunyavirus , Phlebovirus and Nairovirus , which include arboviruses , viruses capable of alternately replicating in vertebrates and arthropods [1] . Orthobunyaviruses are classified into immunologically distinct groups based on their immunological relationships , including Bunyamwera , Bwamba , group C , Guama , Simbu , Mapputta , and California . Little is known of the pathogenicity of some of these groups for human beings or of the clinical manifestations of the associated disease [2] . In Brazil , dozens of orthobunyaviruses have been isolated including Oropouche virus ( OROV ) , the most medically important orthobunyavirus in the country involved in explosive human outbreaks mainly in rural areas of northern Brazil [3] . Recently , OROV has also been reported in urban areas of the west-central region of the country [4] . The west-central region of Brazil encompasses the states of Goiás , Mato Grosso ( MT ) , Mato Grosso do Sul ( MS ) , and the Federal District and is the second largest region in the country . Located within the west-central region of Brazil is the Pantanal , one of the world's largest floodplains with diverse and abundant wildlife located within the states of MS and MT . Despite the evidence of various medically important arboviruses including flaviviruses and alphaviruses in the Pantanal region , the circulation of orthobunyaviruses in the region is poorly known . The only investigation for orthobunyaviruses in the region was conducted in horses in the early 1990s [5] . Based on our previous experience exploring the prevalence of arboviruses in the Pantanal , we expected to encounter a great diversity of orthobunyaviruses there as well . Accordingly , we selected 13 Brazilian othobunyaviruses of potential medical importance for serodetection studies in a series of vertebrate specimens collected throughout the region from 2009–2011 . In particular , we selected OROV , Marituba ( MTBV ) , Guaroa ( GROV ) , Guama ( GMAV ) , Nepuyo ( NEPUV ) , Murutucu ( MURV ) , Oriboca ( ORIV ) , Xingu ( XINV ) , Caraparu ( CARV ) , Catu ( CATUV ) , Apeu ( APEUV ) , Itaqui ( ITQV ) and Tucunduba ( TUCV ) , all viruses known to cause human illness . Maguari virus ( MAGV ) , presumably a non-medically important arbovirus that had been previously reported from Pantanal [5] , was also included totaling 14 orthobunyaviruses investigated .
Equines , sheep and caimans were sampled in 17 cattle ranches visited in February and October 2009 , April , September and October 2010 and January 2011 in the Nhecolândia Sub-region of Pantanal , municipality of Corumbá , MS , west-central Brazil . Equines , including horses , donkeys and mules were sampled in 15 ranches , sheep in nine and caimans in two ranches . Four equine samples collected from a ranch in Nabileque , a different sub-region of Pantanal , were also tested . Serum samples of equines , sheep and caimans , used in the present study had been previously tested for arthropod-borne viruses resulting in serological evidence of four alphaviruses and five flaviviruses [6] , [7] . Considering that orthobunyaviruses are also transmitted by arthropods , a subset of sera from the same samplings was selected for the orthobunyavirus investigation . The collections for this study were authorized by oral consent by the residents of the sampled properties after previous contact with the owners of the sampled properties . This study was approved by the Animal Ethics Committee of Fundação Oswaldo Cruz of Ministry of Health of Brazil ( License CEUA-Fiocruz LW-1/12 , protocol P-74/10-5 ) in compliance with the requirements of Brazilian Law 11794/2008 , which rules on the scientific use of animals , including the principles of the Brazilian society of Science in laboratory animals . The caiman sampling was also approved by the Instituto Chico Mendes de Conservação da Biodiversidade of Ministry of Environment of Brazil ( licenses ICMBio 18363-1/2009 and 18363-2/2010 ) . Blood samples from equines ( n = 375 ) , and from sheep ( n = 232 ) were taken by jugular venipuncture . Information including age , health condition and travel history outside of Pantanal were recorded for each animal sampled . All the equine samples tested were seropositive for flaviviruses by blocking ELISA in a previous study indicating exposure of this subset of equines to mosquito bites [6] . Briefly , blocking ELISA evaluated the ability of the sera to block the binding of the flavivirus group-reactive monoclonal antibody 6B6C-1 to the cell lysate-derived antigen for West Nile virus ( WNV ) [8] . All equines and sheep were apparently healthy at time of venipuncture . Caimans ( n = 66 ) were captured from sites where a high concentration of these animals was observed , such as lentic systems that were formed by ephemeral rivers . Caimans were captured from boats or from the riverbanks , brought to shore and sampled by venipuncture of the internal jugular vein . All serum samples were heat-inactivated and tested by the 90% plaque-reduction neutralization test ( PRNT90 ) for their ability to neutralize plaque formation by OROV , MTBV , GROV , GMAV , NEPUV , MURV , ORIV , XINV , CARV , CATUV , APEUV , ITQV , TUCV and MAGV following standard protocols [9] . Viruses were provided by the Arbovirus Diseases Branch of the Division of Vector-Borne Diseases , Centers for Disease Control and Prevention ( CDC ) , from its arthropod-borne virus reference collection . Low-passage preparations of the following virus strains were used in this study: OROV ( TRVL9760 ) , MTBV ( BeAn15 ) , GROV ( CoH352111 ) , GMAV ( BeAn277 ) , NEPUV ( BeAn10709 ) , MURV ( BeAn974 ) , ORIV ( BeAn17 ) , XINV ( BeH388464 ) , CARV ( SPAn26550 ) , CATUV ( BeH151 ) , APEUV ( BeAn848 ) , ITQV ( BeAn12797 ) , TUCV ( BeAn278 ) and MAGV ( BeAn7272 ) . Mouse hyperimmune ascitic fluids ( MHIAF ) were used as positive control , and diluent media and serum samples of wild animals that had PRNT90 titers <10 for the tested viruses , as negative controls . Briefly , in a biosafety level three facility ( BSL3 ) , serum samples were initially screened at a dilution of 1:10 and those that neutralized virus challenge by at least 90% were further tested at serial two-fold dilutions that ranged from 1:20–1:320 to determine 90% endpoint titers . Serum samples were considered seropositive in a monotypic reaction when a serum dilution of at least 1:10 reduced at least 90% of the formation of viral plaques of only one of the 14 orthobunyaviruses tested in Vero cells . Sera that presented PRNT titer ≥ 10 for more than one orthobunyavirus were considered heterotypic reactions . Serum samples with PRNT titers < 10 for all 14 orthobunyaviruses tested were considered seronegative . To save resources and considering monotypic reactions to be the most reliable with no indication of cross-reaction , for XINV that was one of the last orthobunyaviruses tested , most samples that were XINV-positive in the screening and had PRNT90 titers ≥10 for any other orthobunyavirus were considered heterotypic reactions and not further tested to determine XINV-endpoint titers . The same approach was used for samples that were XINV-positive in the screening and that presented PRNT90 titer <10 for all the other orthobunyaviruses characterizing monotypic reactions for XINV .
For 375 equines tested , 285 ( 76% ) were heterotypic reactions with PRNT titers ≥ 10 for more than one orthobunyavirus , 69 ( 18 . 4% ) showed monotypic reactions for MAGV , 19 ( 5 . 1% ) were seronegative , and two ( 0 . 5% ) showed monotypic reactions for XINV ( Table 1 ) . Seventeen ( 4 . 5% ) equines presented heterotypic reactions for three or more orthobunyaviruses ( Table 2 ) . For 232 sheep tested , 115 ( 49 . 6% ) were seronegative , 95 ( 41% ) were heterotypic reactions , 14 ( 6% ) showed monotypic reactions for XINV , three ( 1 . 3% ) for APEUV , one ( 0 . 4% ) for GROV , one ( 0 . 4% ) for MAGV , one ( 0 . 4% ) for MURV , one ( 0 . 4% ) for ORIV and one ( 0 . 4% ) for OROV ( Table 1 ) . Twelve ( 5 . 2% ) sheep presented heterotypic reactions for three or more orthobunyaviruses ( Table 2 ) . For 66 free-ranging caiman samples , 63 ( 95% ) were seronegative , two ( 3% ) showed monotypic reactions for NEPUV and one ( 1 . 5% ) was heterotypic reaction ( Table 1 ) .
The Pantanal , which presents vast wetland habitat in a subtropical climate , present a set of factors that supports the introduction , maintenance , and evolution of arthropod-borne viruses . The region has abundant biodiversity and is the most important waterbird breeding area in South America [10] . The Nhecolândia Sub-region of the Pantanal is the world’s largest and most biodiverse region of subtropical lakes [11] , [12] . Recent arbovirus studies conducted in the region have found serological evidence of at least nine medically important arboviruses , including WNV and Mayaro virus . Moreover , Ilheus virus and six novel viruses were recently isolated from local mosquitoes [6] , [7] , [13] , [14] , [15] , [16] , [17] . In the absence of direct viral detection , diagnosis of arbovirus infections is performed by indirect serological tests . However , some cross-reactivity in primary infections has been reported among certain bunyaviruses . For instance , using the complement-fixation test , ORIV cross-reacted broadly with MURV antibody [18] . In neutralization tests using NEPUV and guinea pig immune sera for different group C viruses , NEPUV reacted mainly with immune serum of MURV , but also with immune sera of MARV , CARV and ITQV [19] . Additionally , antibody responses in vertebrates sequentially infected with orthobunyaviruses are not well described [20] . There is only one report that describes the antibody responses in vertebrates experimentally inoculated with two different orthobunyaviruses [21] . Moreover , other bunyaviruses may also circulate in the region , including novel orthobunyaviruses , which could theoretically generate cross-reacting neutralizing antibodies and lead to misinterpretation . A novel orthobunyavirus closely related to CARV has been recently reported in febrile patients in Peru [22] . In August of 2010 , after an epizootic of illness among sylvatic monkeys , an OROV species reassortant named Madre de Dios virus was isolated from a sick monkey collected in a forest near a small rural village in Venezuela [23] . Therefore , we used a conservative threshold for detection of neutralizing antibodies ( 90% ) in the region’s equines , sheep and wild caimans and we considered monotypic serologic responses to be the most reliable , as these samples reacted with just one of the fourteen viruses employed in the tests , with no indication of cross-reaction . This criterion may appear conservative , but the main objective is to prevent the introduction of false positives in the data , even at cost of some false negatives [24] . In the present study , monotypic reactions were encountered for local circulation of XINV , MAGV , APEUV , ORIV , OROV , MURV , NEPUV and GROV based on the detection of neutralization antibodies in equines , sheep and free-ranging caimans ( S1 Data ) . Except for one equine that exhibited monotypic reaction to MAGV and had travel history to Sidrolândia , outside of the Pantanal , all of the seropositive animals that showed monotypic reactions , lacked travel history indicating autochthonous transmission of these orthobunyaviruses in the region . OROV is a member of the Simbu serogoup and it has been involved in explosive human outbreaks in northern Brazil since the 1960s . Oropouche fever is characterized by an abrupt onset and fever , headache , myalgia , arthralgia , dizziness , chills and photophobia . Some cases can be severe , including neurologic disorder characterized as meningitis mainly in immunocompromised patients [25] , [26] . From 2000 to 2007 , OROV was detected in febrile patients from Bolivia , Ecuador and Peru [27] . Recently , OROV was reported during a dengue outbreak in MT , the west-central region of the country [4] . OROV is thought to be maintained in nature in a sylvatic cycle in which primates , sloths and birds are the amplifying hosts and the biting midge Culicoides paraensis is the main vector . OROV has also been isolated from mosquitoes , including Aedes serratus and Culex quinquefasciatus [26] . In the present study , only one sheep had monotypic reaction for OROV . APEUV , CARV , ITQV , MTBV , MURV , ORIV and NEPUV are arboviruses of the group C encountered in the Amazon Basin . The group C arboviruses are maintained in nature mainly between rodents and Culex spp . mosquitoes . With the exception of NEPUV , all have been isolated from human beings in the Amazon Basin [26] . In the early 1970s , neutralizing antibodies for NEPUV were detected in bats from Trinidad , and the virus has been involved in human disease in Central America [28] , [29] . CARV is the most widely distributed group C virus in the Amazon Basin and it causes human disease in southeast Brazil [30] . Group C arboviruses produce a febrile syndrome with sudden onset , including high fever , headache , chills , myalgia , photophobia and retrobulbar pain [31] . From 1996 to 2001 , CARV , ITQV , and MURV were isolated from numerous pools of Culex spp . from Peru [32] . In 2001–2002 , APEUV was detected in monkeys from the Brazilian Amazon [33] . CARV and MURV were also recently detected in febrile patients from Bolivia and Peru [27] . Among the group C orthobunyaviruses tested in the present study , sheep had monotypic reactions for APEUV , MURV and ORIV , and two caimans had monotypic reactions for NEPUV . All equines had heterotypic reactions or were negative for group C orthobunyaviruses . GROV , a member of the California encephalitis serogroup , is considered one of the most widely distributed orthobunyaviruses in the Amazon region . Several strains of GROV have been isolated from febrile patients and Anopheles spp . mosquitoes in Brazil , and birds are suspected to be the vertebrate amplifying hosts . GROV has been involved in sporadic cases of disease with acute onset and high fever , chills , headache , myalgia and malaise in rural areas of the Amazon region [26] . GROV was recently isolated in febrile patients from Bolivia and Peru [27] , [34] . In the present study , only one sheep had monotypic reaction for GROV . GMAV and CATUV are members of the Guama serogroup and have been isolated from human blood samples in the Amazon Basin . The ecoepidemiology of these viruses is similar to the group C arboviruses [35] . Both of them have been isolated mainly from rodents and Culex spp . mosquitoes [26] . In the early 1970s , neutralizing antibodies for GMAV were detected in bats from Trinidad [28] . When symptomatic , infections by GMAV and CATUV have a sudden onset of mild fever , dizziness , headache , muscle pains , arthralgia , photophobia and malaise [31] , [35] . All animals tested in the present study were seronegative or had heterotypic reactions for GMAV or CATUV . TUCV and XINV are genetically characterized as members of the Wyeomyia and Bunyamwera groups , respectively [36] , [37] . TUCV has been isolated from different mosquitoes , including Wyeomyia sp . , Sabethes sp . and Trichoprosopon digitatum and was also isolated from a patient with meningoencephalitis in Brazil [38] . The vertebrate hosts of TUCV remain unknown . No monotypic reactions for TUCV were detected in the present serological inquiry . Regarding XINV , the vectors and wild vertebrate amplifiers are unknown and the virus was only isolated from a hepatitis B case with a fatal outcome in Brazil [31] . XINV was the most prevalent orthobunyavirus in Pantanal sheep . The detection of 14 ( 6% ) sheep with monotypic reactions for XINV in seven ( 78% ) ranches sampled suggests widespread circulation of XINV in sheep of Pantanal ( Table 1 ) . Equines with monotypic reactions for XINV were detected in two Pantanal ranches . MAGV ( Bunyamwera serogroup ) was first isolated in Brazil in the 1950s from a mixed mosquito pool containing Aedes spp . , Mansonia sp . and Psorophora ferox and since then has never been reported to cause disease . MAGV was previously classified as a subtype of Cache Valley virus , but some strains of MAGV have been shown to differ antigenically from the prototype . MAGV is now regarded as a closely related , but distinct virus [39] . The enzootic transmission cycle of MAGV is unknown , but serological evidence has been reported in birds , sheep , water buffalo , man , cattle and mainly horses , from which MAGV has been isolated in Colombia and Guyana [39] . An arbovirus investigation conducted also in the Pantanal in the 1990s detected serological evidence for MAGV in 28% of the equines tested [5] . In the present study , one sheep and 69 ( 18 . 4% ) equines from 14 ( 93% ) ranches had monotypic reactions for MAGV . Together , these studies suggest that the circulation of MAGV in equines from Pantanal has been active for at least three decades and that MAGV is now widely distributed in the region ( Table 1 ) . Evidence reported here also suggest current or recent circulation of MAGV in the Pantanal . Among the equines with monotypic reactions for MAGV , three animals were two-year-old at the moment of the venopuncture in 2009 suggesting that MAGV circulated in equines of the region between 2007 and September 2009 . Except for MAGV , which was previously detected in horses in the 1990s [5] , the detection of monotypic reactions for seven other orthobunyaviruses provides the first evidence of their circulation in the Pantanal . Considering the potential for cross-reaction complicates the interpretation of serological tests , more investigation is needed to confirm their circulation by virus isolation , to determine the public health burden and understand the ecology of transmission of these viruses in the Pantanal region , including identifying amplifier hosts and vectors . The biased selection of equines for our study ( positive by blocking ELISA for flaviviruses ) may have contributed to our high prevalence for MAGV result in equines . If so , an explanation is that the vector or vertebrate hosts involved in transmitting or amplifying flaviviruses may play a similar role for MAGV . The vertebrate amplifying hosts of these viruses in the Pantanal have not been studied and testing only equines , sheep and caimans may provide a biased view of the relative amounts of orthobunyavirus transmission because these hosts may not attract all vectors equally . In the present study , there was a significant difference in the number of orthobunyaviruses detected according to the host tested . For instance , sheep had monotypic reactions for seven orthobunyaviruses , while equines only for two . Sheep may attract more orthobunyavirus vectors than equines in the region and may be useful surrogates for detection of orthobunyavirus activity in the Pantanal . Interestingly , the same is not true for flaviviruses . A previous study reported that equines are more exposed to flaviviruses than sheep in the region [6] . A serosurvey of free-ranging rodents , as well as , non-human primates and/or local human residents would be interesting as an additional investigational tool for some of these orthobunyaviruses . In fact , a recent survey conducted among free-living non-human primates in MS outside the boundaries of the Pantanal found one animal with OROV-reactive hemagglutination-inhibiting antibodies [40] . Caimans were selected for inclusion in our study because reptiles may play a larger role in the transmission cycle of arboviruses than previously assumed [41] . However , we found that only two caimans were seropositive for NEPUV , suggesting that the participation of caimans in other orthobunyavirus transmission cycles in the Pantanal is unlikely . However , interestingly monotypic reactions for NEPUV were detected only in caimans . The prevalence of MAGV in equines and XINV in sheep suggests that both viruses are widespread in the Pantanal . Another explanation would be cross-reaction between XINV and MAGV , which are considered indistinguishable by some classical tests [26] . Both viruses are members of the Bunyamwera serogroup , but the molecular characterization of XINV confirms them as different orthobunyaviruses [36] . In fact , the number of MAGV-seropositive equines in the present study was higher than the number of XINV-seropositive equines . Thirty animals , including equines and sheep , presented heterotypic reactions not only for both MAGV and XINV , but also for a third orthobunyavirus , including TUCV ( n = 14 ) , APEUV ( n = 10 ) , NEPUV ( n = 4 ) , GMAV ( n = 1 ) , MURV ( n = 1 ) and ORIV ( n = 1 ) . One sheep presented neutralizing antibodies to four orthobunyaviruses ( Table 2 ) . Together these findings might indicate either cross-reactivity among the orthobunyaviruses tested or multiple exposure of Pantanal animals to various orthobunyaviruses throughout their lifetimes . Conservative serologic criteria were used to present evidence of local circulation of orthobunyaviruses primarily in equines , but also in sheep and caimans in the Pantanal , Brazil . The detection of seropositive animals for seven medically important orthobunyaviruses are novel findings for the Pantanal . However , because detection of antibodies is indirect evidence of virus circulation and because unknown orthobunyaviruses may circulate in the region , we encourage efforts to isolate viruses to confirm the circulation of these orthobunyaviruses in the Pantanal . Despite the low titers for most of the orthobunyaviruses tested , monotypic reactions for eight orthobunyaviruses suggest the Pantanal as a region of great orthobunyavirus diversity . The present data , in conjunction with previous studies that detected local circulation of various flaviviruses and alphaviruses , confirm the Pantanal as an important natural reservoir for zoonotic arboviruses in Brazil .
|
In the present study , we report the evidence of various orthobunyaviruses of medical importance in domestic and wild animals of the Pantanal , a large floodplain located in West-Central Brazil . Although various other arboviruses as flaviviruses and alphaviruses have been reported in the region , orthobunyaviruses are not commonly investigated in Pantanal . Positive results for eight orthobunyaviruses , including Maguari , Xingu , Apeu , Guaroa , Murutucu , Oriboca , Oropouche and Nepuyo viruses were detected in equines , sheep and caimans in the region . The findings reported here suggest the Pantanal as a region of great orthobunyavirus diversity and ratify the Pantanal as hotspot for arbovirus studies in South America .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"ruminants",
"pathogens",
"immunology",
"geographical",
"locations",
"microbiology",
"vertebrates",
"animals",
"mammals",
"viruses",
"rna",
"viruses",
"antibodies",
"insect",
"vectors",
"bunyaviruses",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"sheep",
"south",
"america",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"brazil",
"disease",
"vectors",
"insects",
"arthropoda",
"people",
"and",
"places",
"biochemistry",
"mosquitoes",
"eukaryota",
"arboviruses",
"flaviviruses",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"equines",
"amniotes",
"organisms"
] |
2017
|
Neutralizing antibodies for orthobunyaviruses in Pantanal, Brazil
|
The success of combination antiretroviral therapy is limited by the evolutionary escape dynamics of HIV-1 . We used Isotonic Conjunctive Bayesian Networks ( I-CBNs ) , a class of probabilistic graphical models , to describe this process . We employed partial order constraints among viral resistance mutations , which give rise to a limited set of mutational pathways , and we modeled phenotypic drug resistance as monotonically increasing along any escape pathway . Using this model , the individualized genetic barrier ( IGB ) to each drug is derived as the probability of the virus not acquiring additional mutations that confer resistance . Drug-specific IGBs were combined to obtain the IGB to an entire regimen , which quantifies the virus' genetic potential for developing drug resistance under combination therapy . The IGB was tested as a predictor of therapeutic outcome using between 2 , 185 and 2 , 631 treatment change episodes of subtype B infected patients from the Swiss HIV Cohort Study Database , a large observational cohort . Using logistic regression , significant univariate predictors included most of the 18 drugs and single-drug IGBs , the IGB to the entire regimen , the expert rules-based genotypic susceptibility score ( GSS ) , several individual mutations , and the peak viral load before treatment change . In the multivariate analysis , the only genotype-derived variables that remained significantly associated with virological success were GSS and , with 10-fold stronger association , IGB to regimen . When predicting suppression of viral load below 400 cps/ml , IGB outperformed GSS and also improved GSS-containing predictors significantly , but the difference was not significant for suppression below 50 cps/ml . Thus , the IGB to regimen is a novel data-derived predictor of treatment outcome that has potential to improve the interpretation of genotypic drug resistance tests .
Despite an increasing arsenal and improved potency of antiretroviral drugs , the optimal use of combination antiretroviral therapy against HIV-1 infection remains challenging [1] . Complicating factors include drug interactions and toxicities , adherence to therapy , and development of drug resistance [2] . Because genotypic drug resistance testing is performed on a routine basis today and because mutational patterns are unique for each patient , treatment choices are , in principle , highly personalized . In practice , however , it can be difficult to identify an optimal drug combination for each individual patient due to the combinatorial complexity of both the set of feasible drug combinations and of viral mutational patterns . In addition to controlled clinical trials , analyzing data from large observational cohort studies is a promising way to identify predictors of treatment outcome , even if the availability of drugs and therapeutic strategies change over time [3] . This approach can be based on modeling the risk of acquiring additional mutations [4] , on estimating future drug options [5] , on predicting the time to virological failure [6] , [7] , or on classifying the regimens of treatment change episodes ( TCEs ) as successful versus failing , depending on the patient's response to therapy . A TCE consists of predictor variables including the applied drug combination , viral genotype , treatment history , demographic and clinical parameters , and a response variable such as the change in viral load . HIV-1 genotype has been shown to be a strong predictor of therapeutic success in retrospective and prospective studies [8]–[14] , but the large number of mutations complicates prediction . TCE classification is a noisy , high-dimensional prediction problem with unobserved confounding factors and sparse data . It has been addressed by several statistical learning methods [15]–[25] . Comparative studies have emphasized the importance of selection and representation of features , especially of the viral genotype , over the choice of the learning algorithm [26]–[28] . In order to directly correlate genotype with clinical response , rules-based approaches , such as the genotypic susceptibility score ( GSS ) [29]–[34] and statistical models [23] , [26] , [28] have been proposed , often outcompeting human experts [35] . Drug resistance development is driven by viral evolution and thus models of viral evolutionary escape from drug pressure have been proposed to improve therapy response prediction [16] , [22] , [36] . Specifically , the individualized genetic barrier ( IGB ) to drug resistance has been suggested as a predictor of treatment outcome . The IGB is defined as the probability of the virus not to become resistant to a certain drug [37]–[39] . A high IGB means that viral evolutionary escape from the selective pressure of the drug is unlikely . Related quantities are the average number of mutations and the average time to reach drug resistance derived from simulated HIV-1 evolutionary trajectories on an estimated fitness landscape [36] , [40] , [41] . This approach has been explored for treatment with zidovudine plus lamivudine and with nelfinavir [42] , but it does not scale to the variety of combination therapies observed in clinical databases , because sufficient data for estimating fitness landscapes is available only for a few drug combinations . Earlier , the term ‘calculated genetic barrier’ has been used to assess the number of mutations necessary to acquire specific drug resistance-associated mutations , which were found to be similar among HIV-1 subtypes [43] . In the present study , we apply a simplified definition of the IGB which can be computed efficiently for any drug combination based on a statistical model that captures the order and the dynamics of accumulating mutations and the associated levels of phenotypic drug resistance [44] . The IGB to resistance to a certain drug is the probability that the virus will not accumulate additional mutations leading to a resistant strain . This drug-specific IGB has been demonstrated to be a strong predictor of virological response in two large observational cohort studies [26] , [28] . Here , we derive a novel predictor , the IGB to the entire drug combination which measures the genetic potential for evolutionary escape of the virus from the selective pressure of combination therapy . In order to assess the performance of the IGB as a predictor of treatment outcome , we analyzed TCE data from the Swiss HIV Cohort Study ( SHCS ) database , a large , long-term observational , multi-center , clinical database with integrated results of genotypic drug resistance tests [45] , [46] . We identified risk factors of therapeutic failure and constructed models of treatment outcome considering as predictors the applied regimen , treatment history , viral genotype , GSS , drug-specific IGBs , IGB to regimen , and demographic and clinical variables including patient adherence . Overall , we found the IGB to the entire regimen to be the strongest and most significant predictor . Our results demonstrate that the viral genotype is represented efficiently by the IGB to regimen , a single , interpretable probability summarizing the predicted dynamics of viral evolutionary escape .
For each drug , viral evolutionary escape from its selective pressure was modeled using Isotonic Conjuctive Bayesian Networks ( I-CBNs ) . In these probabilistic graphical models , dependencies among mutations are described by a partial order , which defines the genotype lattice , i . e . , the set of genotypes compatible with the order constraints , and hence the set of possible mutational escape pathways ( Figure 1 ) . To each genotype , its level of phenotypic drug resistance is associated using isotonic regression , such that drug resistance is monotonically non-decreasing along any mutational pathway from the wild type towards the genotype carrying all mutations . Using cross-sectional matched genotype-phenotype pairs from the Stanford HIV Drug Resistance Database , I-CBN models were learned for a total of 18 antiretroviral drugs ( Supporting Figures S4 , S5 , S6 , S7 , S8 , S9 , S10 , S11 , S12 , S13 , S14 , S15 , S16 , S17 , S18 , S19 , S20 , S21 , Supporting Table S2 ) . Each model includes up to eleven pre-selected mutations ( see Methods ) . From the I-CBN models , transition probabilities among genotypes were derived and the individualized genetic barrier ( IGB ) to resistance development to each drug was computed as the probability of the observed genotype not acquiring additional mutations that would transform it into a genotypic state predicted to be resistant . For a drug combination , the IGB was obtained as the sum over all drugs of the regimen of the drug-specific IGBs . Thus , the IGB to regimen can be regarded as the expected number of active components in the drug cocktail taking viral evolutionary escape mechanisms into account . To assess the predictive power of the IGB in a clinical setting , we analyzed a large cohort of HIV-1-infected patients and compared the IGB to several known predictors of therapy response ( Figure 2 ) , including the GSS , obtained from the Stanford HIV Drug Resistance Database website ( HIVdb 6 . 2 . 0 ) . TCEs from the time period 1988–2010 were derived from the SHCS database ( Table 1 and 2 ) and labeled as either failure or success ( see Methods ) . Therapy success was defined as viral load reduction below 50 cps/ml ( 400 cps/ml ) during treatment . We obtained 2185 ( 2631 ) genotype-therapy pairs , including 73% ( 63% ) failures . The usage of individual drugs and the 30 most frequent drug combinations are shown in Supporting Figures S2S and S3 , respectively . The historical development of drug usage patterns is reported in Supporting Table S3 , where the regimens are annotated as either being recommended as first-line or alternative regimens according to current treatment guidelines [47] , or as past first-line or second-line recommended regimens that are still in use in developing countries or occasionally used if drug resistant virus is present at baseline or as salvage regimens , or as regimens that are not in use anymore as first-line regimens but were before , including those still used under special circumstances , such as unusual tolerability . In order to predict the outcome ( failure versus success ) of each therapy , we considered applied drugs , demographic and clinical variables , viral genotype , IGBs to received drugs , and IGB to regimen ( Figure 2 , Table S1 ) . Univariate logistic regression resulted in a total of 50 ( 44 ) features that were significantly associated with therapy outcome ( Figure S22 ) . Among the predictive drugs , the use of ZDV , d4T , 3TC , and NFV were associated with increased risk of therapeutic failure , while ABC , TDF , FTC , EFV , RTV , LPV/r , ATV , and ATV/r increased the odds of therapeutic success . Most of the significant amino acid changes in the viral protease ( PR ) gene ( 10I , 30N , 33F , 46I , 54V , 71V , 82A , 84V , 90M ) and reverse transcriptase ( RT ) gene ( 39A , 41L , 44D , 67N , 74V , 103N , 118I , 123S , 210W , 215Y , 297R ) have been associated with resistance to multiple PR inhibitors ( PIs ) and RT inhibitors ( RTIs ) , respectively , and all except PR 30N and RT 123S increased the risk of treatment failure . A higher IGB to any of 15 ( 16 ) individual drugs increased the chance of successful virological response . The IGB to the entire drug combination and the GSS were also significant predictors . In the multivariate analysis , only 12 ( 14 ) variables were significant , nine ( ten ) of which are indicating the inclusion of individual drugs in the regimen ( Figure 3 ) . The usage of the nucleoside RTIs ( NRTIs ) ZDV , ddI , d4T , and 3TC , and of the PIs APV and SQV , were associated with negative treatment outcome , whereas the four boosted PIs ( i . e . , given together with low-dose RTV to improve their bioavailability ) SQV/r , IDV/r , LPV/r , and ATV/r had positive predictive power . Among the many genotype-derived predictors , only GSS and IGB to regimen reached statistical significance at the 1% level in the multivariate model . For the 50 cps/ml success definition , the odds ratio ( OR ) of therapeutic success was ten-fold higher for the IGB ( OR 23 . 6 , 95% confidence interval [CI] 12 . 21–45 . 4 , ) as compared to the GSS ( OR 2 . 1 , 95% CI 1 . 6–2 . 7 , ) , and similarly for 400 cps/ml ( IGB OR 25 . 0 , 95% CI 14 . 7–42 . 5 , versus GSS OR 1 . 8 , 95% CI 1 . 5–2 . 2 , ) , indicating that the IGB provides an effective summary of the risk of treatment failure due to viral genetic changes . In addition , increased overall maximum ( peak ) viral load before treatment remained a significant predictor of therapy outcome in the multivariate logistic regression model . For optimal treatment outcome prediction , we also explored the use of regularized logistic regression models . Specifically , the elastic net , which combines and regularization was applied to identify sparse classifiers of therapy outcome . Classifier performance was evaluated in ROC curves summarized by the area under the ROC curve ( AUC ) , and analyzed according to the historical drug usage patterns ( Table S3 ) . The competitive models ( high AUC ) are only those using all clinical and demographic variables , mutations , and drugs ( Tables S5 , S6 , Figure 4 ) . When comparing IGB to GSS as predictors in this setting , we found a significant advantage of the IGB for 400 cps/ml if all other features are included in the models ( , Wilcoxon rank sum test ) . Furthermore , the IGB also improves treatment outcome prediction if added to models that already contain the GSS ( ) . For 50 cps/ml , we did not find significant differences in AUC between IGB and GSS when used in prediction models that included all other covariates , nor did the GSS-containing model improve upon adding IGB . The significant increase for the larger dataset with the 400 cps/ml success definition demonstrates the predictive power of the IGB and indicates that GSS and IGB , although correlated , contain some orthogonal information , which , if combined , can further improve treatment outcome prediction .
We have comprehensively analyzed factors of therapy outcome in the SHCS database using univariate , multivariate , and regularized multivariate logistic regression models . As predictors of therapeutic success we identified the applied drugs , the GSS , and as the strongest predictor the IGB to regimen , a novel predictor derived from viral genotype . Including genotype information into treatment outcome prediction is challenging because of the large number of observed mutations and the complexity of the genotype-phenotype relationship . Here , we have explored the IGB to drug resistance as a summary measure of the escape dynamics of the virus under treatment . The underlying idea of this modeling approach is that the IGB captures how difficult it is for the virus to escape from the selective pressure of individual drugs or from the entire drug combination . This piece of information is different from assessing the current genotypic or phenotypic drug resistance state of the virus , as intended , for example , by the GSS . The IGB makes a prediction about the expected escape dynamics of the virus population given its current genetic state . The computation of the IGB involves an evolutionary model of genetic progression under selective drug pressure along multiple mutational pathways and a notion of evolutionary escape , which was based here on the predicted level of phenotypic drug resistance . We applied I-CBN models for jointly describing genetic progression and associated phenotypic change of the virus . In particular , phenotype predictions are non-linear in the mutations , which allows for capturing epistatic effects , i . e . , the same mutation can have different effects on the resistance phenotype depending on the genetic background of the virus ( Figure 1 ) . The I-CBN models were estimated from independent genotype-phenotype data . Using these models , the complex , high-dimensional , genotypic data of each virus can be summarized efficiently by the IGB to resistance to each drug . Thus , rather than modeling interactions between drugs and individual mutations , the IGB provides a comprehensive model of drug-genotype interaction . In the present study , we have extended the concept of the IGB to the entire regimen in a fashion that allows for computing this quantity for any drug combination and hence for large clinical datasets . The IGB to regimen can be regarded as the expected number of active drugs in the regimen . Assuming independent effects among drugs , we compute the regimen IGB from the drug IGBs . These simplifying assumptions are made for computational feasibility . They present a conceptual limitation of the approach and more elaborate models are conceivable . In addition , other variables not included in this study might be important , for example , pharmacological properties of drug combinations and host genetic factors . Here , the IGB , a single interpretable quantity , was found to be the strongest genotype-derived predictor of virological response and hence the most efficient representation of the viral genotype with respect to therapy outcome . We have used throughout two definitions of virological success of treatment , namely reduction of viral load below 50 cps/ml and below 400 cps/ml . The latter less stringent cutoff was included because in the past it represented the limit of detection of viral load assays . Today viral load values of 50 cps/ml and lower can be measured and reduction below 50 cps/ml ( or below the limit of detection ) is an accepted therapeutic goal . We generally found very similar results for the two datasets , but the advantage of using IGB over GSS ( the de facto standard genotype interpretation tool ) reached statistical significance only for 400 cps/ml , but not for 50 cps/ml . This finding may , in part , be due to the larger dataset and hence increased statistical power for 400 cps/ml as compared to 50 cps/ml . In the future , larger datasets will be required to further evaluate the IGB and its potential to predict treatment outcome without the need for expert rules . This property of the IGB is particularly appealing for new drugs , for which reliable rules are not readily available before evidence has accumulated in published studies . Larger datasets and more elaborate statistical variable importance methods [48] will also increase the power to detect other factors of therapeutic outcome , but the general consistency between the 50 cps/ml and 400 cps/ml success definitions suggests that a sizable fraction of important variables have been identified . In addition , larger TCE databases will allow for analyzing alternative endpoints , such as time to virological failure or virological response after a fixed period of time . In the univariate analysis , most drugs had a positive effect on treatment outcome , with the exception of ZDV , d4T , 3TC , and NFV . The negative associations might be due to the prominent use of the drug combinations ( ZDV or d4T ) +3TC+ ( IDV or NFV ) , 90% of which were failures . The four drugs were among the first to be approved for antiretroviral therapy and used in early suboptimal regimens . Moreover , they were poorly tolerated and therefore one can expect a general lower adherence to treatment . A similar observation was made in the multivariate analysis , where ZDV , ddI , SQV , 3TC and d4T were significant predictors decreasing the odds of therapeutic success . This effect might also be due to the common early use of these drugs in mono therapy and their later use in salvage regimens , even if multiple resistance mutations had already accumulated [49] . Among PIs , a pronounced trend was that boosting with RTV increased the odds of successful treatment . The fraction of PI boosting in the dataset is reported in Supporting Table S4 . A few variables did not show significant association with therapy outcome although they might have been expected to . For example , adherence is a well-known predictor of treatment success [50] , [51] , but it failed to reach significance in the multivariate model , most likely due to lack of adherence data for about 45% of the patients . The missing data resulted from collecting adherence data within the SHCS only since January 2003 . Indeed , in a multivariate analysis restricted to the subset of 1183 TCEs with observed adherence a more pronounced effect can be observed . We have not included a set of variables in this study that are known to be predictors because of the construction of the dataset . The definition of the dataset of genotype-therapy pairs allows for including several sequential TCEs from the same patient . Most TCEs are actually derived from unique patients , but some patients occur multiple times . Each TCE gives rise to two therapy cases , a failure , which had given rise to the switch , followed by a salvage regimen , which can be a failure or a success . Therefore , we did not include variables that are affected by the sequential ordering of therapies , such as the total time a patient was under therapy with a certain drug or the calendar year of treatment . In summary , the IGB to regimen is a new predictor of treatment outcome that captures , in a single quantity , the virus' genetic potential for developing drug resistance under the selective pressure of the combination therapy . The IGB can be computed efficiently for any viral genotype and any drug combination . It may thus contribute to improved interpretation of genotypic drug resistance tests and to the rational design of individualized therapies . Future prospective studies are required to apply these results to other patient populations and to eventually integrate them into clinical practice .
Founded in 1988 , the SHCS is a nationwide , prospective , multicenter , clinic-based cohort with continuous enrolment and semi-annual study visits representing approximately 50% of all HIV-infected and 75% of all treated patients in Switzerland [46] . The SHCS has been approved by ethical committees of all participating institutions , and written informed consent has been obtained from all participants . The SHCS drug resistance database contains the results of 13 , 201 genotypic resistance tests from 9 , 231 patients , stored in a central database [45] . Resistance data stem from routine clinical testing ( 60% ) and from tests performed retrospectively from frozen repository plasma samples ( 40% ) ( Table 1 and 2 ) . The SHCS has been approved by the following ethical committees of all participating institutions: Kantonale Ethikkommission Bern; Ethikkommission beider Basel; comité d'éthique du département de médicine de Hôpitaux Universitaires de Genéve; commission d'éthique de la recherche clinique , Lausanne; comitato etico cantonale , Bellinzona; Ethikkommission des Kanton St . Gallens; and Ethik-Kommission Zürich , all Switzerland . Written informed consent has been obtained from all participants [46] . TCEs were obtained from the SHCS database as follows . Each TCE consists of a failing therapy followed by a salvage therapy ( Supporting Figure S1 ) . We required that the failing therapy was at least four month long and that the genotype was measured no more than 90 days before and no more than 30 days after onset of the uninterrupted salvage therapy [26] . In order to restrict to failing regimens due to viral rebound and to exclude convenience treatment changes or single determinations of low-level viremias ( blips ) , a failing therapy was defined by either two consecutive viral load measurements above 500 cps/ml , or a single viral rebound followed by therapy switch , or single rebound after 180 days and lack of viral suppression below the limit of detection . Therapies were labeled ‘success’ versus ‘failure’ as follows . Any failing therapy was considered a failure . Salvage therapies were considered successful , if viral load dropped below 50 cps/ml at any time point during treatment , otherwise they were considered failures . Because viral load assays with a sensitivity of 50 cps/ml were not available for the whole observation period , we also considered an alternative definition of therapy success as a viral load reduction below 400 cps/ml . The TCE dataset spans the time period 1988–2010 , but 75% of TCEs date from 2000 or later . Genetic progression of the virus under selective drug pressure and the resulting phenotypic drug resistance changes were modeled jointly using I-CBNs [44] . In this model , mutations occur subject to partial order constraints which define the genotype lattice , the set of genotypes compatible with the constraints , and drug resistance is non-decreasing along any mutational pathway ( Figure 1 ) . Formally ( see [44] for details ) , let be a partially ordered set of mutations . Each genotype is identified with the subset of mutations it carries . The genotype lattice induced by is the set of all genotypes for which it holds that implies whenever in . We denote by the set of accessible mutations from genotype under the given partial order constraints . The I-CBN is a statistical model for the random variables , describing observed genotypes , and , describing associated drug resistance phenotypes , both of which are observed from true hidden genotypes subject to noise . The probability of an unobserved genotype is defined as ( 1 ) where the parameters denote the conditional probabilities of mutation given that all of its predecessor mutations have occurred , . The observed random variables and are independent given . The genotype observation error is modeled as ( 2 ) where denotes the Hamming distance and errors are assumed to occur independently among sites at rate . The observed drug resistance phenotype is the log fold-change in susceptibility . For each genotype , it follows a normal distribution ( 3 ) subject to the monotonicity contraints for all genotypes . The complete model for and is then the marginalization ( 4 ) Parameter estimation for this model was performed using the EM algorithm described in [44] . The model was applied separately to 18 antiretroviral drugs , using between 280 and 2303 ( median 1448 ) cross-sectional genotype-phenotype pairs , i . e . , observations of , obtained from the Stanford HIV Drug Resistance Database , restricted to subtype B sequences and to Phenosense or Antivirogram assays [52] . For each drug , we selected its resistance-associated mutations reported on the Stanford HIVdb website lumping together mutations occurring at the same site , or if unavailable , applied -penalized ( lasso ) linear regression [53] , [54] to select from all PR or RT mutations occurring at least ten times a sparse set of predictor mutations . The performance of the models is reported as the Pearson correlation coefficient between true and predicted phenotypes , estimated from a separate , random subset of 20% of the data . Phenotypic cutoff values were derived from the distribution of fold-change values as described previously [15] , [26] and used to dichotomize resistance predictions ( Supporting Table S2 ) . Given an I-CBN model , transition probabilities among genotypes , can be computed as ( 5 ) Using these transition probabilities and the predicted drug resistance phenotypes , we define the IGB of genotype to resistance to drug as the probability of the virus not reaching any genotypic state predicted as resistant , ( 6 ) where is the subset of all genotypes predicted to be resistant to drug , i . e . , for which is greater than the resistance cutoff ( Supporting Table S2 ) . Genotypes outside the lattice ( not complying with the partial order constraints ) are regarded as erroneous observations of the genotypes in the lattice . The IGB of such a genotype is ( 7 ) where is the probability of the actual genotype being given that has been observed . By Bayes' theorem , ( 8 ) where is modeled as in Eq . 2 . The genetic barrier to escape from a regimen is defined as the sum of the drug-specific barriers over all drugs in the regimen ( 9 ) Because the IGB to each drug can be regarded as an estimate of the activity of the drug ( the probability of not escaping ) , the IGB to a regimen may be interpreted as the expected number of active drugs in the regimen . Note that , that means that evolutionary escape is almost certain , and that adding a drug to a regimen can only increase the genetic barrier to the regimen . For classifying therapies as failures versus successes , univariate , multivariate , and regularized multivariate logistic regression was used . For a set of precitors , the therapeutic success probability is modeled by the regression ( 10 ) where are the regression coefficients . The odds ratio of therapeutic success associated with a one-unit increase in predictor is . P-values for the predictors are corrected for multiple testing using the Benjamini-Hochberg procedure . For regularization , we applied the elastic net [55] , which combines an ( lasso ) penalty encouraging sparse solutions with an ( ridge ) penalty that tends to average across correlated features . Classifier performance was evaluated using ROC curves and is reported as the area under the ROC curve ( AUC ) . The data was ten times randomly split into 40% for estimation of the two hyperparameters ( one for the degree of each type of regularization ) and 60% for model fitting and testing , which was done by 10-fold cross-validation [56] . The R language for statistical computing ( http://www . r-project . org/ ) was used for all analyses , including the R packages icbn , glmnet , and ROCR . An R script for computing the IGB is available at: http://www . cbg . ethz . ch/software/igb . The Stanford HIVDB Sierra web service was used for GSS computation .
We thank the patients who participated in the SHCS; the physicians and study nurses for excellent patient care; the resistance laboratories for high-quality genotypic drug resistance testing; SmartGene , Zug , Switzerland , for technical support; Brigitte Remy , Martin Rickenbach , F . Schoeni-Affolter , and Yannick Vallet from the SHCS Data Center in Lausanne for data management; and Daniéle Perraudin and Mirjam Minichiello for administrative assistance . The members of the Swiss HIV Cohort Study are: Aubert V , Barth J , Battegay M , Bernasconi E , Böni J , Bucher HC , Burton-Jeangros C , Calmy A , Cavassini M , Egger M , Elzi L , Fehr J , Fellay J , Francioli P , Furrer H ( Chairman of the Clinical and Laboratory Committee ) , Fux CA , Gorgievski M , Günthard H ( President of the SHCS ) , Haerry D ( deputy of “Positive Council” ) , Hasse B , Hirsch HH , Hirschel B , Hösli I , Kahlert C , Kaiser L , Keiser O , Kind C , Klimkait T , Kovari H , Ledergerber B , Martinetti G , Martinez de Tejada B , Metzner K , Müller N , Nadal D , Pantaleo G , Rauch A ( Chairman of the Scientific Board ) , Regenass S , Rickenbach M ( Head of Data Center ) , Rudin C ( Chairman of the Mother & Child Substudy ) , Schmid P , Schultze D , Schöni-Affolter F , Schüpbach J , Speck R , Taffé P , Tarr P , Telenti A , Trkola A , Vernazza P , Weber R , Yerly S .
|
Drug resistance remains a challenge in the management of HIV-infected patients . The accumulation of mutations during ongoing viral replication is the origin of drug resistance development . Understanding this evolutionary process in a quantitative manner is an important prerequisite for minimizing the risk of resistance development and for the optimal selection of drug combinations for each individual patient . We present probabilistic graphical models for describing the evolution of drug resistance , and we derive the individualized genetic barrier ( IGB ) , a single quantity summarizing the genetic potential of the virus for evolutionary escape from selective drug pressure . The predictive power of the IGB is demonstrated on a large well characterized clinical cohort of HIV patients and compared to classical predictors .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods",
"Acknowledgments"
] |
[] |
2013
|
The Individualized Genetic Barrier Predicts Treatment Response in a Large Cohort of HIV-1 Infected Patients
|
All eukaryotic genomes are packaged as chromatin , with DNA interlaced with both regularly patterned nucleosomes and sub-nucleosomal-sized protein structures such as mobile and labile transcription factors ( TF ) and initiation complexes , together forming a dynamic chromatin landscape . Whilst details of nucleosome position in Arabidopsis have been previously analysed , there is less understanding of their relationship to more dynamic sub-nucleosomal particles ( subNSPs ) defined as protected regions shorter than the ~150bp typical of nucleosomes . The genome-wide profile of these subNSPs has not been previously analysed in plants and this study investigates the relationship of dynamic bound particles with transcriptional control . Here we combine differential micrococcal nuclease ( MNase ) digestion and a modified paired-end sequencing protocol to reveal the chromatin structure landscape of Arabidopsis cells across a wide particle size range . Linking this data to RNAseq expression analysis provides detailed insight into the relationship of identified DNA-bound particles with transcriptional activity . The use of differential digestion reveals sensitive positions , including a labile -1 nucleosome positioned upstream of the transcription start site ( TSS ) of active genes . We investigated the response of the chromatin landscape to changes in environmental conditions using light and dark growth , given the large transcriptional changes resulting from this simple alteration . The resulting shifts in the suites of expressed and repressed genes show little correspondence to changes in nucleosome positioning , but led to significant alterations in the profile of subNSPs upstream of TSS both globally and locally . We examined previously mapped positions for the TFs PIF3 , PIF4 and CCA1 , which regulate light responses , and found that changes in subNSPs co-localized with these binding sites . This small particle structure is detected only under low levels of MNase digestion and is lost on more complete digestion of chromatin to nucleosomes . We conclude that wide-spectrum analysis of the Arabidopsis genome by differential MNase digestion allows detection of sensitive features hereto obscured , and the comparisons between genome-wide subNSP profiles reveals dynamic changes in their distribution , particularly at distinct genomic locations ( i . e . 5’UTRs ) . The method here employed allows insight into the complex influence of genetic and extrinsic factors in modifying the sub-nucleosomal landscape in association with transcriptional changes .
Within the nucleus , higher eukaryotic genomes are packaged in the form of chromatin in which the structural unit is a nucleosome , a ~150 bp of double-stranded DNA wound around an octameric histone protein core . Nucleosomes are arranged in a repetitive manner spaced by a linker region of DNA with a length of ~30 bp , varying by species and genomic location [1 , 2] . The positioning of nucleosomes relative to DNA sequence is a dynamic process , regulating access to the DNA sequence of various factors and is essential for proper functional biological processes such as transcription , replication , DNA repair and recombination [3] . Moreover , both post-translational modifications of nucleosomal histones and positioning of nucleosomes on the DNA sequence contribute to epigenetic genomic information [4] . Technologies combining digestion of unbound double-stranded DNA using enzymes such as DNAse I or micrococcal nuclease ( MNase ) , and next-generation sequencing ( NGS ) of the remaining protected material allow the generation of genome-wide maps of nucleosome positions . These studies have previously been carried out in a wide range of eukaryotic organisms such as Saccharomyces cerevisiae [5] , Caenorhabditis elegans [6 , 7] , Drosophila melanogaster [8] , Zea mays [9] , and also in human cell culture systems [7 , 10] . Chodavarapu et al . ( 2010 ) [11] presented the first nucleosome-positioning map of mononucleosomes in Arabidopsis thaliana . They showed that ( 1 ) exons are nucleosome-enriched , ( 2 ) the intron-exon boundaries are demarcated by strongly positioned flanking nucleosomes , and ( 3 ) nucleosomal DNA is methylation-enriched . However their findings focus more on genome region rather than the underlying DNA sequence . Similarly it was shown that genome-wide , nucleosome patterning is uniform in protein-coding genes but not in pseudognes , transposable element genes and transposable elements [12 , 13] . Nucleosome patterning is uniform in euchromatin , whereas pericentromeric heterochromatin shows nucleosome enrichment [11] . Additionally , the periodic distribution of nucleosomes is independent of the level of expression of the gene , but the unoccupied distance between nucleosome positions is higher in transcribed genes [12] . Typically in these studies , digestion of chromatin was performed to completion with optional gel purification steps utilised to enrich for mono-nucleosomal DNA fragments . This gives a limited and static picture of the chromatin landscape focused on the positioning of stable nucleosomes . However , the dynamic structure of chromatin is required for changes in gene expression [14] . Partial digestion of chromatin in yeast and flies reveals variability in nucleosome occupancy profiles explained by the presence of MNase-hypersensitive nucleosomal DNA regions [15–17] . The existence of both hypersensitive and hyper-resistant nucleosomes was shown in the crop plant Zea mays [18] . These regions are found in the non-coding DNA around active genes and are colocalised with regulatory elements found in converse noncoding sequences , sequences such as KNOTTED 1 transcription factor binding site [19–21] . Moreover , MNase sensitive regions also correlate with recombination hotspots and hypomethylation [21] . These studies highlight the complex nature of labile DNA-bound proteins . In addition to the visualisation of nucleosome dynamics , partial digestion approaches allow size-resolution of nuclease protected chromatin particles using modified paired-end mode sequencing protocols to reveal the positions of sub-nucleosome sized particles ( subNSPs ) . These subNSPs protect DNA fragments of a size smaller than 120bp and may represent non-nucleosomal chromatin-associated proteins including DNA replication machinery such as DNA polymerase Origin Recognition complex ( ORC ) [20] or the transcriptional machinery such as sequence-specific TFs and complexes containing TFs [22] . Further examples include chromatin remodeler complexes [23] , and RNA polymerase or proteins involved in chromatin structure [24 , 25] in DNA recombination [26] or DNA repair [27] . Here we combine partial differential MNase digestion and size-resolved chromatin-seq with transcript profiling from the same samples in Arabidopsis thaliana for the first time . Low levels of MNase treatment reveals complexity regarding small particle recruitment and the lability of bound factors which has only previously been suggested in yeast [16] and human sperm [28] through comparisons of mutant types . Here we examine for the first time the dynamics of the overall chromatin landscape in a single cell type to changes in environmental conditions correlated to RNA-seq data , giving insight into the dynamic relationship between chromatin and the changing transcriptome .
Eight replicate Arabidopsis thaliana Col-0 cell cultures were sampled after 16-hour passage ( see Methods ) ( 4x Light grown , 4x Dark grown ) and subjected to a “higher” or “lower” level of chromatin digestion with MNase followed by Illumina Paired-End sequencing modified to take a wide DNA fragment size range . This allowed us to perform chromatin particle spectrum analysis ( CPSA ) as previously described in yeast [29] and enabled determination of regions of genomic DNA protected by nucleosomal , multi-nucleosomal , and sub-nucleosomal sized particles ( NSP , multiNSP , subNSP ) by retaining size information when mapping to genomic DNA . Spatial mapping of all protected DNA sequences demonstrated a homogeneous coverage of the A . thaliana genome with nucleosome sized bound particles , albeit with some regions of lesser chromatin coverage ( Fig 1A ) . Regions such as the centromeres , the heterochromatic knob on chromosome 4 [30] , and the rDNA encoding regions demonstrate abnormal structures likely due to their highly repetitive state inhibiting accurate mapping [31] . Mapping of the midpoints of all 150bp ( +/-15bp ) fragments across the genome ( Fig 1B: black track ) provides an NSP pattern remarkably consistent with previous results [11 , 12 , 32] . However , the approach of mapping according to size provided novel insight into the range of DNA-bound material , as utilising low intensity MNase digestion revealed not only NSPs , but also multiNSP regions resistant to MNase . Coverage of the underlying genome by nucleosomal ( ~150bp +/- 30bp ) or multi-nucleosomal ( N * ( 150bp + linker ) ) chromatin demonstrated a regularly occurring structure throughout the genome , with higher association to genic regions ( described further below ) . Interpolation between sequenced paired reads and cumulative measurement of all mapped fragments equal to or larger than nucleosomal and multi-nucleosomal sized particles generated a single nucleotide resolution map of the total coverage of the underlying DNA at genome scale ( Fig 1B: blue track ) . Furthermore , mapping by particle size in addition to genomic location demonstrated the continuous range of bound particle sizes apparent at this low digestion , as is observed directly in the 1 . 2kb exemplar region in Fig 1C . This 3D plot shows the non-discrete range of fragment sizes and abundances , revealing the absence of subnucleosomal fragments mapping to its centre , periodic mono-nucleosomes and a range of multi-nucleosomal protected regions . The differential MNase digestion method was previously implemented in yeast and revealed DNA fragments protected by subNSPs such as TFs [16 , 29] . To profile the subNSPs across the genome , we computationally subsampled all sequenced material where paired-end reads were less than 120bp i . e . sub-nucleosomal sized regions protected from MNase cutting activity . This revealed that subNSP binding factors have distinct organisation to key positions throughout the genome , specifically positions immediately upstream of , or coinciding with 5’ untranslated region ( UTR ) sites , as observed in the example of Fig 1B ( red track ) . This is consistent with the profiles expected by the binding of TFs and promoter complexes directly upstream of TSSs as observed in yeast [16 , 29] . Comparison of the otherwise equivalent but differentially digested A . thaliana samples between ‘high’ and ‘low’ level of MNase treatment demonstrated the effect of digestion extent on the DNA-bound material . At a genomic level , overt variation was not observed in NSP or multiNSP counts between differentially digested samples , consistent with view of chromatin as an essential stabilizer of the genome [5 , 33] ( S1 Fig ) . However , the marked variation in number of subNSP identified between digestion levels indicates the sensitive nature of these labile factors and suggests that their binding is more readily displaced from the underlying genome than nucleosomes ( S1 Fig ) . Under high digest , subNSPs prominent in low digest samples appear to have a reduced detection indicating greater susceptibility of the underlying DNA to nuclease-mediated degradation due to weaker protection from the bound protein in comparison to nucleosomes . Subsequently , high digest resulted in subNSPs frequently indistinguishable from the likely noise of surrounding low-level small-particle structure , particularly at TSSs ( Fig 2A—compare red tracks ) . Differential digestion levels in Zea mays has previously enabled description of hyper-sensitive sites in the genome where fragile nucleosomes are found [19] . We here observe a similar fragile nucleosome throughout the genome at the ‘-1’ position directly upstream of the TSS ( Fig 2B & 2C ) . Our comparative analysis was consistent with this outcome , which however was not previously reported in A . thaliana , likely due to high levels of MNase digestion [12] . For this reason , we believe that the low-digestion level and wider particle size analysis implemented here is essential in producing a holistic view of the nucleosomal landscape , and a facet often omitted from other genome-accessibility studies . The positioning of nucleosomes was observed to be most conserved at the TSS in individual gene examples ( Fig 1B ) . This periodicity is less defined downstream , with the less stringent chromatin organisation being consistent with previous results [12] . Plotting the average structure surrounding the TSS of all A . thaliana genes ( with consistent expressed isoforms between samples , N = 21 , 314 ) , we observed periodical ~150bp mono-nucleosomal organisation patterning as previously described in Arabidopsis ( Fig 2B ) [11] and typical of other eukaryotic organisms [6 , 7 , 17 , 29] . However , nucleosome positioning in the average profile surrounding the TSS also incorporates the variation between gene lengths and intron-exon structure when surmising multiple gene regions , which contributes to high genic average coverage but lower definition of nucleosomal periodicity . We note the comparative low height of the +2 nucleosome peak which is a consistent feature of our analyses , but currently unexplained . Consideration of MNase digestion level on mono-nucleosomal organisation shows higher digestion directly correlates with increased mono-nucleosomal structure in the genic region downstream of the TSS . Fig 2B indicates the areas where increased digestion correlates with strengthened or weakened monoNSP positions surmised through the genome , where the colorimetric background shows the strength of correlation between digestion level and NSP abundance per 10bp region . We observe a particularly sensitive nucleosome ( negatively correlated to digestion ) directly upstream of the TSS which has significantly lower occupancy in the high digested samples ( Fig 2C ) . This is in concordance with the hypersensitive minus-1 nucleosome identified in maize [18] . Additionally , the partially resistant multiNSP and sensitive subNSP positions are less represented under high digest conditions . For nucleosomes this is presumably due to cleavage of di- or tri-nucleosomes to monomers , however the comparative susceptibility of subNSPs results from greater access of MNase to digest under higher levels due to weaker binding or lower levels of DNA protection . The intron-exon boundary demonstrated a strong defining role for nucleosomal positioning as previously described [10] , but is not as strong a factor as that by which nucleosomal structure is organised at the TSS ( S2 Fig ) . Observing the nucleosomal organisation at the TSS regions of individual genes supports variable access of subNSPs due to widening of the open chromatin region between nucleosomes . This pattern is replicated at genomic scale , where open chromatin regions appear at the TSS and Coding Start Sites ( CSS ) ( Fig 2D & 2E ) , suggesting access for TF recruitment . The differential occupancy signals found at these open regions supports that differential digestion is required to demonstrate the exclusion of the labile and sensitive factors in these regions . The irradiance change in environmental conditions of the diurnal cycle is central to the cyclic expression changes in the plant transcriptome and the normal phenotype [34] . The changing accessibility of key genes has an impact on this expression . Nucleosome spacing at specific genes was correlated with their expression level in targeted analysis [12] . We here addressed the light-responsive gene networks to further understand the role that chromatin has on the activation and inactivation of genes . In addition to NSP binding , we analysed subNSP binding to get a more thorough global view of chromatin dynamics . Chromatin is understood to be more rigorously structured in exonic regions of genes than intergenic and intronic [11 , 12 , 32] . Utilising RNAseq data derived from the same samples used for chromatin-seq , the genome was divided into quartiles based upon gene expression level and the NSP positioning was plotted four bio-replicates of light grown cell cultures subjected to low and heavy digest ( see Material and methods ) . Genome-wide , genes under median to high expression showed a strongly focused positioning of NSPs surrounding the TSS ( Fig 3A , 50–75% , 75–100% ) . This structure was less pronounced in low-expressed genes ( Fig 3A , 25–50% ) and a solitary strongly positioned nucleosome at the point of TSS initiation ( up to 200bp downstream ) was typical in the lowest quartile of expressed genes ( Fig 3A , 0–25% ) . This loss of mononucleosomal-sized patterning supports the assertion that typical control of gene expression requires well-regulated chromatin [12] . Both the immobile NSP at the TSS of inactive genes , and a strongly defined TSS+1 NSP have been previously reported [12] , which was showed to inhibit the binding of transcription complexes and other factors , and to inhibit RNA Pol II activity [35] . Consideration was given to lateral movement of the TSS+1 nucleosome in response to changes in expression , allowing for increased upstream access , however we did not observe convincing evidence for this ( S3 Fig ) . We explored the position of subNSPs for the same range of light grown samples . SubNSP binding was enriched at TSS regions throughout the genome . Genes were separated into quartiles by degree of kurtosis of the subNSP peak at the TSS , showing that stronger recruitment correlates with higher levels of expression across the genome ( Fig 3B—upper: example peak structure separation , lower: 4 samples with RNAseq separated by subNSP recruitment ) . The increased expression in the higher quartiles of subNSP recruitment could either be explained by more frequent binding within the cell population at that position , or stronger binding to DNA by the subNSP , or more defined MNase-hypersensitive linkers at the TSS . Having establishing the nucleosomal and sub-nucleosomal structural changes within single samples , we contrasted the chromatin profiles between isolates grown in different irradiance conditions . We did not identify a statistically significant change in nucleosomal patterning genome-wide between samples grown in light or in dark ( following two weeks of continuous culture without light ) . We therefore conclude that the NSP structure is not significantly changed in response to this extrinsic environmental change , even when applied for an extended period . Consequently , we investigated the binding of subNSPs to discern any effect of the environmental changes to the DNA-bound landscape . The correlation between changing gene expression ( RNAseq , log2-fold change ) and the recruitment of subNSP ( TSS subNSP abundance , log2-fold change ) was assessed for each shared-isoform gene ( Fig 3C ) . We identified that genes under significantly increased expression in either condition ( p>0 . 05 , FDR ) correlated with stronger recruitment of subNSP binding at the TSS consistent with DNA-binding complexes such as TF being present at the time of sample harvest . Hence , whilst no change in gross NSP pattern was seen , changes of expression caused by applied environmental stimuli were correlated with a genome-wide changes at the level of subNSP binding . We therefore sought to understand whether the average landscape observed across the genome was representative of the local profile at well-regulated genes . Within genes of relevance to the light response , we assessed key genes involved in photosynthesis or photosensing that showed strong expression in the RNA-Seq data . For instance , Photosystem II Subunit S ( CP22 , AT1G44575 , fold change: -10 . 8 , FDR = 2 . 59E-136 ) showed that subNSP recruitment was observed around the TSS and the UTR region directly 5’ to the CSS , but was not identified in samples grown in dark ( Fig 4A , dashed boxes ) . CP22 encodes a photoprotective pigment-binding protein associated with the photosystem II in the grana thylakoids [36] . To prevent the formation of reactive species and the damage to membranes , a quenching of excited chlorophyll occurs by transferring the energy to carotenoid pigments [37] . Another example is the disappearance of a 5’UTR associated subNSP in dark compared to light for the Rubisco Small Subunit 3B ( AT5G38410 , fold change: -9 . 46 , FDR = 3 . 01E-076 , Fig 4B ) , encoding a subunit of the key Calvin cycle enzyme ribulose 1 , 5-bisphophate carboxylase-oxygenase ( RuBisCO ) [38] . Again , the presence of subNSP at the 5’UTR appears to coincide reliably with the increased expression . We then examined whether these changes in subNSPs could be related to known positions of TF binding to specific genomic locations . Previously identified TF Binding Sites ( TFBS ) were sourced from Athamap [39 , 40] ( S4 Fig ) or O’Malley et . al ( 2016 ) [41] to assess subNSP binding to their corresponding genomic targets . PHYTOCHROME INTERACTING FACTOR 3 ( PIF3 ) is a helix-loop-helix TF interacting with photoreceptors phyA and phyB as part of the plant light response [42] . Analysis of the PIF3 target binding motif across the genome demonstrated significantly more occupancy in light conditions ( Fig 5A , Number of sites = 1 , 930 ) . This was despite a non-significant change ( log2FC: -0 . 25 , FDR = 0 . 7 ) in expression of the PIF3 gene , consistent with post-translational regulation of the TF through phosphorylation [43] . The related PIF4 TF was however associated with a significant downregulation in dark-grown cells ( AT2G43010 , log2FC: -1 . 4 , FDR = 0 . 079 ) and revealed a similar subNSP binding structure at identified sites ( Fig 5B , N = 20 , 252 ) [44] . In contrast to the light responsive PIF TFs , the Myb-type CIRCADIAN CLOCK ASSOCIATED ( CCA1 , AT2G46830 ) TF similarly demonstrated increased binding at known TFBS [45] in light-grown samples , but presents an alternative saddle-shaped binding pattern . This heightened accumulated around the binding site indicates the motif region having susceptibility to MNase digestion and the subNSP being offset in dark conditions ( Fig 5C , N = 59 , 249 ) . TF encoding genes with higher expression in dark grown samples ( ERF1 , WRKY59 , AT4G18450 ) did not demonstrate an increased abundance of subNSP binding and patterning was retained , although some variability can be observed due to variation in digestion level ( S5 Fig ) . While the majority of subNSP binding positions are associated with genic promotor regions , approximately 7 . 91% of positions identified were greater than 10kb from an annotated coding sequence and 1 . 18% greater than 100kb distant . These sites can clearly be observed as protected DNA fragments of unknown function , which supports the role of the subNSP/TF detection approach as a new methodology to reveal not only the potential binding of transcription factors but also indicating other regions of interest for further analysis . However , while these positions may function as binding locations for long distance promotor elements , there are also potential roles unrelated to transcription factor binding which must be considered including the spatial organisation of DNA , and underlying physical structure of DNA which make regions resistant to MNase digestion i . e . secondary structures . Furthermore , investigation into associations with underlying genomic factors and histone modification statuses will be of significant future interest due to the connection of modified histones to transcriptional control [46] . One aspect to be assessed under matched conditions is the connecting the exclusion of H2A . Z and methylation of DNA in actively transcribed DNA [47] , and positions found to be enriched for subNSP binding ie . the 5’UTR and Transcription Start Sites . As ChIP-seq approaches to mapping TFs in A . thaliana advance , we should come to learn the identities of the factors we have observed protecting DNA and their role in the genomic landscape .
Dispersed cell suspension cultures , ( prepared from Arabidopsis thaliana leaves Columbia ecotype , Col-0 [48 , 49] and were a kind gif of Dr . Linda Hanley-Bowdoin obtained from North Carolina State University , USA ) were used . The cultures are a homogeneous population of physiologically and morphologically identical cells [50 , 51] . The cultures were maintained in 250 mL Erlenmeyer flasks , filled with 50 mL of cell culture medium ( Gamborg’s B5 basal medium with minor organic ( Sigma G5893 ) in 1 . 1 mg/L 2 , 4-D , 3 mM MES and 3% sucrose ) as previously described [49 , 52] . Cells were grown on a rotary shaker 160 rpm at 23°C ( LS-X ( Lab Shaker ) , Kuhner Shaker X ) . Constant light was used whereas dark grown cells were incubated under the same shaking and temperature condition but the flask were covered of aluminium foil . Cell line was subcultured every 7 days with a 2:50 ( inoculum: fresh medium ) dilution ratio . Cultures were carried out in duplicates . Light grown cells were sampled 5 days or 16 hours after subculture . Dark grown cells were adapted to dark conditions for 2 weeks with a 7-day subculture period . Dark grown cells were sampled 16-hour after the 2nd subculture . Cells were sampled by harvesting 30mL of cell culture and washed with autoclaved ultrapure water . Cells were frozen in liquid nitrogen and ground into a fine powder . The cell powder was either stored at -80°C or used directly for MNase digestion and RNA isolation . Pelleted cell culture was ground in a liquid nitrogen-cooled pestle and mortar to generate a cell powder . 1mL of the powder was suspended in a 0 . 5mL modified spheroplast digestion buffer & Nonidet P40 ( SDBN: 1M sorbitol , 10mM NaCl , 50 mM TrisŸHCl pH7 . 5 , 5mM MgCl2 , 1 mM CaCl2 , 1 mM 2-mercaptoethanol , 0 . 5 mM spermidine , 0 . 075% Nonidet P40 ) . 300 μl of cells were transferred into 1 . 5mL microcentrifuge tube containing 30 or 120 units of MNase ( Affymetrix ) . Cells were digested with MNase at 37°C for 3 min . The MNase digestion reactions were stopped by addition of and thorough mixing 30 μl of STOP solution containing 5% SDS and 250 μl EDTA . DNA was extracted by addition of an equal volume of 2X CTAB supplemented with PVP ( 200 mM Tris-HCl , 40 mM EDTA , pH 8 . 0; 2 . 8 M NaCl , 4% w/v CTAB , 2% PVP-40 CAS number 9003-39-8 ) was added to the digestion reaction and incubated at 45°C for 15 mins . DNA was separated from protein by two phenol: chloroform ( 1:1; 600μl total ) steps . DNA was precipitated with sodium acetate and propan-2-ol , washed in 80% ethanol and dried . Samples were incubated with 10X RNase A at 37°C for 1h with 100U unmodified T4 polynucleotide kinase ( NEB ) for further 30 min at 37°C . MNase digested DNA was separated on 1 . 5% agarose gel , stained with ethidium bromide for 10 min at 80V , DNA fragments ranging from 20 bp to 1 kb were excised and gel pieces containing DNA were placed in a Costar Spin-X 0 . 45 μm cellulose acetate centrifuge tube filter ( Sigma CLS8136 ) . Two series of freeze-thaws ( -80C for 10 mins/RT for 10 mins ) were performed to macerate the gel . Filter tubes were centrifuged twice ( 14kg; 5 min; RT ) with a 180 degree rotation of the tubes between the two spins . DNA in the aqueous phase was then extracted with 400μl phenol: chloroform ( 1:1 ) , and precipitated at -80°C for 30 mins with sodium acetate to propan-2-ol and finally washed with 80% ethanol before being resuspended in ultrapure milliQ water . Total RNA was extracted using the RNeasy Plant mini Kit ( Qiagen , http://www . qiagen . com ) . DNA samples were quantified on QubitTM 2 . 0 Fluorometer with dsDNA BR assay and RNA with Qubit RNA BR assay according to manufacturer’s instructions . RNA quality was checked on 1 . 5% agarose gel . Exeter Sequencing Service and Computational core facilities at the University of Exeter performed the Genomic and RNA library preparation , quantification and 50bp paired end sequencing on HiSeq 2500 . Additional information can be found on the Exeter Sequencing Service website ( http://sequencing . exeter . ac . uk/ ) . A total of ~861M paired end reads were obtained from the MNase-seq genomic library and ~173M for the RNAseq library ( S6 Fig ) . Genomic mapping was performed with bowtie allowing for 3 nucleotide errors [53] , resulting in 85 . 09% alignment to the TAIR10 reference Columbia-0 genome [54] . Aligned BAM files were converted to wig trace files for monoNSP midpoints ( 150bp +/- 10% ) using bespoke perl scripts . DNA coverage was produced using bamToBed and coverageBed from the bedtools package [55] . Datasets were either complete ( total ) or where specified split into <120bp , and >120bp subsamples , and converted to wig traces ( S7 Fig ) . Gene feature alignments were produced using danpos [56] and the ARAPORT11 annotation [57] . Where alternate splice variants were present between samples as determined from RNAseq analysis ( see below ) only genes primarily sharing the same model were utilised to ensure accurate boundary comparisons i . e . Transcription Start Sites . Continuous size variation plots ( 2D and 3D ) were produced using a bespoke package ( BAM2SizePlot . py ) producing visualisations utilising paired-end insert distance to determine particle size and can be obtained at https://github . com/ChromatinCardiff/ALD . RNAseq mapping was performed with Tophat2 [58] , resulting in 78 . 40% alignment to the TAIR10 reference Colombia-0 genome [54] and the ARAPORT11 annotation [57] . Determination of isoform presence was achieved with cufflinks [59] and only primary isoforms shared between all samples were utilised as gene feature boundaries to ensure error was not induced between splice variants ( n = 21 , 314 ) . Quantitative assessment was performed with HTseq [60] and analysed with edgeR [61 , 62] .
|
DNA is packaged by proteins into chromatin , of which the fundamental unit is a complex of histone proteins that wraps ~150bp of DNA into a nucleosome . Digestion of chromatin with enzymes such as micrococcal nuclease cuts the DNA between the protein particles , and by sequencing the cut sites , they can be mapped across the entire genome . Whilst nucleosomes are the most stable feature , less intensive digestion reveals a wider range of protein particles bound to DNA , in particular small sub-nucleosomal particles located most frequently upstream of genes . Here we show that these techniques can be used in the plant Arabidopsis to map the chromatin landscape of the range of particles sizes across the genome . We show for the first time in a multicellular higher eukaryote that this chromatin landscape is dynamic in response to environmental changes , in this case light or dark growth . Whereas the nucleosome positioning does not change significantly , we show profound changes in the smaller more labile factors under these different conditions . These changes in many cases correspond to the known binding sites of transcription factors that regulate genes in response to light , leading us to propose that full-spectrum chromatin landscape analysis can reveal directly the changes in transcription factor complex binding across the genome .
|
[
"Abstract",
"Introduction",
"Results",
"&",
"discussion",
"Materials",
"&",
"methods"
] |
[
"biotechnology",
"brassica",
"nucleosome",
"mapping",
"plant",
"science",
"model",
"organisms",
"experimental",
"organism",
"systems",
"plant",
"genomics",
"epigenetics",
"molecular",
"biology",
"techniques",
"plants",
"chromatin",
"structural",
"genomics",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"genome",
"complexity",
"gene",
"mapping",
"chromosome",
"biology",
"gene",
"expression",
"plant",
"genetics",
"molecular",
"biology",
"nucleosomes",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"plant",
"biotechnology",
"computational",
"biology",
"organisms"
] |
2017
|
Genome-wide chromatin mapping with size resolution reveals a dynamic sub-nucleosomal landscape in Arabidopsis
|
During neural circuit formation , most axons are guided to complex environments , coming into contact with multiple potential synaptic partners . However , it is critical that they recognize specific neurons with which to form synapses . Here , we utilize the split GFP-based marker Neuroligin-1 GFP Reconstitution Across Synaptic Partners ( NLG-1 GRASP ) to visualize specific synapses in live animals , and a circuit-specific behavioral assay to probe circuit function . We demonstrate that the receptor protein tyrosine phosphatase ( RPTP ) clr-1 is necessary for synaptic partner recognition ( SPR ) between the PHB sensory neurons and the AVA interneurons in C . elegans . Mutations in clr-1/RPTP result in reduced NLG-1 GRASP fluorescence and impaired behavioral output of the PHB circuit . Temperature-shift experiments demonstrate that clr-1/RPTP acts early in development , consistent with a role in SPR . Expression and cell-specific rescue experiments indicate that clr-1/RPTP functions in postsynaptic AVA neurons , and overexpression of clr-1/RPTP in AVA neurons is sufficient to direct additional PHB-AVA synaptogenesis . Genetic analysis reveals that clr-1/RPTP acts in the same pathway as the unc-6/Netrin ligand and the unc-40/DCC receptor , which act in AVA and PHB neurons , respectively . This study defines a new mechanism by which SPR is governed , and demonstrates that these three conserved families of molecules , with roles in neurological disorders and cancer , can act together to regulate communication between cells .
Perception , thought , and behavior all rely on the faithful transfer of information between neurons . During development , neurons form circuits through a series of well-characterized steps , including neuronal migration , guidance of axons and dendrites into target regions , and finally the formation of synapses between presumptive neuronal partners . However , within a target region , most neurites contact many potential partners . To form functional circuits , neurons must faithfully recognize and form synapses only with the correct neuronal partners [1 , 2] . Relatively little is understood about this process of synaptic partner recognition ( SPR ) , and many of the molecular mechanisms involved remain unknown . To discover molecular pathways that mediate SPR , we focus on the phasmid sensory circuit in Caenorhabditis elegans hermaphrodites , which mediates avoidance of toxin-producing Streptomyces bacteria [3] and other repulsive cues [4 , 5] . Specifically , we study synapses between the PHB sensory neurons and AVA interneurons in C . elegans ( Fig 1A ) . The left and right PHBs extend axons through a neurite bundle containing approximately 30 potential partners , yet selectively form the majority of their synapses with AVA and PVC interneurons , which control backward and forward movement [6–9] . Here , we study PHB-AVA synapses utilizing the split GFP-based trans-synaptic marker NLG-1 GRASP to visualize specific sensory synapses ( Fig 1B ) [10 , 11] , and a circuit-specific behavioral assay to probe circuit function ( Fig 1C and 1D ) [4 , 10] . The stably expressed NLG-1 GRASP marker specifically labels PHB-AVA synapses in live animals without affecting the behavioral output of the circuit [10] . We previously determined that UNC-6/Netrin acts as a retrograde juxtacrine signal from presumptive postsynaptic AVA interneurons to presumptive presynaptic PHB neurons . PHB neurons receive this signal via the UNC-40/DCC receptor , specifying PHBs as AVAs’ synaptic partners [10] . However , it was not known if other molecules were required for SPR , and the mechanism by which AVA neurons receive the SPR signal remained unknown . Here , we show that the clr-1 gene plays a crucial role in SPR . clr-1 encodes a receptor protein tyrosine phosphatase ( RPTP ) with extracellular domains similar to those in the Leukocyte common Antigen-Related protein ( LAR ) family of RPTPs [12] . We demonstrate that clr-1/RPTP is necessary for formation of PHB-AVA synapses , and overexpression promotes increased synapse formation between the two neurons . CLR-1/RPTP acts in postsynaptic AVA neurons to direct SPR , and is enriched in AVA neurites in the region of synapse formation with PHB neurons . Finally , we find that clr-1/RPTP acts in the same pathway as unc-6/Netrin and unc-40/DCC to mediate SPR . Our findings demonstrate a new role for clr-1/RPTP in SPR , and indicate that these three conserved families of proteins can act together to mediate communication between cells , which may provide insight into their roles in neurological disorders [13 , 14] and cancer biology [15–17] .
To discover molecules that might act with UNC-6/Netrin and UNC-40/DCC in SPR , we introduced the NLG-1 GRASP marker into a series of strains with mutations in molecules that have neuronal function and/or expression . We found that in clr-1 mutants , cell migration , axon guidance , and contact between PHB and AVA neurites appear normal ( Fig 1E and 1F ) , but NLG-1 GRASP intensity is severely reduced , indicating a reduction in synapse formation ( Fig 1G and 1H ) . We tested three loss-of-function alleles of clr-1 and all showed significantly reduced NLG-1 GRASP intensity ( Fig 1H ) . clr-1 encodes a receptor protein tyrosine phosphatase ( RPTP ) with immunoglobulin-like and fibronectin III domains , similar to the Leukocyte common Antigen-Related protein ( LAR ) family ( type IIa ) RPTPs [12 , 18] . In clr-1/RPTP ( e1745 ) mutants , PHB axon length is slightly reduced , but other mutants with similarly reduced axon length , such as ced-10/Rac1 ( n1993 ) , have normal NLG-1 GRASP intensity , indicating that a slight reduction in PHB-AVA contact is unlikely to result in decreased synapse formation ( S1A and S1B Fig ) . clr-1/RPTP promotes proper extension of highly branched PVD dendrites [19] . However , the extension of the AVA neurites was not decreased in clr-1/RPTP mutants ( S1C and S1D Fig ) . When PHB-AVA SPR is disrupted , PHB circuit function should be compromised . We tested this using a PHB circuit-specific behavioral assay . PHB sensory neurons mediate the nematode’s avoidance of sodium dodecyl sulfate ( SDS ) [4] . Wild-type animals will move backward into a control buffer for over one second , but stop backward movement into diluted SDS in approximately one third that time [10] . We compare these times using a response index ( RI ) , dividing the average time backing into SDS by the average time backing into control buffer , and normalizing wild-type to 100% . A larger RI indicates impaired PHB circuit function [10] . clr-1/RPTP mutants have a severe defect in SDS avoidance , indicating that PHB circuit function is disrupted ( Fig 1I ) . This defect in neural circuit function also indicates that reduction of NLG-1 GRASP fluorescence cannot be explained simply by a reduction in marker expression . However , to control for this possibility in another way , we also drove expression of GFP in PHB neurons and mCherry in AVA neurons in wild-type and clr-1/RPTP mutant animals with the same promoters used in the PHB-to-AVA NLG-1 GRASP marker , and measured average fluorescence intensity in the posterior region of AVA and the anterior region of PHB neurites . Expression levels were comparable between wild-type and clr-1 in both pre- and postsynaptic neurons , indicating that the reduction in NLG-1 GRASP fluorescence is not the result of a reduction in marker expression in either cell ( S2A Fig ) . We previously found that nlg-1 mutants respond to SDS normally [10] , thus impaired nlg-1/Neuroligin 1 expression or localization cannot explain the defect in PHB circuit function observed in clr-1/RPTP mutant animals . To determine if synaptic defects in clr-1/RPTP mutants are due to a defect in trafficking synaptic components to the distal synaptic region of PHB and AVA neurites , or if these defects might be due to a later recognition defect in SPR , we examined levels of synaptic components in the distal region of the neurites . Levels of the presynaptic vesicle marker RAB-3::mCherry and the presynaptic active zone marker GFP::ELKS-1 in PHB , and the postsynaptic marker NLG-1::YFP in AVA , are not reduced in clr-1/RPTP mutants ( S2B–S2N Fig ) . This is similar to observations in unc-6/Netrin and unc-40/DCC mutants , and consistent with a role for clr-1/RPTP in SPR [10] . To determine if clr-1/RPTP was expressed in PHB or AVA neurons , we conducted expression studies utilizing a pclr-1::GFP transcriptional fusion transgene in wild-type animals . GFP expression was observed in AVA neurons , but not in PHB neurons ( S3 Fig ) . To test if clr-1/RPTP can function in AVA neurons to promote PHB-AVA SPR , we generated a construct driving the expression of the clr-1 cDNA under control of an AVA-selective promoter ( pAVA::clr-1 ) and introduced it into clr-1/RPTP mutants . NLG-1 GRASP intensity and the behavioral output of the PHB circuit were rescued to wild-type levels in these transgenic animals , indicating that clr-1/RPTP can function in AVA neurons to direct SPR . Expression of the clr-1 cDNA in PHB neurons did not rescue NLG-1 GRASP intensity or the behavioral output of the PHB circuit in clr-1/RPTP mutants , consistent with a role for clr-1/RPTP in AVA neurons ( Fig 2A–2C ) . To test if the extracellular domain of clr-1/RPTP is necessary for SPR , we generated a construct in which the extracellular domain of the clr-1/RPTP cDNA was deleted and expressed under the direction of an AVA-selective promoter ( pAVA::clr-1/RPTPΔxcd ) , and introduced it into clr-1/RPTP mutants . Unlike the full-length clr-1/RPTP cDNA under the direction of the AVA promoter , pAVA::clr-1/RPTPΔxcd did not rescue either NLG-1 GRASP fluorescence or the behavioral output of the circuit to wild-type levels; NLG-1 GRASP fluorescence was 51% of wild-type levels , and the relative response index was 206% of wild-type levels ( Fig 2A–2C ) . This indicates that the extracellular domain is required for full SPR function . To determine if the phosphatase activity of clr-1/RPTP is required for PHB-AVA SPR , we introduced a point mutation into the pAVA::clr-1/RPTP construct that inactivates the membrane-proximal phosphatase domain [18] ( pAVA::clr-1/RPTPpd ) . Again , the NLG-1 GRASP intensity and behavioral output of the PHB circuit were not rescued to wild-type levels; NLG-1 GRASP fluorescence was 41% of wild-type levels , and the relative response index was 221% of wild-type levels ( Fig 2A–2C ) . This indicates that clr-1/RPTP phosphatase activity is required for full SPR function as well . To better understand the mechanism by which CLR-1/RPTP directs SPR , we visualized the subcellular localization of CLR-1/RPTP in AVA neurons . We generated a translational fusion of the clr-1/RPTP cDNA to YFP under the direction of an AVA-selective promoter ( pAVA::clr-1/RPTP::YFP ) . We found that CLR-1/RPTP was localized throughout the AVA axon , but was brightest in the preanal ganglion , where AVA neurons contact PHB neurons . Interestingly , within the region of the preanal ganglion , it was usually concentrated in the anterior half , where the majority of synapses between PHB and AVA normally form ( Fig 2D ) . We generated a similar translational fusion of the clr-1/RPTP cDNA to mCherry ( pAVA::clr-1/RPTP::mCherry ) and introduced it into animals expressing PHB-AVA NLG-1 GRASP . We similarly found that the NLG-1 GRASP signal was usually confined to the anterior half of the preanal ganglion where CLR-1/RPTP localized ( Fig 2E ) . This localization is consistent with a role for CLR-1/RPTP in SPR . CLR-1/RPTP could function in embryogenesis during establishment of correct SPR , later to maintain proper synaptic connectivity , or at both times . To test when CLR-1/RPTP functions , we took advantage of the temperature-sensitive nature of the clr-1/RPTP ( e1745 ) allele [18 , 20] . Interestingly , behavioral function was impaired in clr-1 ( e1745 ) animals placed at the restrictive temperature only during embryogenesis , or placed at the restrictive temperature only after embryogenesis , indicating that clr-1/RPTP function is required both during and after embryogenesis ( Fig 3A ) . To test when clr-1/RPTP function is necessary and sufficient , we shifted the animals immediately before the last stage of embryogenesis ( the 3-fold embryo ) and after the first larval stage ( L1 ) . Animals placed at the restrictive temperature only during the 3-fold embryo and L1 stages exhibited impaired behavior , while animals placed at the permissive temperature during the same stages exhibited normal behavior ( Fig 3A ) . This indicates that clr-1/RPTP function is necessary during the 3-fold embryo and L1 stages , the period when synaptogenesis is likely initiated and the stage directly after , consistent with a role in SPR . In addition , we observed NLG-1 GRASP fluorescence intensity in animals placed at the permissive or restrictive temperature during the 3-fold embryo and L1 stages . We observed a reduction in NLG-1 GRASP fluorescence intensity in clr-1/RPTP animals shifted from the restrictive to the permissive temperature at the 3-fold embryo stage , and shifted from the permissive to the restrictive temperature at the end of the L1 stage , possibly due to stress on the mutant animals from multiple temperature shifts . Even so , NLG-1 GRASP intensity was severely and significantly reduced from this level in animals moved to the restrictive temperature for the 3-fold embryo and L1 stages , compared with animals kept at the permissive temperature during the same periods ( Fig 3B and 3C ) . This is also consistent with a function during the 3-fold and L1 stages . If CLR-1/RPTP functions in the SPR signaling event between PHB and AVA neurons , increasing the expression of clr-1/RPTP in AVA neurons should direct additional synapse formation between these neurons . In fact , overexpression ( OE ) of pAVA::clr-1/RPTP was sufficient to drive a significant increase in PHB-AVA synaptogenesis ( Fig 4A and 4B ) . This is similar to the increase in PHB-AVA synaptogenesis observed in animals overexpressing unc-6/Netrin in AVA neurons [10] , and demonstrates an ability to drive synaptogenesis . To test if these synapses were functional , we also conducted behavioral analysis . We found that the response index was significantly smaller , consistent with potentiation of the circuit ( Fig 4C ) . To test if UNC-6/Netrin and UNC-40/DCC act with CLR-1/RPTP in promoting SPR , we generated double-mutants between clr-1/RPTP and unc-40/DCC , and clr-1/RPTP and unc-6/Netrin . These double-mutants did not have more severe SPR defects when compared with clr-1 single mutants , indicating that unc-40/DCC , unc-6/Netrin , and clr-1/RPTP function in the same SPR pathway ( Fig 5A and 5B ) . If clr-1/RPTP receives the unc-6/Netrin signal , clr-1/RPTP should act downstream of unc-6/Netrin . To test this , we introduced the clr-1/RPTP ( e1745 ) mutation into animals overexpressing unc-6/Netrin to determine if the high levels of synaptogenesis were suppressed . Indeed , PHB-AVA synapses were dramatically reduced ( Fig 5A and 5B ) , consistent with a role for clr-1/RPTP downstream of unc-6/Netrin in SPR . In addition , we generated trans-heterozygotes between the recessive clr-1/RPTP and unc-40/DCC mutants , and clr-1/RPTP and unc-6/Netrin mutants . NLG-1 GRASP fluorescence was significantly reduced in both trans-heterozygous strains , indicating that these genes likely function together in SPR ( Fig 5C and 5D ) . To determine if clr-1/RPTP affects other synaptic connections , we introduced the clr-1/RPTP mutation into a NLG-1 GRASP marker labeling synapses between AVA neurons and their postsynaptic partners , the VA and DA motorneurons [11] . Specifically , we assayed synapses between AVA and VA10 motorneurons . A cluster of AVA-VA10 synapses is localized between the VA10 and DA7 neurons in wild-type animals [11 , 21] , but was lost or reduced in most clr-1/RPTP mutants ( S4 Fig ) . This indicates that SPR defects in clr-1/RPTP mutants are not specific to PHB-AVA synapses , and that clr-1/RPTP has broader functions in synaptic partner recognition .
Here , we demonstrate a novel role for clr-1/RPTP in promoting SPR between neurons in complex in vivo environments . In clr-1/RPTP mutants , fluorescence of the NLG-1 GRASP marker labeling synapses between PHB sensory neurons and AVA interneurons is severely reduced , indicating a reduction of synapses between these neurons . In addition , a PHB circuit-specific behavioral response is compromised , consistent with a loss of synaptic function . A transcriptional fusion of clr-1/RPTP showed expression in AVA , but not PHB neurons , and expression of clr-1/RPTP in AVA neurons was sufficient to rescue the clr-1/RPTP mutant defects , indicating a postsynaptic role . Deletion of the clr-1/RPTP extracellular domain or a mutation in the catalytic site of the phosphatase domain compromises this rescue , indicating that both domains are required for full SPR activity . Overexpression of clr-1/RPTP in AVA neurons results in increased NLG-1 GRASP fluorescence and potentiates circuit function , suggesting that postsynaptic clr-1/RPTP is sufficient to promote synaptogenesis . Our genetic analysis indicates that clr-1/RPTP acts in the same SPR pathway as unc-6/Netrin and unc-40/DCC . clr-1/RPTP is also required for synaptogenesis between AVA and its postsynaptic partner , the VA10 motorneuron , indicating that clr-1/RPTP may have a broader role in SPR between other neurons . Our work describes the first role in synaptogenesis for clr-1 , and the first postsynaptic role in synaptogenesis for a LAR family member in C . elegans . The LAR family member ptp-3 has also been studied in C . elegans . ptp-3 isoform A is presynaptic and required for proper presynaptic morphology at neuromuscular junctions , indicating a role in presynaptic assembly [22 , 23] . In Drosophila and vertebrate systems , LAR family members have also been found to regulate synaptogenesis , although the focus of most work has been on presynaptic roles for these proteins [24–26] . Yet , LAR has also been found at postsynaptic sites in vertebrate systems [27–29] . PHB-AVA synapses can be visualized and their circuit function tested in intact , live animals , in a genetically tractable system , making them a powerful system in which to understand the roles of LAR family members in postsynaptic cells , and identify their interactors . LAR family members can act with molecules in the same cell and in their presumptive synaptic partners to mediate synapse formation . Studies in invertebrate and vertebrate systems have demonstrated that one mechanism by which LARs can promote presynaptic assembly is via their interaction with liprins [30–32] . Liprin-α interacts with several other active zone proteins , providing a mechanism by which presynaptic components may be recruited to sites of LAR binding [26 , 32] . LAR family RPTPs can act in trans with several proteins including Netrin-G ligand , IL1RAPL , neurotrophin receptor tyrosine kinase C , Slit- and Trk-like proteins , and synaptic adhesion-like molecule 3 , and in cis with the heparin sulfate proteoglycans glypican and syndecan [33–40] . Work in cultured cells demonstrates that the Netrin family member Netrin-G1 can interact in cis with LAR to mediate synaptogenesis [41] . This is similar to what we observe at PHB-AVA synapses , where both UNC-6/Netrin and CLR-1/RPTP function in the same cells , and genetic evidence shows that they act in the same pathway to promote SPR . However , in cultured cells , the receptor on the opposing cell is the Netrin G Ligand NGL-1 [41] . In PHB-AVA SPR , we find that the receptor functioning in the opposing partner is Netrin’s canonical receptor DCC . CLR-1/RPTP , UNC-6/Netrin and UNC-40/DCC have previously been shown to act together in mechanosensory neuron axon guidance . However , rather than promoting UNC-40/DCC function in trans as in SPR , CLR-1/RPTP acts in cis to negatively regulate UNC-40/DCC function or the function of downstream signaling molecules [12] . Thus , this study defines a new mechanism by which synaptic partner recognition is mediated . Our previous work suggests a model in which limiting amounts of UNC-6/Netrin are secreted from postsynaptic AVA neurons , binding UNC-40/DCC in presynaptic neurons to specify them as the correct presynaptic partners [10] . Our current work indicates CLR-1/RPTP likely transduces the SPR signal into presumptive postsynaptic AVA neurons , promoting synaptogenesis . We propose that CLR-1/RPTP may interact in trans with UNC-6/Netrin , UNC-40/DCC , or a ligand or receptor that is yet to be identified , to generate PHB-AVA SPR ( Fig 6 ) . The requirement of the clr-1/RPTP extracellular domain for full rescue of the clr-1/RPTP defect is consistent with this model . The small degree of SPR rescue is consistent either with a low level of CLR-1/RPTP activation in the absence of its extracellular domain , or by a minority of CLR-1/RPTP synaptogenic activity not requiring a trans-interaction . CLR-1/RPTP requires phosphatase activity for the majority of its function in SPR . Although phosphatase activity of Drosophila LAR is not required for photoreceptor axon targeting [42] or viability [43] , LAR phosphatase activity is required for many processes , including growth of neuromuscular junctions [40] . Several LAR family substrates have been identified , including N-cadherin , β-catenin , Abelson kinase , Enabled , Trio , p250RhoGAP , and multiple tyrosine kinases , and their regulation by LAR may modulate synaptic adhesion and actin dynamics [44] . This study defines a new pathway by which SPR is governed . The ability of CLR-1/RPTP to function in postsynaptic cells with secreted UNC-6/Netrin and presynaptic UNC-40/DCC also demonstrates a new mode of action for these conserved molecules . The distinct molecular pathways and sites of action discovered for LAR family members may allow these same molecules to govern different processes in neural circuit formation and other developmental processes . Understanding how Netrin , DCC , and LAR-RPTP family proteins act together to regulate cell-cell signaling may be important for human health , as these genes are associated with neurological disorders , such as schizophrenia [13 , 14] , as well as cancer [16 , 17 , 45] .
All worms were maintained according to standard protocols [46] and were raised on 6 cm NGM plates seeded with OP50 Escherichia coli at 20°C , except for the worms used for the temperature shift experiment , which were raised at 16°C when noted . Wild-type strains were C . elegans variety Bristol , strain N2 . Mutants used for this study include clr-1 ( e1745 ) II , clr-1 ( e2530 ) II , clr-1 ( n1992 ) II , [18 , 20] , unc-6 ( ev400 ) X [47] , unc-40 ( e271 ) I [20] and ced-10 ( n1993 ) IV [48] . Except for strains containing wyEx1364 , wyEx1402 , iyEx323 , and iyIs8 ( see below ) , all strains contain the integrated PHB-AVA NLG-1 GRASP marker wyIs157 IV ( pSM::pgpa-6::nlg-1::spGFP1-10 ( 60 ng/μl ) , pSM::pflp-18::nlg-1::spGFP11 ( 30 ng/μl ) , pSM::pnlp-1::mCherry ( 10 ng/μl ) , pSM::pflp-18::mCherry ( 5 ng/μl ) and podr-1::DsRed2 ( 20 ng/μl ) ) [10] . Transgenic lines used in this study include lines carrying pAVA::clr-1/RPTP for both cell-specific rescue of clr-1 ( e1745 ) II ( Fig 2A–2C ) , and overexpression in AVA neurons ( Fig 4A–4C ) : iyEx97 ( pSM::prig-3::clr-1 ( 40 ng/μl ) , punc-122::RFP ( 20 ng/μl ) ) and iyEx101 ( pSM::prig-3::clr-1 ( 60 ng/μl ) , punc-122::RFP ( 20 ng/μl ) ) . Transgenic lines carrying pPHB::clr-1/RPTP include iyEx364 , iyEx365 , and iyEx366 ( pSM::pnlp-1::clr-1 ( 1–5 ng/μl ) , punc-122::RFP ( 22 ng/μl ) ) ( Fig 2A–2C ) . Transgenic lines carrying pAVA::clr-1/RPTPΔxcd include iyEx133 , iyEx134 , iyEx135 , and iyEx362 ( pSM::prig-3::clr-1Δxcd::mCherry ( 60 ng/μl ) , punc-122::RFP ( 20 ng/μl ) ) ( Fig 2A–2C ) . Transgenic lines carrying pAVA::clr-1/RPTPpd include iyEx169 , iyEx170 and iyEx174 ( pSM::rig-3::clr-1pd::mCherry ( 60 ng/μl ) , punc-122::RFP ( 21 ng/μl ) ) ( Fig 2A–2C ) . The transgenic line generated to determine the subcellular localization of CLR-1/RPTP in AVA neurons was iyEx121 ( pSM::prig-3::clr-1::YFP ( 85 ng/μl ) , punc-122::RFP ( 20 ng/μl ) ) . The transgenic line generated to visualize CLR-1/RPTP co-localization with NLG-1 GRASP contained iyEx175 ( pSM::prig-3::clr-1::mCherry ( 60 ng/μl ) , punc-122::RFP ( 20 ng/μl ) ) and iyIs2 ( pSM::pgpa-6::nlg-1::GFP1-10 ( 30 ng/μl ) , pSM::pflp-18::nlg-1::GFP11 ( 15 ng/μl ) , podr-1::DsRed2 ( 50 ng/μl ) ) , in a clr-1 ( e1745 ) background ( Fig 2D and 2E ) . For overexpression of unc-6/Netrin in AVAs in wild-type and clr-1/RPTP mutant backgrounds , the transgenic line used was iyEx47 [10] ( Fig 5A and 5B ) . For measurement of AVA neurites length , we generated the transgenic line wyEx1364 ( pSM::pgpa-6::GFP ( 50ng/μl ) , pSM::pflp-18::mCherry ( 10 ng/μl ) and punc-122::RFP ( 20 ng/μl ) ) ( S1C and S1D Fig ) . Transgenic lines used to assay expression levels from the promoters that drive PHB-AVA NLG-1 GRASP were wyEx1402 ( pSM::pgpa-6::GFP ( 50 ng/μl ) and podr-1::DsRed2 ( 20 ng/ μl ) ) and wyIs157 ( S2A Fig ) . The transgenic lines used to visualize localization of presynaptic and postsynaptic markers in wild-type and clr-1 mutant animals were wyEx2309 for RAB-3 , iyEx82 for NLG-1 , and iyEx83 for ELKS-1 [10] ( S2B–S2N Fig ) . The transgenic lines generated to determine the expression pattern of clr-1 in the posterior region was iyEx61 ( pSM::pclr-1::GFP ( 50ng/μl ) , punc-122::RFP ( 20 ng/μl ) ) with wyIs157 , and in the head was iyEx323 ( pSM::pclr-1::GFP ( 20 ng/μl ) , pSM::prig-3::mCherry ( 17 ng/μl ) , punc-122::RFP ( 22 . 5 ng/μl ) ) ( S3 Fig ) . To test if clr-1/RPTP is required for AVA-VA synapses , we stably integrated the wyEx1845 AVA-VA/DA NLG-1 GRASP marker ( pSM::punc-4::nlg-1::spGFP1-10 ( 20 ng/μl ) , pSM::pflp-18::nlg-1::spGFP11 ( 30 ng/μl ) , pSM::punc-4::mCherry ( 5 ng/μl ) , podr-1::DsRed2 ( 50 ng/μl ) ) [10] into the genome to generate iyIs8 ( S4 Fig ) . Constructs were generated using standard molecular techniques . To generate pSM::prig-3::clr-1 ( referred to as pAVA::clr-1/RPTP in the text ) , an N2 C . elegans cDNA library was generated by isolating C . elegans mRNA using Sigma Tri Reagent ( TRIzol ) to break the worm cuticle , chloroform to isolate RNA , and the Qiagen RNeasy Mini kit to purify the RNA . The Invitrogen SuperScript II Reverse Transcriptase kit was used to generate cDNA . clr-1 cDNA was amplified using clr-1-specific primers ( MVP281: AGACGTCGACATGCGAATAAATCGATGGATC and MVP282: TATTTGGTACCCTACCTATATGTCTTAGAGATA ) that introduced the SalI and Acc65I restriction sites . The clr-1 cDNA was subcloned into the SalI-Acc65I fragment from pSM::prig-3 , which was made by subcloning the rig-3 promoter from pSM::prig-3::MT::unc-6 [10] using SphI and AscI restriction sites into the SphI-AscI fragment from pSMΔ ( a gift from S . McCarroll ) . To generate the pSM::pclr-1::GFP construct , ( referred to as pclr-1::GFP in the text ) the clr-1 promoter ( 4480 bp upstream of the clr-1 start site ) was amplified from genomic DNA using pclr-1 specific primers ( MVP227: TAACGGCGCGCCGAGAATGAGGTTACGATCTAC and MVP228: ACATACCCGGGGTTTCCGCGTTAATTTAAAAGCC ) that introduced AscI and SmaI restriction sites . Then , the pclr-1 fragment was subcloned into the AscI-SmaI fragment from pSM::GFP ( a gift from S . McCarroll ) . To generate pSM::prig-3::clr-1::YFP ( referred to as pAVA::clr-1/RPTP::YFP in the text ) , the clr-1 cDNA ( without its stop codon ) was amplified from pSM::prig-3::clr-1 using clr-1 specific primers that introduced the SalI and Acc65I restriction sites ( MVP344: TATTGGTACCCCTATATGTCTTAGAGATATAG and MVP281: AGACGTCGACATGCGAATAAATCGATGGATC ) . The clr-1 cDNA was subcloned into the SalI-Acc65I fragment from pSM::prig-3::YFP . pSM::prig-3::YFP was made by amplifying the YFP fragment from pSM::prig-3::unc-6::YFP [10] using primers that introduced a 10GS linker as well as the Acc65I and SacI restriction sites ( MVP338: AGACGGTACCGGATCTGGATCTGGATCTGGATCTGGATCTATGAGTAAAGGAGAAGAACTT and MVP339: TATTGAGCTCCTATTTGTATAGTTCATCCATG ) . The 10GS linker-YFP fragments were then subcloned into the Acc65I-SacI fragment of pSM::prig-3 . To generate pSM::prig-3::clr-1::mCherry ( referred to as pAVA::clr-1/RPTP::mCherry in the text ) , the clr-1 cDNA ( without its stop codon ) was amplified from pSM::prig-3::clr-1 using clr-1 specific primers that introduced the SalI and Acc65I restriction sites ( MVP344: TATTGGTACCCCTATATGTCTTAGAGATATAG and MVP281: AGACGTCGACATGCGAATAAATCGATGGATC ) . The clr-1 cDNA was subcloned into the SalI-Acc65I fragment from pSM::prig-3::mCherry . pSM::prig-3::mCherry , was made by amplifying mCherry from pttx-3::mCherry [49] using primers that also introduced a 10GS linker and Acc65I and SacI restriction sites ( MVP340: AGACGGTACCGGATCTGGATCTGGATCTGGATCTGGATCTATGGTCTCAAAGGGTGAAGA and MVP341: TATTGAGCTCCTTATACAATTCATCCATGCC ) . The 10GSlinker::mCherry was then subcloned into the Acc65I-SacI fragment of pSM::prig-3 to generate pSM::prig-3::mCherry . To generate pSM::prig-3::clr-1Δxcd::mCherry ( referred to as pAVA::clr-1/RPTPΔxcd in the text ) , the transmembrane and intracellular domains of the clr-1 cDNA were amplified from pSM::prig-3::clr-1 , adding SalI and Acc65I sites ( MVP342: AGACGTCGACGCGTATGGATATTCTGCATACT and MVP344: TATTGGTACCCCTATATGTCTTAGAGATATAG ) and subcloned into pSM::prig-3::mCherry using the SalI and Acc65I sites . pSM::prig-3::clr-1pd::mCherry ( referred to as pAVA::clr-1/RPTPpd in the text ) was generated using the Stratagene QuikChange site-directed mutagenesis kit to make a point mutation in the catalytic sequence of the D1 intracellular phosphatase domain of pSM::prig-3::clr-1 , the domain predicted to be active . The cysteine at position 1013 was replaced with a serine ( C1013S ) . Specifically , we changed G at 3038 to C ( G3038C ) [18] . pSM::pnlp-1::clr-1 ( referred to as pPHB::clr-1/RPTP in the text ) was generated by subcloning the nlp-1 promoter , flanked by the SphI and SmaI restriction sites , from pSM::pnlp-1 into the SphI-SmaI fragment of pSM::prig-3::clr-1 . A Zeiss Axio Imager . A1 compound fluorescent microscope and a Zeiss LSM710 confocal microscope were used to capture images of live C . elegans under 630X magnification . Worms were anesthetized on 2% agarose pads using a 2:1 ratio of 0 . 3 M 2 , 3-butanedione monoxime ( BDM ) and 10 mM levamisole in M9 buffer . All micrographs taken were of larval stage 4 ( L4 ) animals , except micrographs in S2B to S2E Fig were of L2 animals . All data from micrographs were quantified using NIH ImageJ software [50] , as previously described [10] . Briefly , PHB-AVA NLG-1 GRASP intensity was measured by outlining each cluster of puncta and measuring the intensity at each pixel . To account for differences in background fluorescence , background intensity was estimated by calculating the minimum intensity value in a region immediately around the puncta . This approximated value was then subtracted from the intensity for each pixel , and the sum of the adjusted values was calculated . Median intensity values were normalized to wild-type levels measured on the same day using the same settings . To measure neurite intensity , we used segmented line tool with width 10 to measure the average intensity in the anterior PHB axon and posterior AVA neurite . To account for differences in background fluorescence , background intensity was estimated by drawing a similar line in a region next to the neurite , and subtracting this value from the average intensity per pixel . PHB axon length was measured from the cell body to the distal tip of the axon . AVA axon extension was assessed by measuring the distance from the posterior end of the AVA dendrite to the anal sphincter cell . To test the PHB-AVA circuit function , we utilized a high-throughput SDS avoidance assay based on the previously published SDS dry drop test [4 , 10] . Briefly , a day one adult hermaphrodite is placed on a dry , unseeded NGM plate . The worm is touched by a hair pick on its nose to induce backward motion by stimulating the ASH sensory neuron . Once the animal starts backing , a drop of M13 buffer or repellent ( M13 buffer with 0 . 1% SDS ) is placed on the agar behind the tail of the moving worm using a mouth pipette . The droplet placed on the agar is absorbed into the agar and the animal backs into the dry drop . The response time of the worm is the time the animal takes to stop backing into the dry drop . The response time for a minimum of 40 worms with the control M13 buffer and 40 worms to 0 . 1% SDS ( in M13 ) for each genotype was recorded . The relative response index was measured by dividing the mean backing time into SDS by the mean backing time into buffer . This calculated value was then divided by the same value for wild-type animals assayed on the same day to normalize the wild-type response index to 100% . Only animals that are able to move backwards can be tested . clr-1 ( e2530 ) and ( n1992 ) did not back sufficiently for behavioral testing . In the figures , the results are reported in the form of P-values ( *P < 0 . 05 , ** P < 0 . 01 , *** P < 0 . 001 , NS P > 0 . 05 ) . P-values provide accurate information about whether two samples differ significantly . P-values are generated by a procedure that incorporates both the sample sizes and variability in samples , so that the reader is not required to multiply each standard error of the mean ( obtained by assessing the length of error bars ) by a factor that depends on each sample size . Error bars ( standard error of the mean , or SEM ) are included for the behavioral measurements throughout the manuscript . However , the NLG-1 GRASP intensity data is not normally distributed , as previously described [10] , and therefore it would not be meaningful to include the SEM . Instead , we have calculated 95% confidence intervals for the medians through the bootstrap method ( using the DescTools package in R [51] ) for the NLG-1 GRASP data throughout the manuscript , and included them in S1 Table . For comparing more than two NLG-1 GRASP relative median intensity values , a Kruskal-Wallis test , a nonparametric alternative to ANOVA that does not rely on a normality assumption , was first used . If a Kruskal-Wallis test yielded a P-value less than 0 . 05 , or when comparing only two NLG-1 GRASP relative median intensity values , pair-wise comparisons were made using the Mann-Whitney U-test , which is a non-parametric significance test that compares medians of two independent groups . If more than one independent test was performed , P-values were adjusted for multiple comparisons using the Hochberg procedure . If pair-wise comparisons are not independent , the Hochberg procedure is no longer appropriate , as in Fig 5D . The Hochberg procedure is a standard method used to adjust for the tendency to incorrectly reject a null hypothesis for multiple comparisons , and can conservatively increase P-values . For behavior analysis , relative SDS response indices were compared using the t-test , and P-values were adjusted for multiple comparisons using the Hochberg method . Statistical significance was confirmed by conducting a multi-way ANOVA model with appropriate interaction terms using the linear model procedure ( R Development Core Team , 2009 ) . For axon length , mean values were compared using an ANOVA F-test , and pair-wise comparisons were made using the t-test , and adjusted for multiple comparisons using the Hochberg method . Microsoft Excel and R statistical computing software [51] were used for all statistical tests .
|
The nervous system is required for many body functions including perception , behavior and thought . Cells in the nervous system called neurons function in interconnected groups called circuits to carry out these basic functions . While we have learned a great deal about how circuits function , we know much less about how they are set up during development . Discovering the mechanisms that organize these neural circuits could help us to understand neurological disorders that may result from defects in this process . Our work has identified a key role for the cell surface molecule CLR-1 in a critical step in the formation of neural circuits: the recognition between neurons that must link together . We find that CLR-1 acts with the ligand Netrin and its receptor Deleted in Colorectal Cancer ( DCC ) to mediate communication between adjacent cells . Interestingly , all three of these molecules have been linked to schizophrenia and to cancer , indicating that our discovery may help inform our understanding of these diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"nervous",
"system",
"enzymes",
"caenorhabditis",
"enzymology",
"electrophysiology",
"neurites",
"neuroscience",
"animals",
"phosphatases",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
"animal",
"behavior",
"experimental",
"organism",
"systems",
"nerve",
"fibers",
"zoology",
"neuronal",
"dendrites",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"proteins",
"axons",
"behavior",
"neural",
"pathways",
"biochemistry",
"cellular",
"neuroscience",
"eukaryota",
"neuroanatomy",
"cell",
"biology",
"anatomy",
"synapses",
"physiology",
"neurons",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neurophysiology",
"organisms"
] |
2018
|
The receptor protein tyrosine phosphatase CLR-1 is required for synaptic partner recognition
|
Hedgehog signaling plays conserved roles in controlling embryonic development; its dysregulation has been implicated in many human diseases including cancers . Hedgehog signaling has an unusual reception system consisting of two transmembrane proteins , Patched receptor and Smoothened signal transducer . Although activation of Smoothened and its downstream signal transduction have been intensively studied , less is known about how Patched receptor is regulated , and particularly how this regulation contributes to appropriate Hedgehog signal transduction . Here we identified a novel role of Smurf E3 ligase in regulating Hedgehog signaling by controlling Patched ubiquitination and turnover . Moreover , we showed that Smurf-mediated Patched ubiquitination depends on Smo activity in wing discs . Mechanistically , we found that Smo interacts with Smurf and promotes it to mediate Patched ubiquitination by targeting the K1261 site in Ptc . The further mathematic modeling analysis reveals that a bidirectional control of activation of Smo involving Smurf and Patched is important for signal-receiving cells to precisely interpret external signals , thereby maintaining Hedgehog signaling reliability . Finally , our data revealed an evolutionarily conserved role of Smurf proteins in controlling Hh signaling by targeting Ptc during development .
Hedgehog ( Hh ) signaling is evolutionarily conserved and is essential for patterning of organs of both invertebrates and vertebrates [1] , [2] . Dysregulation of Hh signaling activity leads to developmental abnormalities and cancers [3] . In Drosophila , genetic and biochemical evidence revealed that two multi-span transmembrane proteins , Patched ( Ptc , 12-span ) and Smoothened ( Smo , 7-span ) , serve as a reception system for Hh signal transduction in Hh-receiving cells [4] , [5] . In the absence of Hh ligands , Ptc inhibits Smo activity , thereby blocking Hh signaling transduction , while in the presence of Hh , the Hh ligands , in concert with its co-factor iHog , physically interacts with Ptc and alleviates its inhibition on Smo , resulting in Smo accumulation on the cell surface and a conformational switch of Smo to an active form that regulates the distinct downstream target genes in an Hh concentration-dependent manner [6]–[12] . In Hh signaling , the mechanisms by which Smo activation leads to downstream target gene expression are generally understood [2] , [3] . However , less is known about the regulation of the Hh signaling receptor Ptc . Given that ptc is a direct target of the Hh pathway and that Ptc itself negatively regulates Hh signaling [9] , Ptc expression must be tightly controlled to ensure proper Hh signal transduction . Previous studies have also shown that endogenous Ptc protein in Hh-receiving cells exhibits both plasma membrane and punctate-distribution patterns upon Hh ligand stimulation [13] , [14] , suggesting that Hh signal potentially promotes Ptc turnover . However , the molecular mechanism underlying Ptc degradation in response to Hh signal remains largely unknown . Protein turnover mediated by ubiquitin modification plays important roles in the regulation of numerous cellular processes during development . The enzymatic reaction of protein ubiquitination is a highly ordered multi-step process involving three classes of enzymes , including ubiquitin-activating enzymes ( E1s ) , ubiquitin-conjugating enzymes ( E2s ) , and ubiquitin ligases ( E3s ) [15] , [16] . E3 ubiquitin ligases are crucial in the ubiquitin conjugation cascade because of their roles in the recruitment of ubiquitin-loaded E2s and their selective recognition of target proteins . Generally , the E3 ubiquitin ligases are classified into three subfamilies: the really interesting new gene ( RING ) finger domain containing E3s , the homologous to E6-AP carboxyl terminus ( HECT ) domain containing E3s , and the U box E3s [15] , [17] . Previous studies have shown that Neural precursor cell expressed , developmentally downregulated 4 ( Nedd4 ) , one member of the C2-WW-HECT family proteins , could physically associate with the Ptc protein [13] , [18]; however , whether the Nedd4 is involved in the regulation of Hh signaling activity through its interaction with Ptc remains unknown . Smad ubiquitin regulatory factor ( Smurf ) proteins are other members of the C2-WW-HECT E3 family of proteins that contain typical WW and HECT domains . Smurf proteins ( including Smurf1 and Smurf2 in mammals ) were originally identified as an E3 ubiquitin ligase for the degradation of R-Smad proteins and type I receptors to negatively regulate TGFβ/BMP signal [19]–[25] . Recently , Smurfs have also been shown to regulate cell motility by targeting RhoA for ubiquitin-mediated degradation [26] , [27] , and are involved in the non-canonical Wnt signaling to regulate planar cell polarity by degrading the PCP core component , Prickle1 protein [28] . We recently uncovered that Smurf functions in concert with Fused , a serine/threonine kinase that regulates Hedgehog signaling , to degrade the BMP type I receptor Tkv allowing for bam expression in differentiating cystoblast cells , thereby determining the fate of Drosophila germline stem cells [22] , [24] . These studies revealed that Smurf proteins have diverse biological functions through regulating multiple signal pathways in different cellular contexts . In this study , we identified a novel role of Smurf E3 ligase in the regulation of Hh signaling by directly controlling Ptc protein turnover . Moreover , we found that Smurf mediates Ptc degradation in a manner that depends on Smo signaling activity . These findings revealed a novel mechanism by which an Hh signaling-dependent bidirectional control mechanism involving Ptc and Smurf is important for signal-receiving cells to precisely interpret external signal , thereby maintaining the reliability of Hh signaling transduction . Finally , we provided evidence that zebrafish Smurf proteins also modulate Hh signaling by controlling Ptc1 protein degradation to regulate late somitogenesis during zebrafish embryogenesis . Thus , our data support the conclusion that Smurf proteins have an evolutionarily conserved role in controlling Hh signaling during development .
To explore novel biological functions of Smurf , we searched for Smurf-interacting partners by performing a yeast two-hybrid screen using the full-length Drosophila Smurf as bait . From this screen , we found three positive clones that encode two different C-terminal fragments of the Drosophila Ptc protein , Ptc-B4/5 and Ptc-L1 ( Figure 1A; Table S1 ) , indicating that Ptc could be a potential candidate as a Smurf-interacting partner . To confirm the interaction between Ptc and Smurf proteins , we then performed co-immunoprecipitation experiments , and found that Flag-tagged Smurf and Myc-tagged Ptc co-immunoprecipitated with each other in transfected S2 cells ( Figure 1B and 1C ) . Consistent with this , endogenous Smurf protein could be co-immunoprecipitated by anti-Ptc antibody ( Figure 1D ) , revealing that Smurf and Ptc proteins are able to form a complex together in S2 cells . In agreement with the yeast-two-hybrid assay , domain-mapping analysis of Ptc revealed that the C-terminal tail was necessary and sufficient for Ptc to interact with Smurf ( Figure 1E , 1G , and 1H ) . To further define the essential domain ( s ) in Smurf required for its interaction with Ptc , we employed a series of previously described truncated forms of Smurf ( Figure 1F ) [24] in co-immunoprecipitation experiments . As shown in Figure 1I , while the C2 and HECT domains of Smurf were dispensable for its interaction with Ptc , the WW domains were essential for Smurf to form a complex with Ptc , suggesting that Smurf interacts with Ptc through its WW domains . WW domains are typical domains in C2-WW-HECT subfamily E3 ligases for substrate recognition through their interaction with a proline-rich sequence ( PPXY motif ) in substrate proteins [15] , [29] . Given that the C-terminal region of Ptc contains a consensus PPXY motif that has been shown to be required for the Ptc interaction with Nedd4 , the other member of C2-WW-HECT subfamily E3 ligases [13] , [18] , [30] , and that mutations of PPXY in the C-terminal region of Ptc abolished the Ptc-Smurf interaction , as indicated in the GST pull-down and yeast-two-hybrid assays ( Figures S1A and S8A ) , we reasoned that Ptc might be a strong candidate as a substrate of Smurf and determined whether Smurf regulates Ptc stability in S2 cell cultures . As shown in Figure 2A and 2B , overexpression of Smurf apparently downregulated the full-length Ptc , but not mutant Ptc ( PtcΔCTD ) that lacks the C-tail of Ptc , which showed more stability than the full-length Ptc ( Figure S1B ) . Interestingly , the level of Ptc protein could be significantly increased when cells were treated with either MG132 , a proteasome inhibitor , or NH4Cl , a lysosome inhibitor ( Figure 2C and 2D ) , and Smurf promotes both Ub K48-linked and Ub K63-linked ubiquitination of Ptc , as indicated by multiple co-immunoprecipitation experiments followed by Western blot assays ( Figures 2E , 2F , and S1C–S1F ) , suggesting the stability of Ptc is regulated through both proteasome and lysosome pathways . Consistent with this observation , the downregulation of the full-length of Ptc protein mediated by co-expression of Smurf was markedly blocked by treatments with both MG132 and NH4Cl ( Figure 2G ) . Collectively , our data argue that Smurf promotes Ptc degradation through its C-tail in cell cultures . We then determined whether Smurf promotes Ptc ubiquitination by performing ubiquitination assays in S2 cells according to the methods described previously [31]–[33] . As shown in ubiquitination assays , overexpression of the wild-type form of Smurf significantly increased the ubiquitin conjugation of full-length Ptc or the C-tail of Ptc , but the catalytic dead form of Smurf ( SmurfC1029A ) [19] did not ( Figures 2H and S2A–S2E ) . We noted that overexpression of SmurfC1029A reduced ubiquitination of Ptc , suggesting its dominant negative role in regulating Ptc ubiquitination in S2 cells , likely due to the fact that SmurfC1029A interacted with Ptc in cultured cells , and thus competitively inhibited endogenous Smurf function ( Figure S2F ) . Consistent with these observations , we found that knockdown of Smurf markedly reduced the full-length Ptc ubiquitination in S2 cells ( Figure 2I ) . To ask whether the C-tail of Ptc is responsible for its ubiquitination mediated by Smurf , we co-overexpressed PtcΔCTD and Smurf in S2 cells . As shown in Figure 2J , this truncated form of Ptc was apparently resistant to ubiquitination by Smurf E3 ligase . In agreement with these findings , the in vitro ubiquitination assays revealed that the C-tail of Ptc could be directly ubiquitinated by Smurf E3 ligase ( Figure S2G ) . Taken together , our results support that Smurf promotes Ptc ubiquitination and degradation through its C-tail . To determine the biological link between Smurf and Hh signaling , we used Drosophila wing discs and adult wings as an assay system to assess the in vivo role of smurf in Hh signaling during development . In wing discs , Hh produced in the posterior compartment cells induces the anterior compartment cells adjacent to the A/P boundary to express dpp and ptc , which respond to low levels and higher levels of Hh activity , respectively . It has been documented that , while activation of Hh signaling is essential for the proper patterning of the intervein region between L3 and L4 of adult wing , downregulation of Hh signaling causes the reduction of the intervein region between L3 and L4 [34] . To test whether smurf is involved in the regulation of Hh signaling , we first measured the area between L3 and L4 space and the whole wing in wild type and smurf mutants . As shown in Figure 2K–2M , loss of smurf led to a slight but evident reduction ( n = 18 , p<0 . 01 ) of the intervein region between L3 and L4 , compared to the wild-type control , suggesting that smurf might function as a positive modulator in Hh signaling pathway . To confirm this observation , we then overexpressed Flag-tagged Smurf or Flag-tagged SmurfC1029A in the wing discs using the MS1096-gal4 , a driver that has higher activity in the dorsal compartment of wing discs , and measured the expression levels of Ci and dpp-lacZ , the responsive reporter to the low level of Hh signaling . As shown in Figure 3A , 3B–3B″ , and 3C–3C′″ , overexpression of the wild-type form of Smurf , rather than SmurfC1029A , was able to not only increase expression of dpp-lacZ , but also slightly expanded the domain of both Ci and dpp-lacZ toward the P compartment . Similar results were obtained when the ap-gal4 , a dorsal compartment specific driver , was employed ( Figure S3A–S3A″ , S3B–S3B″ , and S3C–S3C″ ) . We then expressed smurf RNAi transgene , P{uasp-shmiR-smurf} [35] , to specifically knockdown smurf in wing discs using the MS1096-gal4 or the ap-gal4 driver , and found that inactivation of smurf decreased expression of dpp-lacZ ( Figures 3D–3D′″ and S3C–S3C″ ) . To further confirm these observations , we then performed the mosaic clonal analysis . Consistently , dpp-lacZ expression was significantly downregulated in smurf mutant cell clones located in the A/P boundary ( Figure 3E and 3F–3F″ ) . Taken together , our findings suggest that Smurf functions as a positive regulator in the Hh signaling pathway in wing discs . We noted that overexpression of smurf by MS1096-gal4 driver markedly increased dpp-lacZ expression , but did not apparently affect the levels of both Ptc protein and ptc-lacZ , which responds to a higher level of Hh signaling ( Figure 3G–3H′ ) . Accordingly , knockdown of smurf did not dramatically affect the global levels of Ptc protein either ( Figure 3D′ ) . One possibility could be that ptc transcripts and Ptc protein can compensate for each other , owing to the negatively regulatory role of Ptc protein in Hh signaling . To validate this issue , we blocked new Ptc protein synthesis from Ptc transcripts by pre-treatment of the discs with the protein synthesis inhibitor cycloheximide ( CHX ) , and then assessed whether Smurf controls the Ptc protein turnover . As shown in Figure S3D–S3K , after the CHX treatment , overexpression of smurf by the MS1096-gal4 greatly reduced the Ptc protein expression level , whereas knockdown of smurf evidently increased Ptc levels , compared to the discs without the CHX treatment . To confirm these findings , we then expressed the Ptc by the en-gal4 driver in smurf mutants . As shown in Figure 3J–3L , expression of PtcWT led to much more severe wing phenotype in smurf mutants than that in wild-type background . These results reveal that Smurf positively regulates Hh signaling through controlling Ptc turnover . Given that the function of Smurf that targets Ptc is restricted to the A/P boundary cells , which exhibit high levels of Hh signaling activity , we reasoned that Smurf might regulate Ptc protein turnover via responding to the high levels of Hh signaling activity . However , a recent study proposed that Ptc degradation might depend on Ptc itself , since a high level of Ptc is also present in the A/P boundary cells [30] . To distinguish these two potential mechanisms by which Smurf mediates Ptc degradation , we employed the P{hs-ptc:gfp} transgene to overexpress Ptc:green fluorescent protein ( GFP ) in the wing discs by heat-shock treatment , which is independent of Hh signaling . As shown in Figure 4A–4A′ and 4B–4B′ , overexpression of either the wild-type Smurf or SmurfC1029A by the ap-gal4 driver did not affect the expression of Ptc:GFP induced by heat-shock treatment in the wing discs , suggesting that expression of Ptc itself is not sufficient to induce Smurf-mediated Ptc degradation . To test whether Smurf could regulate Ptc turnover in response to Hh signaling , we next established an ectopic Hh signaling activation system by overexpression of the activated forms of Smo ( SmoSD12 or SmoSD123 ) [11] in the A-compartment of wing discs . As shown in Figure 4C–4J , overexpression of either SmoSD12 or SmoSD123 was sufficient to upregulate Ptc expression ( Figure 4D and 4G ) . In particular , the high level of Ptc in the A-compartment cells induced by SmoSD123 was comparable with that in the A/P boundary cells ( Figure 4G ) . Given that Hh is absent in A-compartment cells away from A/P boundary and that activated Smo-induced Hh signaling is independent of Ptc expression , this system circumvents the compensatory feedback regulation between Ptc expression and Hh signaling in the A-compartment cells . We then probed whether Smurf degrades Ptc in the cells with ectopic Hh signaling activity . As shown in Figure 4E , 4F , 4H , and 4I , overexpression of Smurf but not SmurfC1029A by the MS1096-gal4 significantly downregulated the levels of Ptc protein ectopically induced by the expression of SmoSD12 or SmoSD123 in A compartment cells ( Figure 4D and 4H ) . Moreover , we found that overexpression of Smurf by the MS1096-gal4 did not affect Hh signaling activity induced by the activated Smo , as indicated by the fact that the expression levels of ptc-lacZ or dpp-lacZ were not evidently changed in Figure 4K–4N , although overexpression of Smurf slightly reduced SmoSD123 levels ( unpublished data ) . Thus , activation of Smo is sufficient to promote Ptc protein degradation in vivo . To test whether the Smurf-mediated Ptc degradation is attributed to activation of Smo downstream signaling , we performed genetic assays by overexpression of constitutively activated form of Ci ( Ci103 ) in the wing disc [8] , [11] . As shown in Figure S4A–S4D , overexpression of Smurf had no detectable effect on Ptc protein stability induced by Ci103 expression in wing discs . Thus , our results suggest that Smurf-mediated Ptc protein degradation largely depends on activation of Smo itself but not its downstream signaling . To gain biochemical evidence , we then performed the ubiquitination assay in S2 cells . As shown in Figure 5A and 5B , overexpression of SmoSD123 significantly increased the Ptc ubiquitination by Smurf; however , knockdown of Smurf markedly reduced Ptc ubiquitination promoted by overexpression of SmoSD123 . Thus , our findings further confirm that activated Smo promotes Smurf-mediated Ptc ubiquitination and degradation . Having seen that Smo promotes Smurf-mediated Ptc ubiquitination , we then sought to investigate whether activated Smo and Smurf act in a common pathway to regulate Ptc ubiquitination by forming a complex . As shown in co-immunoprecipitation assays ( Figures 5C , 5D , and S4E ) , Smurf could be associated with Smo through the C-tail of Smo , emphasizing that activated Smo promotes Smurf-mediated Ptc ubiquitination in a Smo activation-dependent manner . Since Smurf and Smo form a complex , we then asked whether Smurf regulates ubiquitination and stability of Smo in S2 cells and wing discs . As shown in Figure S4F and S4G , knockdown of smurf did not affect Smo ubiquitination and stability in S2 cells , and knockdown of smurf in wing disc clone cells caused no apparent change of endogenous Smo protein expression , when compared to the neighbor wild-type cells ( Figure S4H–S4H″ and S4I–S4I′″ ) . Thus , our results suggest that Smurf regulates Hh signaling primarily through targeting Ptc , but not Smo , for ubiquitination . To further understand the molecular basis of how Ptc is ubiquitinated through its C-tail , we sought to search for the specific site ( s ) in Ptc C-tail that respond to the Smurf E3 ligase . We thus generated a series of mutant forms of Ptc , in which the K sites in the Ptc C-tail were mutated to R individually or in combination . As shown in ubiquitination assays , both PtcK1261R and PtcK1211R , K1261R , K1269R mutants were evidently resistant to being ubiquitinated by Smurf ( Figures 5E and S5A ) , suggesting the K1261 site is specifically regulated by Smurf in S2 cell cultures . To determine the biological function of the K1261 site of Ptc , we generated a ubiquitin-resistant form of the Ptc transgene line , P{UAS-PtcK1261R} , in which the K1261 site was mutated to R . As shown in Figure 5F–5I , expression of PtcK1261R by the en-gal4 driver in the posterior compartment of wing disc caused extreme reductions in the intervein region between L3 and L4 veins , whereas expression of PtcWT led to a less severe wing phenotype . In support of this idea that the greater severity of the PtcK1261R phenotypes was attributed to the resistance of PtcK1261R to ubiquitination by Smurf E3 ligase , expression of PtcWT by the en-gal4 driver in the smurf mutants caused much more severe intervein phenotype than that in wild type ( Figures 3J–3L and S5A ) , demonstrating the biological importance of the K1261 site of Ptc in the regulation of Hh signaling . The previous study has shown that residues ( T1260 , T1263 , T1265 ) are vital for Ptc trimerization . Interestingly , the K1261 we identified is very close to these residues , and disruption of these residues resulted in a similar phenotype to that in the wing expressing PtcK1261R [13] , raising a possibility that K1261 might contribute to the Ptc trimerization . To test this , we used the native gel and performed Western blot assays , and found that the K1261R mutation had no effect on the Ptc trimerization ( Figure S5B ) . Taken together , our results strongly argue that Smurf regulates Ptc ubiquitination through its K1261 site in vivo . It has been proposed that the ratio of Hh ligand bound to unbound Ptc protein is important in determining Hh signaling activity in Drosophila wing discs [36] . However , the molecular mechanism of how Ptc degradation is regulated remains elusive . In this study , we found that Hh ligand-independent Smo activation is sufficient to promote Smurf-mediated Ptc ubiquitination and degradation . Interestingly , as shown in Figure 4D and 4H , we noted that downregulation of Ptc by Smurf in A compartment cells appears much more effective than that in the A/P boundary . Given that the Hh molecules are highly presented in A/P boundary but absent in A-compartment cells away from the A/P boundary , our findings thus raised a possibility that Smurf might have a preference for targeting unbound Ptc in wing discs . To test this , we performed ubiquitination experiments in S2 cells . S2 cells were co-transfected with Ptc , Smurf , and SmoSD123 , and then treated with Hh-conditioned or control medium . As shown in Figure 5J , the levels of Ptc ubiquitination from cells with Hh treatment were apparently lower than that from control cells , suggesting that SmoSD123 promotes Smurf-mediated Ptc ubiquitination more efficiently in the absence of Hh . To further verify this data in vivo , we co-expressed Hh with or without Smurf in wing discs using MS1096-gal4 driver . As shown in Figure 4O–4P′ , overexpression of Hh resulted in high levels of Ptc expression in A-compartment cells , which were comparable with the A/P boundary cells . However , we found that , in contrast to the fact that Smurf overexpression significantly downregulated SmoSD-induced Ptc expression in A-compartment cells , compared with that in A/P boundary cells , overexpression of Smurf had no apparent effect on Hh ligand-induced Ptc expression in A-compartment cells , compared with that in A/P boundary cells . Thus , our data strongly argue that Smurf has a preference for targeting ligand-unbound Ptc . How does the preferential targeting of ligand-unbound Ptc by Smurf contribute to Hh signaling transduction ? The previous study proposed an elegant model by which the ratio of ligand bound to unbound Ptc protein is important in determining Hh signaling activity in Drosophila wing discs [36] . Our experimental data reveal that Smurf could control the ratio of ligand bound to unbound Ptc protein through preferentially targeting ligand-unbound Ptc in wing disc , thereby precisely determining Hh signaling . Additionally , this regulation apparently depends on Hh signaling , particularly promoted by activation of Smo . Given the fact that the high level of ptc expression is directly induced by Hh but negatively regulates Hh signaling , collectively , our findings imply that a bidirectional regulation mechanism is involved in Ptc and Smurf to control and balance the signaling activity when the cells respond to external Hh molecules . Our findings also raise an intriguing issue as to how the bidirectional regulation contributes to the dynamics of Hh signaling transduction . Hh signaling dynamics has been studied previously [37] , however , how Ptc ( or the ratio of ligand bound to unbound Ptc ) is precisely regulated in the model remains elusive . In order to better understand the dynamic basis of how Smurf-mediated Ptc turnover contributes to the reliability of the Hh signaling transduction , we developed a mathematical model below based on the current simplified Hh signaling pathway network established in this study ( Figure 6A ) . As pointed out by Nahmad and Stathopoulos [37] , the synthesis rate of Smo/Ci is regulated by the ratio [Hh_Ptc]/[Ptc] rather than [Ptc] directly since Hh-dependent gene expression depends on the ratio of liganded to unliganded Ptc . The concentrations of the Hh , Ptc , and Hh_Ptc complex , activated Smo/Ci , and activated Smurf are denoted in the equations below by [Hh] , [Ptc] , [Hh_Ptc] , [activated Smo/Ci] , and [activated Smurf] , respectively , and , for simplicity , the synthesis rate of Hh is assumed to be a constant CHh [36] , [37] . However , for convenience , let x1 = [Hh] , x2 = [Ptc] , x3 = [Hh_Ptc] , x4 = [activated Smo/Ci] , and x5 = [activated Smurf] . Thus , according to Figure 6A , the rate equations of [Hh] , [Ptc] , [Hh_Ptc] , [activated Smo/Ci] , and [activated Smurf] can be given by ( 1 ) where the terms , , and are the Hill-type functions regarding the synthesis rates of Ptc , activated Smo/Ci , and activated Smurf , respectively , where and are the fundamental synthesis rates of Ptc and Smurf; is the formation rates of Hh_Ptc complex; and the terms is the Hill-type functions regarding the degradation rates of Ptc , where is the fundamental degradation rate of Ptc; and , , , and are the degradation rates of Hh , Hh_Ptc complex , activated Smo/Ci , and activated Smurf , respectively , which are assumed to be constants . Basically , the stability analysis of Equation 1 ( see Table S2 ) shows that the dynamics has a unique equilibrium and is globally asymptotically stable . We also noticed that although the change in the level of activated Smurf cannot change the dynamical properties of the system ( i . e . , no multi-equilibrium and periodic or chaotic solutions can exist ) , the levels of Ptc ( or the ratio [Hh−Ptc]/[Ptc] ) and activity of Smo/Ci sensitively depend on the change in the levels of activated Smurf . That is , the increase of activated Smurf will result in the decrease of Ptc ( or the increase of the ratio [Hh−Ptc]/[Ptc] ) and increase in the activity of activated Smo/Ci , and conversely . Numerical solutions of Equation 1 are shown in Figure 6B–6D , in which the effects of different activated Smurf levels on Ptc and activated Smo/Ci are considered . It appears that the theoretical and numerical analysis of Equation 1 matches our experimental data . The above dynamic analysis reveals how the bidirectional regulation of activated Smo/Ci on Ptc , namely , activated Smo/Ci directly promotes the synthesis of Ptc through transcriptional regulation , but or indirectly promotes the degradation of Ptc through promoting the activity of Smurf . This bidirectional control appears to affect the dynamical properties of the Hh signaling pathway , and this mathematic model reveals that Ptc ( or the ratio of [Hh−Ptc]/[Ptc] ) not only responds to the Hh signaling but also depends more sensitively on the bidirectional regulation of activated Smo/Ci . Thus , theoretically , the bidirectional regulation occurring in Ptc and Smurf should be able to ensure the reliability of Hh signaling transduction when cells respond to an external Hh signal . The role of Smurf in targeting Ptc protein indicates the biological importance of its potential localization on plasma membrane . On the basis of our mathematic model , we predicted that an increase of Smurf concentration ( or activity ) on plasma membrane would accelerate the turnover of Ptc protein , and subsequently reduce the level of Ptc . To further test our model , we generated transgenic lines , P{uas-Src:Flag:smurf} , in which a membrane-tethered form of Smurf was under the control of the UAS promoter . Expression of the membrane-tethered form of Smurf ( Src:Flag:smurf ) appeared to accumulate Smurf on plasma membrane , compared to the wild type of Smurf ( Figure S6A and S6B–S6B′ ) . As shown in Figure 6E–6H″ , while expression of wild-type Smurf by the ap-gal4 driver in the dorsal region of wing discs caused a weak reduction of Ptc protein ( Figure 6E and 6F–6F″ ) , expression of the membrane-tethered form of Smurf almost abolished Ptc protein expression ( Figure 6H–6H″ ) and increased dpp-lacZ expression by MS1096-gal4 ( Figure S6C , S6D–6D″ , and S6E–S6E″ ) , suggesting that the membrane anchorage of Smurf appears to increase its activity in controlling Hh signaling , and further supporting our mathematic model that Smurf plays an important role in targeting Ptc protein degradation , thereby balancing Hh signaling . To this end , we have identified a novel role of Smurf in regulating Hh signaling in Drosophila . Given that both Smurf and the Hh signal pathway are evolutionarily conserved from fly to vertebrates , we then turned our attention to test whether Smurf proteins are also involved in regulating Hh signaling in vertebrates using zebrafish embryos as an assay system . The zebrafish genome harbors two smurf genes , smurf1 and smurf2 , which encode zebrafish Smurf1 and Smurf2 proteins , respectively . As shown in an in situ hybridization assay , both smurf1 and smurf2 mRNAs were ubiquitously present in all stages during embryogenesis ( Figure S7A ) . To investigate the biological functions of Smurfs in zebrafish embryonic development , we performed morpholino ( MO ) -based knockdown experiments to inactivate smurf1 or smurf2 in zebrafish embryos . As shown in Figure 7A , both smurf1 and smurf2 morphants displayed curved tails and U-shaped somite phenotypes , compared with control MO injected embryos . Injection of another splicing blocking morpholino , smurf1 s1MO , which blocks the intron 3 splicing from the primary smurf1 transcript , thereby causing intron 3 retention , gave similar somitic phenotypes ( Figure S7B ) ( unpublished data ) . Whole-mount F59 immunofluorescence staining analysis clearly showed that somitic muscles in the mid-trunk region were disrupted in the smurf1 or smurf2 morphants ( Figure 7A ) . These Hh-related defects were typically observed in the smo mutant or in embryos treated with cyclopamine ( Figure 7A and 7D ) , which specifically blocks Hh signaling in zebrafish [38] , indicating a potential role of Smurf proteins in regulating Hh signaling that controls the somite development ( i . e . , specification of distinct muscle cell types ) . To validate whether these phenotypes resulting from inactivated smurfs are attributed to disruption of Hh signaling , we first examined the expression levels of several Hh responsive target genes , such as hhip , fkd4 , and nkx2 . 2b in smurf morphants . Consistent with the phenotypic analysis , the expression levels of these Hh responsive target genes were significantly reduced in either smurf1 or smurf2 morphants when compared with the control morphants ( Figure 7B ) , suggesting that Hh signaling activity was indeed downregulated when either smurf1 or smurf2 was inactivated . Expression of Engrailed , a direct target of Hh signaling , was also reduced or absent in smurf1 or smurf2 morphants ( Figure S7C ) . We then elevated Hh signaling in smurf morphants by injection of smo mRNA . Interestingly , we found that co-injection of smo mRNA appeared to suppress the phenotypes resulting from either smurf1 or smurf2 MOs ( Figure 7C ) . To further confirm the role of Smurfs in regulating Hh signaling in zebrafish , we performed a dosage-modified genetic interaction between Smurfs and Hh by employing smo mutants . As shown in Figure 7D , both embryos injected with low doses of smurf1 or smurf2 MOs ( 2 ng/embryo ) and embryos carrying heterozygous smo mutant exhibited no apparent phenotype during embryo development . However , smo heterozygous embryos showed strong Hh-related phenotypes when the embryos were injected with the same dose of smurf1 or smurf2 MOs . To test whether Smurfs have a role in the degradation of Ptc1 in zebrafish , we made a zebrafish Ptc1c and GFP fusion mRNA ( Ptc1c:GFP ) . As shown in Figure S4D , co-injection with either Smurf1 or Smurf2 mRNA resulted in much weaker fluorescence , compared with Ptc1c:GFP mRNA injection alone , suggesting that Smurfs might play a conserved role in degrading Ptc protein . Similarly , knockdown of Ptc1 could efficiently rescue the somite defects in smurf morphants ( Figure 7C ) , suggesting that Smurfs negatively regulates Ptc1 . Collectively , our results support the notion that zebrafish Smurf1 and Smurf2 also function as positive modulators in Hh signaling via targeting the Ptc1 protein for its degradation , thereby controlling late somite development ( i . e . , distinct muscle cell types ) during zebrafish embryogenesis .
We have previously shown that Smurf functions in concert with Fu to degrade the BMP Tkv receptor in differentiating cystoblasts ( CBs ) , thereby generating a steep BMP responsive gradient between germline stem cell ( GSC ) and CBs , and determining the GSC fate in Drosophila ovary [24] . Thus , our prior findings reveal an important in vivo role of Smurf E3 ligase in targeting to the membrane receptor for proper BMP signal transduction . In this study , we provide both biochemical and genetic evidence to show a novel role of Smurf in mediating Hh signal transduction by targeting to Ptc for its degradation in wing discs . Additionally , we find that , like their homologue in Drosophila , the zebrafish Smurf proteins are also involved in Hh signaling via targeting to the Ptc1 protein to control late somitogenesis during zebrafish embryogenesis . Thus , our study reveals a novel and evolutionarily conserved role of Smurf proteins in controlling Hh signaling . Smurf is a member of the C2-WW-HECT E3 family of proteins in Drosophila . In addition to Smurf , Drosophila also has two other members of the C2-WW-HECT family of proteins: Suppressor of deltex ( Su ( dx ) ) and Nedd4 protein . Interestingly , we found that all three of these E3 ligases could individually form a complex with the Ptc protein through the C-tail of Ptc ( Figures S2H , S2I , and S8A ) . However , we observed that Smurf is the only member of the C2-WW-HECT family in Drosophila that plays a positive role in the regulation of Hh signaling through targeting Ptc in the wing disc system , since alteration of Nedd4 expression did not affect Hh signaling activity and the Su ( dx ) has a negative role in regulating the Hh signaling pathway in the wing disc ( Figure S8B–S8K ) . Notably , we also found that zebrafish Nedd4 has no role in Hh signaling during early embryonic development in zebrafish ( Figure S9 ) . Thus , our findings suggest a specific biological role for Smurf in targeting Ptc to balance Hh signaling in the tested systems in this study . Morphogen-mediated signaling gradients have been proposed to regulate differential gene expression in a concentration-dependent manner [39] , [40] . However , the fundamental question of how signal-receiving cells perceive and precisely interpret the environmental cues provided by morphogens has been not well-understood [39] , [41] . The Hh signaling pathway has an unusual signal reception system including the receptor Ptc and the signal transducer Smo . While the activation of Smo and its downstream events have been extensively studied , how the Hh receptor Ptc is regulated and whether activation of Smo also contributes to regulating the expression and/or activity of the receptor Ptc remain unknown . Given that Ptc has a negative role in the regulation of Hh signaling and that ptc itself is a responsive target of Hh signaling , the expression and/or activity of Ptc must be tightly controlled to ensure proper Hh signal transduction . In Drosophila wing discs , previous studies have proposed that Hh ligand binding and self-induced degradation of Ptc might contribute to the Ptc internalization and the regulation of Ptc expression [13] , [30] . In this study , we identify Smurf as a Ptc-interacting partner , and provide evidence that Smurf functions as a negative regulator to control Ptc ubiquitination and degradation . Importantly , we observe that Smo activation promotes the E3 ligase activity of Smurf to ubiquitinate Ptc protein in both wing discs and S2 cell cultures . In particular , we find that Smurf activity largely depends on the activation of Smo to regulate Ptc turnover . Previous studies have shown that plasma membrane accumulation of Smo is essential for Hh signaling transduction [42] . In this study , we show that Smurf was able to form a complex with Smo to regulate Ptc ubiquitination in a manner that depends on the specific Ptc-K1261 site . Thus , in addition to establishing functions for Hh signal transduction by controlling Ci processing , our findings identify a novel role for the Smo protein in targeting the Ptc receptor on plasma membrane for Ptc degradation . Previous studies have proposed an elegant model by which the ratio of ligand bound to unbound Ptc protein is important in determining Hh signaling activity in Drosophila wing discs [36] . However , the molecular mechanism underlying the regulation of the ratio of ligand bound to unbound Ptc remains unknown . Given that the ptc itself is a transcriptional target of the Hh signal , and that the excessive unbound Ptc induced by Hh signaling appears to potentially increase fluctuating levels of cellular Hh signaling , the question becomes how excessive Ptc is restricted within a safe range to control the appropriate ratio of ligand bound to unbound Ptc . Notably , our findings reveal that Smurf has a preference for degrading unbound Ptc ( Figure 4C–4J ) ; thus , integration of Smurf-mediated Ptc degradation with Hh signaling permits us to propose a bidirectional regulatory model that maintains the robustness of cellular Hh responsive activity through controlling a precise ratio of ligand bound to unbound Ptc . In this model , in the presence of Hh molecules , Ptc is highly expressed in signal-receiving cells in response to Hh signaling activation , but negatively regulates Hh signaling through a feedback mechanism . Meanwhile , Smurf also responds to activation of Smo and forms a complex with Smo , thus accumulating on plasma membrane to target Ptc for degrading the excessive unbound Ptc protein , thereby controlling the ratio of bound to unbound Ptc protein and subsequently sustaining the precise level of Hh signaling that ensures distinct target gene activations during development ( Figure 6A ) . The mathematic modeling analysis reveals that Ptc not only responds to the Hh signaling but also depends more sensitively on the bidirectional regulation of activated Smo/Ci , further supporting the importance of the role of Smurf in controlling the reliability of Hh signal transduction . Hh signaling is the fundamental signaling pathway , and is critical for development and stem cell function in adult tissues [2] , [3] . Aberrant regulation of Hh signaling has been implicated in many human diseases , particularly cancers . Clinical examples include basal cell carcinoma ( BCC ) , medulloblastoma , and other human tumors that are usually associated with mutations of Hh signaling components including PTC1 and SMO [43]–[45] . Interestingly , a very recent study has provided a link between the function of Smurf2 and tumorigenesis , since loss of Smurf2 leads to various types of tumors in old age [46] . Given that Smurf proteins are evolutionarily conserved from Drosophila to vertebrates and that the molecular basis of Ptc protein degradation is likely conserved in evolution [13] , [47] , a target of future investigation would be , therefore , whether Smurf is involved in regulating Hh signaling-related tumorigenesis .
Fly stocks used in this study were maintained under standard culture conditions . The w1118 strain was used as the host for all P element-mediated transformations . smurf15c , MS1096-gal4 , ap-gal4 , act>CD2>gal4 , UAS-GFP , dpp–lacZ ( dppZ ) , ptc–lacZ ( ptcZ ) , UAS-Hh , UAS-SmoSD12 , UAS-SmoSD123 , and UAS-HA:Ci103 have been described previously [11] , [20] , [48] , [49] . Su ( dx ) RNAi , ( #N4244R-1 ) , and Nedd4 RNAi ( #V13121 ) were obtained from NIG and VDRC . Fly Strains including P{uast-Flag-smurf} , P{uast-Flag-smurfC1029A} , P{uast-SRC-Flag-smurf} , P{uast-Flag-su ( dx ) } , P{uast-Flag-nedd4} , P{hs-ptc:gfp} , P{uast-ptc} , P{uast-ptcK1261R} , and P{uasp-shmiR-smurf} [35] were generated in this study . The flies y w P{hsFLP}122/y w P{hsFLP}122 or Y; P{FRT ( w[hs] ) }G13 2xP{Ubi-GFP . nls}/P{FRT ( w[hs] ) }G13 Smurf15c; dpp-lacZ were used for generating smurf mosaic clones in wing discs . Larvae were heat shocked 2–3 days after birth at 37°C for 30 min to induce flippase activity , and then cultured at 25°C for another 2–3 days for wing disc dissection . Standard protocols for immunofluorescence staining of imaginal discs were used [50] . The following antibodies were used in this study , including rabbit anti-GFP ( 1∶5 , 000 , Invitrogen ) ; mouse anti-β Gal ( 1∶1 , 000 Promega ) ; rabbit anti-β Gal ( 1∶2 , 000 , Cappel ) ; mouse anti-Flag ( 1∶500 , Sigma ) ; rabbit anti-Flag ( 1∶500 , Sigma ) ; rat anti-Ci ( 1∶200 , DSHB ) ; mouse anti-Ptc ( 1∶200 , DSHB ) ; and mouse anti-Smo ( 1∶500 , DSHB ) . The following secondary antibodies were used at a 1∶2 , 000 dilution: goat anti-mouse Alexa555 , goat anti-rabbit Alexa488 , and goat anti-rat Alexa647 ( Molecular Probes ) . CHX treatments of wing discs were performed in M3 medium containing 2% fetal bovine serum ( Hyclone ) , 2 . 5% fly extract , and 0 . 5 mg/ml insulin ( Sigma ) for 2 h at 25°C . GST and GST:Flag:PtcCTD fusion proteins were expressed in Escherichia coli , and purified with glutathione agarose beads ( GE Healthcare ) by the batch purification method . His:Smurf protein was expressed in E . coli , and purified with Ni Sepharose ( GE Healthcare ) , and GST fusion protein loaded beads were incubated with 200 ng of purified His:Smurf in GST pull down buffer ( 50 mM Tris-Cl [pH 8 . 0] , 200 mM NaCl , 1 mM EDTA , 1% NP-40 , 2% BSA , 10 mM DTT , 1 mM PMSF ) at 4°C for 1 h . The beads were washed three times with lysis buffer . Western blot analysis was then performed to detect the direct interaction between Smurf and PtcCTD . S2 cells were cultured in Schneider's Drosophila medium ( Sigma ) at 27°C . Transfection was performed using the calcium phosphate transfection method according to our previous method [51] . In some experiments , transfected S2 cells were treated with MG132 ( 50 µM ) or NH4Cl ( 50 mM ) for 4 h before harvesting when necessary . Cells were lysed in lysis buffer ( 150 mM NaCl , 50 mM Tris-HCl [at pH 7 . 4] , 10% glycerol , 1% Triton X-100 ) with a protease-inhibitor cocktail ( Invitrogen ) , before binding with affinity gel; the concentration of NaCl in cell lysate was adjusted to 500 mM . Anti-Flag M2 affinity gel ( Sigma ) , anti-Myc affinity gel ( Abmart ) , and protein A/G gel were used for indicated immunoprecipitation experiments . After affinity pull-down , gels were extensively washed in lysis buffer contain 500 mM NaCl for 15 min each time , and for a total of three times . Western blot analysis of full-length Ptc protein was carried out by SDS-PAGE electrophoresis without boiling; the other proteins were analyzed by using standard protocol . The following antibodies were used for Western blotting: rabbit and mouse anti-Myc ( 1∶3 , 000 , MBL ) ; rabbit anti-HA ( 1∶3 , 000 , MBL ) ; rabbit and mouse anti-Flag ( 1∶3 , 000 , Sigma ) ; mouse anti-Ptc ( 1∶1 , 000 , DSHB ) ; mouse anti-Smurf ( 1∶10 , 000 ) [24]; and rabbit anti-α-tubulin ( 1∶5 , 000 , Abcam ) . In vivo ubiquitination assays were performed according to our previous study [31] . Briefly , S2 cells were transfected with indicated DNA constructs . At 48 h post-transfection , cells were treated with MG132 ( at a final concentration of 50 µM ) and/or NH4Cl ( at a final concentration of 50 mM ) for 4 h . Cells were lysed in lysis buffer ( 50 mM Tris-HCl [at pH 7 . 4] , 150 mM NaCl , 10% glycerol , 1% Triton X-100 , 0 . 1% SDS , and 10 mM NEM ) with a protease-inhibitor cocktail ( Invitrogen ) . Supernatant was collected by centrifugation at 13 , 000 rpm for 20 min at 4°C . Before immunoprecipitation with anti-Flag or anti-Myc beads , NaCl concentration in the lysates was adjusted to 500 mM . After pull-down for 4 h , the beads were then extensively washed with lysis buffer containing 0 . 1% SDS and 500 mM NaCl three times for a total 1 h . Samples were then subjected to Western blot analysis . In addition to the method described here , we also used two other methods to measure Ptc ubiquitination levels ( described in Text S1 ) ; we noted that all three methods generated consistent results .
|
Hedgehog ( Hh ) signaling is a pathway renowned for its roles in controlling embryonic development and tumorigenesis . Signaling via this pathway proceeds when Hh ligands bind to the receptor Patched ( Ptc ) , thereby preventing Ptc from inhibiting the signal transducer , Smoothened ( Smo ) , and thus allowing Smo to accumulate on the cell surface where it becomes activated and promotes downstream signal transduction . In the absence of Hh ligands , Ptc inhibits Smo and is a key negative regulator of Hh signaling . In this study , we investigate how protein turnover of Ptc is controlled to ensure tight regulation of Hh signaling . Using Drosophila as a model system , we provide biochemical and genetic evidence to show that the E3 ligase , Smurf , directly controls Ptc protein turnover in developing wing discs . Moreover , we found that Smurf mediates Ptc degradation in a manner that depends on Smo signaling activity: activated Smo forms a complex with Smurf to preferentially promote degradation of the ligand-unbound Ptc receptor . Using mathematic modeling we reveal that the control of Smo activation by the opposing activities of Smurf and Ptc , is important for cells receiving the Hh signal to precisely interpret and relay external signals . We show that this control mechanism is also active in vertebrates with evidence that zebrafish Smurf proteins target Ptc1 protein for degradation to control late somitogenesis during zebrafish embryogenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Activation of Smurf E3 Ligase Promoted by Smoothened Regulates Hedgehog Signaling through Targeting Patched Turnover
|
A comprehensive analysis was done to evaluate the potential use of anti-parasitic macrocyclic lactones ( including avermectins and milbemycins ) for Buruli ulcer ( BU ) therapy . A panel containing nearly all macrocyclic lactones used in human or in veterinary medicine was analyzed for activity in vitro against clinical isolates of Mycobacterium ulcerans . Milbemycin oxime and selamectin were the most active drugs against M . ulcerans with MIC values from 2 to 8 μg/mL and 2 to 4 μg/mL , respectively . In contrast , ivermectin and moxidectin , which are both in clinical use , showed no significant activity ( MIC> 32 μg/mL ) . Time-kill kinetic assays showed bactericidal activity of selamectin and in vitro pharmacodynamic studies demonstrated exposure-dependent activity . These data together with analyses of published pharmacokinetic information strongly suggest that selamectin is the most promising macrocyclic lactone for BU treatment .
Buruli ulcer ( BU ) , caused by Mycobacterium ulcerans , is a chronic debilitating disease of the skin and soft tissue . Although mortality is low , permanent disfigurement and disability is high [1] . BU is mainly found in Africa , South America and the Western Pacific regions and is often linked to poverty . If detected early , BU can be cured in most cases with the standard treatment , a combination of rifampicin and the injectable antibiotic streptomycin [2] , without further adjunct surgical treatment required . However , new treatment regimens are needed to reduce the long median time to healing , treatment-related side effects , and the requirement for on-site health care workers to administer injections [3] . Furthermore , an alternative drug treatment regimen would be required in the event that rifampicin resistant M . ulcerans strains would emerge in the clinic [4] . Traditionally , the discovery of new antimicrobial drugs has focused on designing and screening for new compounds having novel targets , an approach that is costly in time and capital ( up to ~$800M and 15–20 years ) [5] . This is not a viable option for BU , since most large pharmaceutical and biotech companies are primarily interested in blockbuster , broad spectrum antibacterial drugs [6] rather than treatments for neglected tropical diseases . A faster and cheaper alternative to finding new BU treatments is drug repositioning , i . e . using approved drugs for alternative clinical indications [7] . These drugs with known pharmacokinetic and safety profiles could be more rapidly evaluated in clinical trials [8] . Such an approach would also allow for an easier drug introduction , since manufacturing and distribution infrastructures are already available . In the course of screening clinically approved drugs to find new drug combinations for tuberculosis ( TB ) therapy , we discovered anti-mycobacterial activities of the avermectins , a class of macrocyclic lactones [9] . Following up these findings , the in vitro activities of two clinically approved macrocyclic lactones ( ivermectin and moxidectin ) against M . ulcerans were recently reported [10] . The avermectins are a family of macrocyclic lactone derivatives with potent anthelmintic properties , produced by the soil actinomycete Streptomyces avermitilis . Since avermectins are inactive against all other bacterial species tested [9] , oral administration would not affect healthful intestinal microbiome balances . We undertook a comprehensive approach to evaluate additional macrocyclic lactones used in veterinary medicine . Based on our in vitro measurements of their activities and a literature review of their pharmacokinetic ( PK ) properties , we provide strong indications that selamectin ( used in veterinary medicine ) , and not ivermectin ( used in human medicine ) , is the avermectin with the highest potential for clinical efficacy to treat BU .
M . marinum isolates ( 1704 and 1705; kindly provided by Dr . Julian Davies , University of British Columbia ) were routinely propagated at 30°C in Middlebrook 7H9 broth ( Difco ) supplemented with 10% Middlebrook albumin-dextrose-catalase ( ADC ) ( Difco ) , 0 . 2% glycerol and 0 . 05% ( vol/vol ) Tyloxapol or on Middlebrook 7H10 agar plates ( Difco ) supplemented with 10% ( vol/vol ) oleic acid-albumin-dextrose-catalase ( OADC ) ( Difco ) . M . ulcerans strains S1012 , S1013 and S1047 ( isolated in 2010 and 2011 from Cameroonian BU patients ) were routinely grown in BacT/Alert culture bottles using enrichment medium ( bioMérieux ) or on Middlebrook 7H10 agar plates ( Difco ) supplemented with 10% ( vol/vol ) OADC . Macrocyclic lactones were purchased from the following providers: abamectin and doramectin ( Sigma ) , emamectin and eprinomectin ( LKT Labs ) , ivermectin ( Alpha Diagnostic ) , milbemycin oxime ( US Pharmacopeia ) , moxidectin and selamectin ( European Pharmacopoeia ) . Minimal Inhibitory Concentrations ( MIC ) were determined in 7H9 broth supplemented with 0 . 2% glycerol and 10% ADC ( M . marinum ) or 10% OADC ( M . ulcerans ) using two-fold serial dilutions of compounds in triplicate in polystyrene 96-well plates . MTT [3- ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyl tetrazolium bromide] and resazurin were used as the bacterial growth indicators [11] for M . marinum and M . ulcerans , respectively . For M . marinum , cultures were sampled ( 100 μL ) at a cell density of 105 cells/mL and incubated in the presence of the drug for 3 days before addition of 25 μL of MTT ( 5 mg/mL ) . After further overnight incubation , 100 μL of 10% Sodium Lauryl Sulfate ( SLS ) were added to solubilize the formazan precipitate that indicates bacterial growth and the optical density at 580 nm ( OD580 ) was then measured . In the case of M . ulcerans , 100 μL culture samples ( OD600 = 0 . 04 ) were incubated in the presence of the drug at 30°C for 8 days before addition of 20 μL of a resazurin solution ( 0 . 125 mg/mL ) , followed by overnight incubation at 37°C . Compound activity was determine by fluorescence measurements ( λ = 540/588 nm ) . The lowest concentration of drug that inhibited 90% of the MTT or resazurin color conversion ( IC90 ) was used to define MIC values . 96-well polystyrene plates containing 200 μL per well of 7H9 broth supplemented with 10% OADC were inoculated in duplicate with M . ulcerans S1013 to a final OD600 = 0 . 04 ( ca . 105 cells/mL ) . Cultures were grown at 30°C in the presence of 0 . 5 , 2 , 4 , 8 , 16 and 32 μg/mL of selamectin ( 0 . 5 , 1 , 2 , 4 , 8 and 16 fold the selamectin MIC value , respectively ) for 0 , 3 , 7 , 14 and 21 days . At every time point , 100 μL of undiluted and ten-fold serial dilutions were plated on 7H10 agar . Colony-forming units for all plates were determined after 8 weeks of incubation at 30°C .
Eight commercially available macrocyclic lactones used in human and veterinary medicine were tested in vitro against M . ulcerans and M . marinum . Milbemycin oxime and selamectin were the most potent drugs against the M . ulcerans isolates ( MIC in 2–4 μg/mL range ) . Emamectin and moxidectin had intermediate potency ( MIC = ca . 32 μg/mL ) . While it was not possible to determine minimal inhibitory concentrations for ivermectin ( IC90 >64 μg/mL ) , some inhibitory activity was observed in dose response studies . In contrast , most of the macrocyclic lactones showed activity against M . marinum , a faster growing phylogenetic progenitor of M . ulcerans , with milbemycin oxime being the most potent ( Fig 1 and Table 1 ) . The PK properties of selamectin ( described below ) , together with its high in vitro activity against M . ulcerans , strongly indicated it as the most suitable avermectin for further evaluation as a potential new anti-BU treatment . To further characterize this potential new application , the in vitro pharmacodynamic ( PD ) parameters of selamectin were evaluated using kill-kinetic assays ( Fig 2 ) . In vitro kill-kinetic curves for selamectin were obtained by plotting the number of CFU at every time point for every concentration of the drug ( Fig 2A ) . These experiments confirmed the MIC and dose-response data determined by reporters of metabolic activity ( resazurin and the MTT; Table 1 and Fig 1 ) and showed a sharp threshold of bactericidal activity above the MIC ( 2 μg/mL ) . We also used an alternative method to visualize kill kinetics: each selamectin concentration was multiplied by the time of exposure ( CSEL x Tdays ) and then divided by the MIC of selamectin to give the in vitro area under the concentration-time curve ( AUC/MIC ratio ) , a standard measure of drug exposure ( Fig 2B ) . These analyses showed that just seven days of exposure were needed to observe the bactericidal activity of selamectin . The AUC/ MIC needed to achieve a bactericidal effect ( 4-log10 CFU/ml reduction , 99 . 99% killing ) required AUC/MIC ratios between 10 and 15 . These ratios were comparable to those previously observed for M . tuberculosis [9] . In summary , these studies showed that the activity of selamectin against M . ulcerans is exposure-dependent; if a certain concentration is achieved , a bactericidal effect is observed by increasing the time of exposure but not by increasing dose concentrations . This information could have important implications when designing pre-clinical and clinical studies .
The family of anthelmintic macrocyclic lactone drugs is one cornerstone of modern parasite control with annual world sales of US $850 million , indicating a well-established production and distribution pipeline . These drugs share a poly-cyclic lactone chemical moiety and can be divided in two sub-families: avermectins and milbemycins [12] . Because members of this family of natural products have complex structures and specificity for parasites , only a few have been commercialized , mostly for veterinary medicine [13] . Ivermectin is used to treat the human parasitic diseases onchocerciasis and lymphatic filariasis [14] . Moxidectin was also recently evaluated for these indications in clinical trials [15] . The potential use of ivermectin for TB treatment is questionable due to its neurotoxicity at high doses and the low exposure levels achieved using clinically approved doses [16] . We analyzed available literature to compare the pharmacological properties of clinically approved drugs ( ivermectin , moxidectin ) to those with best in vitro activities against M . ulcerans ( milbemycin oxime and selamectin ) ( Table 2 ) . By integrating this information with in vitro data , we propose selamectin as the anthelmintic macrocyclic lactone with the highest potential for anti-BU therapy . In invertebrate nematodes , avermectins specifically bind to glutamate-gated chloride channels present in nerve and muscle cells , causing paralysis and reduced ability to reproduce . In general , macrocyclic lactones have a high margin of safety in mammals because P-glycoproteins ( P-gp ) or other types of efflux pumps , highly expressed at the blood–brain barrier , efficiently restrict their penetration into the central nervous system . In fact , dogs lacking the MDR1 efflux pump , such as collies , have much less tolerance for treatment with an array of avermectin compounds [29] . In contrast , milbemycin oxime , selamectin , and moxidectin can be safely administered at therapeutic doses to dogs having a homozygous MDR1 mutation without any signs of toxicosis [12 , 28] . PK and toxicological profiles of the clinically used macrocyclic lactones ( ivermectin and moxidectin ) have been studied extensively . Using standard dosages for onchocerciasis treatment in humans , ivermectin is extremely well tolerated , effective , orally active , and associated with long-term safety at the current clinical dose ( single dose of 12 mg ) [17] . Clinical studies have shown that it is safe in humans at doses up to 10-fold higher; however , further increased dosage provokes severe neurotoxicity [18] . To catalyze application of ivermectin’s therapeutic potential in needy areas throughout the world , Merck & Co . has donated it for over 20 years to treat patients with river blindness , human onchocerciasis , and lymphatic filariasis [14] . In the case of moxidectin , single doses of up to 36 mg were safe in humans , but not doses of 54 mg [23] . The extensive use of macrocyclic lactones in veterinary medicine has generated valuable pharmacological data that could guide selection of these drugs and facilitate their use in humans . Milbemycin oxime is a broad-spectrum intestinal anti-parasitic drug used to treat roundworm , hookworm and tapeworms in cats and dogs; it is also reported to be safer than ivermectin [12] . Administered routinely at a dose of 0 . 25 mg/kg , it showed no signs of toxicity [25] . Although LD50 values after oral administration in dogs are higher than 200 mg/kg , a single dose of 3 . 8 mg/kg was reported to cause reversible neurological signs ( trembling , ataxia ) in dogs [30] . In contrast , selamectin has fewer neurological side effects , and can be administered topically , subcutaneously , or orally to treat a variety of ecto- and endo-parasitic infections in cats and dogs . It is the drug of choice in avermectin-sensitive collies since it has no adverse effects [REVOLUTION—fact sheet] . A toxicity study in female CD1 mice found that selamectin was well tolerated at up to 300 mg/kg body weight ( bw ) , while similar doses of milbemycin oxime were toxic [31] . In the case of milbemycin , doses up to 24 mg/kg bw were safe in cats and dogs [27] and one study reported that doses up to 94 mg/kg bw were safe in dogs [28] . In addition , a 3-month repeated dose toxicity study in dogs found an oral dose of 40 mg/kg/day to be safe [28] . Extrapolated to humans , this corresponds to a dose of 2 , 800 mg/day ( for a 70 kg adult ) . Confirming this extrapolation , the LD50 in rats and mice could not be demonstrated and it was higher than 1 , 600 mg/kg bw [Stronghold ( selamectin ) —Product profile] . Based on established clinical experience in humans at low dosages , Omansen et al . [10] chose to study the anti-mycobacterial activities of ivermectin and moxidectin . They reported MIC values between 4 and 8 μg/mL against M . ulcerans and inactivity ( MIC ≥32 μg/mL ) against M . marinum . We confirmed the activities of macrocyclic lactones , but found different specificities against bacterial isolates representing these two species ( Table 1 and Fig 1 ) . In contrast to analyses reported by Omansen et al . [10] , we detected little or no activity of ivermectin and moxidectin against M . ulcerans isolates but they were active against M . marinum strains . Such discrepancies could reflect variations in methodology . While Omansen et al . used Mycobacteria Growth Indicator Tubes ( MGIT ) and bioluminescence assays for their inhibitions assays [10] , we performed metabolic-based activity assays in liquid cultures grown in 96 well plates . Subtle differences in methodology are known to play a critical role in quantification of the anti-mycobacterial activity of ivermectin [16] . Our in vitro results can be integrated with available PK data to predict which drug would be more suitable for anti-BU therapy . While no human data are available for milbemycin oxime and selamectin , extensive pharmacological data from animal studies provide valuable information to accelerate clinical testing . Standard oral doses ( in μg/kg bw range ) of ivermectin , moxidectin and milbemycin oxime used to treat helminths in humans only achieve low concentrations in the plasma ( ng/mL range ) . Area Under the Curve ( AUC ) values for moxidectin and milbemycin oxime are higher than those of ivermectin , mainly due to their extended residence times ( higher half-life ) . However , the much higher doses needed to achieve concentrations sufficient to kill mycobacteria might not be possible due to toxicity . In contrast , selamectin toxicity is negligible at comparable doses . Standard dose administration of selamectin is in the mg/kg bw range ( versus μg/kg bw ) and doses as high as 95 mg/kg bw have been administered without any side effects [28] . The ability to deliver such high doses without toxicity is also reflected in the elevated concentrations of selamectin that can be achieved in the plasma . These concentrations in the μg/mL range are several fold higher than MIC values against M . ulcerans [21 , 27] which , together with a long half-life ( in days ) , allows for high AUC values . In fact , AUC/MIC values are the most predictable PK/PD parameter for the anti-mycobacterial activity of the avermectins [9] . Similarly , AUC/MIC ratios between 10 and 15 are also needed for bactericidal activity against M . ulcerans ( Fig 2B ) . Thus , when theoretical AUC/MIC values were calculated by integrating data from available PK literature with those from our in vitro data , only selamectin was predicted to have therapeutic activity against M . ulcerans ( Table 2 ) ( nb , calculations based on the lower in vitro MIC measurement reported by Omamsen et al . [10] generated the same conclusion ) . We would also like to point out that when we made corresponding calculations based on in vitro MIC data for M . tuberculosis [9] , selamectin would also be the avermectin of choice for tuberculosis therapy . A synergistic interaction between rifampicin and ivermectin against M . ulcerans has also been reported [10] . Rifampicin is the cornerstone drug for BU treatment . Thus , co-administration of rifampicin with any synergistic , orally available drug would be ideal . Rifampicin is an inducer of the P-gp and other transporters . P-gp protects mammals not only by excluding macrocyclic lactones from the central nervous system , but also by limiting the uptake of compounds from the gastrointestinal tract and by promoting their excretion in the liver , kidney , and intestine . While ivermectin is a good P-gp substrate , thus further reducing available levels of this drug , selamectin is a poorer P-gp substrate [12 , 21] and its plasma levels would be affected to a lesser extent allowing for a potential co-administration with rifampicin . In summary , drug repositioning is an interesting avenue to provide new treatments for neglected diseases . We have tested the family of commercially available macrocyclic lactones against M . ulcerans and M . marinum and demonstrated that milbemycin oxime and selamectin are the most active drugs ( MIC = 2 μg/mL ) . Integrating these values with information gathered in a literature review of the pharmacological properties ( toxicity and PK/PD profiles ) of ivermectin , moxidectin , milbemycin oxime and selamectin , revealed selamectin as the most promising avermectin candidate for anti-BU treatment . Although selamectin is not approved for use in humans , extensive information is available on its pharmacological properties in animals , thus facilitating its progression into clinical trials . These would be warranted if its activity could be validated using in vivo models of M . ulcerans infection . Pre-clinical and clinical development of any drug is a task that one research group cannot achieve alone . Thus , we urge collaboration among the research communities , pharmaceutical companies , and non-governmental organizations to validate the potential of macrocyclic lactones , especially selamectin , as a new anti-BU treatment .
|
Buruli ulcer ( BU ) is a chronic debilitating mycobacterial disease of the skin and soft tissue caused by Mycobacterium ulcerans . It is mainly found in tropical regions and often linked to poverty . BU can be cured in most cases with the standard treatment , a combination of rifampicin and the injectable antibiotic streptomycin . However , new optimized treatment regimens are needed , especially to prepare for an eventual development of resistance to rifampicin , the most efficacious drug for BU therapy . Since traditional antibacterial drug discovery is not a practical option for BU , using approved drugs for alternative clinical indications would be a more economical and faster way to implement new anti-BU therapies . We reported previously that anti-parasitic avermectins are active against Mycobacterium tuberculosis . Here we show that some are also active in vitro against other mycobacterial species , including M . marinum and M . ulcerans . In this study , we undertook a comprehensive approach to evaluate additional macrocyclic lactones including compounds used in veterinary medicine . Based on our in vitro measurements of their activities and a literature review of their pharmacokinetic properties , we present strong arguments that selamectin is the avermectin with the highest potential for being repurposed for BU treatment .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Selamectin Is the Avermectin with the Best Potential for Buruli Ulcer Treatment
|
The non-specific symptoms of Ebola Virus Disease ( EVD ) pose a major problem to triage and isolation efforts at Ebola Treatment Centres ( ETCs ) . Under the current triage protocol , half the patients allocated to high-risk “probable” wards were EVD ( - ) : a misclassification speculated to predispose nosocomial EVD infection . A better understanding of the statistical relevance of individual triage symptoms is essential in resource-poor settings where rapid , laboratory-confirmed diagnostics are often unavailable . This retrospective cohort study analyses the clinical characteristics of 566 patients admitted to the GOAL-Mathaska ETC in Sierra Leone . The diagnostic potential of each characteristic was assessed by multivariate analysis and incorporated into a statistically weighted predictive score , designed to detect EVD as well as discriminate malaria . Of the 566 patients , 28% were EVD ( + ) and 35% were malaria ( + ) . Malaria was 2-fold more common in EVD ( - ) patients ( p<0 . 05 ) , and thus an important differential diagnosis . Univariate analyses comparing EVD ( + ) vs . EVD ( - ) and EVD ( + ) /malaria ( - ) vs . EVD ( - ) /malaria ( + ) cohorts revealed 7 characteristics with the highest odds for EVD infection , namely: reported sick-contact , conjunctivitis , diarrhoea , referral-time of 4–9 days , pyrexia , dysphagia and haemorrhage . Oppositely , myalgia was more predictive of EVD ( - ) or EVD ( - ) /malaria ( + ) . Including these 8 characteristics in a triage score , we obtained an 89% ability to discriminate EVD ( + ) from either EVD ( - ) or EVD ( - ) /malaria ( + ) . This study proposes a highly predictive and easy-to-use triage tool , which stratifies the risk of EVD infection with 89% discriminative power for both EVD ( - ) and EVD ( - ) /malaria ( + ) differential diagnoses . Improved triage could preserve resources by identifying those in need of more specific differential diagnostics as well as bolster infection prevention/control measures by better compartmentalizing the risk of nosocomial infection .
Prior to the 2013–2015 epidemic of Ebola virus disease ( EVD ) , fifteen outbreaks caused by the virulent Zaire ebolavirus strain had been recorded since the identification of the virus in 1976 [1] . The West African EVD epidemic started in December 2013 , rapidly spreading from Guinea to Liberia and Sierra Leone to infect an estimated 28 , 600 people; over half of whom were in Sierra Leone [2] . Its unprecedented spread revealed a deadly potential to exploit weaknesses in public healthcare infrastructure [3] , and established it as a disease for which low-income countries are at disproportionate risk [4] . As repeat outbreaks are predicted in this region for the near future [5] , accurate , low-cost mechanisms to identify and triage EVD suspect cases are critical to ensure patient safety the sustainability of EVD surveillance . Cumulatively , EVD outbreaks prior to 2013 affected less than 2400 people [1] and yielded limited systematic research on its diagnostic features . One of the more comprehensive studies during this time concluded that many of the differential diagnoses were clinically indistinguishable from Ebola without specific molecular testing [6] . This problem was inherited into the current WHO triage guidelines , which consist of a binary evaluation of non-specific symptoms that are shared by the much more prevalent disease , malaria [7] . Indeed , during the recent outbreak over 50% of “suspect” Ebola patients admitted to ( the potentially contagious environment of ) many ETCs did not have Ebola . From a public health perspective , sensitivity is paramount when screening for highly contagious and fatal diseases such as EVD , and specificity is often sacrificed in favour of a more sensitive detection . However , once these suspect patients arrive at the treatment centres , specificity becomes far more important in order to accurately allocate patients to risk-appropriate wards and better distribute limited resources . During the recent outbreak , patients admitted to the ETC were further triaged into a higher risk “probable” ward on the basis of a clinically subjective assessment known as the “Ebola look”: since proven to have comparable accuracy to flipping a coin [8 , 9] . While compartmentalising risk by stratification is an essential component to infection prevention and control measures , patient triage should be sufficiently accurate to justify to its benefit . Thus far , studies conducted on patient data from Ebola Treatment Centres ( ETCs ) in Sierra Leone , Guinea and Liberia have identified several clinical characteristics as being variably predictive of EVD diagnosis [9–18] . Patients that present with symptoms of confusion , conjunctivitis , intense fatigue , hiccups , vomiting [9] , diarrhoea [9–11] , and anorexia [14] have been noted as having a higher probability of EVD infection over other differential diagnoses . Some of these studies have shown that a combination of symptoms [9] or their inclusion in a disease score prediction model [11] , is able to increase the odds of predicting EVD diagnosis . However , the variability across studies and their low positive predictive values show that further research is required before these strategies could be established as safe or effective triage techniques . Malaria infection is not only a prevalent confounding diagnosis for EVD triage , but it was also shown to kill more people than EVD during the 2013 outbreak [19] , which was likely due to its reduced prevention , diagnosis and treatment [20 , 21] . Consequently , mathematical modelling has shown that the incidence of malaria infection during EVD is estimated to increase [21] . However , despite these statistics , no studies have adjusted the predictive values of individual symptoms according to their statistical association with malaria infection: a strategy , which may not only significantly improve their predictive accuracy for EVD but also possibly identify malaria infection . In this retrospective cohort study , we analyse the clinical and epidemiological data of 566 patients admitted to the GOAL-Mathaska ETC in Port Loko , Sierra Leone . The diagnostic potential of each characteristic was analysed and incorporated into a statistically weighted and easy-to-use predictive score , designed to differentiate between EVD and malaria as well as greatly increase the specificity of EVD risk stratification whilst maintaining maximal detection sensitivity .
Ethical approval for this research was granted by the Sierra Leone Ethics and Scientific Review Committee ( SLESRC ) . This retrospective cohort study uses anonymized patient data collected between December 14 , 2014 and November 15 , 2015 at the GOAL-Mathaska ETC in Port Loko , Sierra Leone . Data comprised patient demographics , geographic location , clinical signs and symptoms , and laboratory results ( a rapid diagnostic test for plasmodium infection and a semi-quantitative RTPCR for Ebola viremia , both performed at triage ) . We evaluate the potential of clinical characteristics to predict EVD diagnosis and use these results to construct a symptom-based diagnostic risk-stratification score , which corresponds to the predictive power of the most prevalent symptoms adjusted for the major differential diagnosis of malaria infection . The ETC was run by the humanitarian organization GOAL Global in cooperation with the Sierra Leonean Ministry of Health and Sanitation ( MoHS ) . It opened in December 2014 and accepted 600 patients from a catchment area spanning 200km ( S1 Fig ) . On arrival at the ETC patients were allocated to “suspect” or “probable” wards according to the WHO guidelines [7] . Here , a “suspect” patient was selected for admission to the ETC based on the WHO guidelines which used various permutations of the following 3 elements: 1 ) acute fever , 2 ) contact history with an Ebola patient , and 3 ) any three of the following symptoms: headache , anorexia , lethargy , aching muscles , breathing difficulties , vomiting , diarrhoea , stomach pain , difficulty swallowing or hiccups ( as summarised in S2 Fig ) . The distinction between “suspect” and the higher-risk category of “probable” was based on subjective clinical assessment and circumstantial epidemiological evidence , as per the WHO recommendation . Blood was drawn from all patients on admission to the ETC and sent for Ebola virus testing at on-site laboratories managed by Public Heath England . An RDT malaria test was also performed at admission . Patients later testing positive for EVD by RT-PCR were transferred to the “confirmed” ward . All EVD ( + ) patients were treated according to standard treatment protocols developed by WHO and Médecins Sans Frontières [22 , 23] . This included empiric antimalarial treatment , broad-spectrum antibiotics , and nutritional supplementation for all patients , as well as oral or intravenous fluid rehydration . Patients were discharged from the ETC only after returning two negative Ebola-specific RT-PCRs spaced 72 hours apart and the final decision was conditional to physician approval . Patients still meeting case definition after 2 negative test results were admitted for longer periods in order to account for possible delayed or prolonged symptom presentations ( S2 Fig ) . The signs , symptoms and epidemiological data of each patient were recorded at triage by trained staff in a comprehensive and standardised questionnaire . Diagnosis was confirmed by semi-quantitative reverse transcriptase-PCR ( RT-PCR ) performed on the Cepheid GeneXpert instrument where the cycle threshold ( Ct ) value was used as an inverse proxy for viral load . Histidine-rich protein-II ( HRP-II ) antigen rapid diagnostic kits were used for the testing of malaria infection . While symptoms were reported by the patient , haemorrhaging , pyrexia , and disorientation were recorded by clinicians after examination . Haemorrhagic signs included visible blood loss such as hematochezia , hematemesis , haematuria , epistaxis , haemoptysis or persistent haemorrhage from an IV catheter site as well as subcutaneous haemorrhage such as purpura and petechiae . Pyrexia was defined as a body temperature over 38°C , measured using an infrared thermal sensor . Disorientation was measured by trained ETC clinicians as per the AVPU alertness scale ( where pain and unconsciousness were considered “disorientated” ) . Additionally , any specific mention of “confusion” or “disorientation” in the medical notes was also considered as positive for this variable . Of the 600 patients assessed , 10 were declared dead on arrival and 24 were classified as late transfers ( treated elsewhere and thus convalescent on arrival ) or had incomplete data . Thus , a total of 34 patients were excluded from this analysis . Of the 566 patients involved in the study , 100% had diagnostic test results for EVD , where , 27 . 5% tested EVD ( + ) ( n = 158 ) . 543/566 patients had malaria test results . The cohort was evaluated for missing values in each variable . Referral time ( the time in days from symptom onset to admission at the ETC ) had 20 cases of missing data . Further analysis was undertaken to evaluate the aetiology of missingness , which included demographic variables ( such as age and sex ) , clinical severity variables ( such as EVD viral load ) as well as the covariates used in the final scoring model . Here , we found that subjects with missing data did not differ systematically from those with observed referral time , which is in favour of the hypothesis that the data were “missing completely at random” . In addition , we performed a sensitivity analysis using the “Hotdeck” imputation technique , which showed that the model coefficients did not change when using complete data [24] . To maximize data fidelity , patient files were entered into a secure Microsoft Excel database and cross-checked by 3 independent and trained analysts . Entry of clinical data was overseen by members of the clinical ETC staff . Graphs were constructed using GraphPad Prism , version 6 . 0 . Univariate and multivariate analysis was conducted using STATA software , version 14 ( StataCorp ) . Score validation was performed using “RMS” R-Package ( R Development Core Team . ISBN 3-900051-07-0 , URL: http://www . R-project . org ) . Results were deemed statistically significant at a p-value of less than 0 . 05 . Epidemiological data and clinical variables were summarized by their frequencies and percentages . Univariate logistic regression was performed to assess the association between each predictor and the outcome of EVD diagnosis ( reported as Odds-Ratios ( OR ) and p-values ) . Potential interactions were tested where the functional form of continuous variables ( age and referral time ) was checked using a fractional polynomial model [25] . The linearity assumption was confirmed for age but not for referral time . To simplify the triage score , referral time was coded into two categories ( [4–9] days and [0–3] + [10–23] days ) . As there was an insufficient number of patients in the EVD ( + ) group ( EVD ( + ) = 158 , EVD ( - ) = 408 ) compared to the number of 29 potential predictors , only those associated to the outcome at a level of p<20% were considered into a Stepwise Backward selection procedure to fit a multivariable logistic regression model . Among the significant symptoms , those with the highest prevalence were favored for inclusion in the score . Model diagnostics was then performed to check for influential observations that impact coefficient estimates and a Hosmer-Lemeshow goodness-of-fit test was performed to assess calibration . Discriminative performance of the final model was assessed by calculating the Area Under the Receiver Operating Characteristics ( ROC ) Curve ( AUC ) and its 95% confidence interval . This value is a representation of the performance of a binary classifier system where the true positive rate ( sensitivity ) is plotted against the false-positive rate ( 1 − specificity ) . On this graph , perfect classification is represented by 100% area under the curve ( AUC ) . The β-coefficient = log ( OR ) for each covariate of the final model was converted into an integer-based point-scoring system . The score was then derived as the sum of the covariates’ weighted scores . Internal validation using the bootstrap method ( repeated 1 , 000 times ) as described in Harrell et al . [26] was used to provide a more accurate estimate of the performance of the original model ( AUCoriginal ) . The algorithm allows calculating the optimism of the predictive discrimination in the original model . The difference ( AUCoriginal−optimism ) gives the bootstrap-corrected performance of the original model . In this analysis the outcome was a categorical dependent variable with three categories: 1 ) EVD ( + ) only , 2 ) Malaria ( + ) only , and , 3 ) Double negative ( EVD ( - ) /Malaria ( - ) ) . To identify factors associated with the outcome , we performed a multinomial logistic regression analysis using the double negative group as a reference . Relative-Risk Ratios ( RRR ) and p-values were calculated to assess the strength of discrimination between the three categories .
Of the 566 patients included in this study , 27 . 5% tested positive for EVD ( n = 158 ) . Malaria test results were available for 543 patients , of whom , 34 . 6% were positive ( n = 188 ) ( Fig 1A ) . Gender was evenly distributed among admissions and there were no significant differences between EVD ( + ) and EVD ( - ) cohorts ( Fig 1B ) . Confirming its role as a major differential diagnosis , malaria infection was 2-fold more likely in the EVD ( - ) cohort than in the EVD ( + ) cohort ( p = 0 . 005 ) ( Fig 1C ) . This quantifies the need for malaria-sensitive triage in order to better separate EVD ( + ) and EVD ( - ) patients . The mean age for all ETC admissions was 32 . 4 years , which was similar for EVD ( + ) and EVD ( - ) cohorts ( 30 . 6 vs . 33 . 1 years respectively ) ( Fig 1D ) . Indeed , probability of being infected with EVD did not vary with age ( Fig 1E ) , unlike malaria , which was more probable at younger ages ( Fig 1E ) . Oppositely , the probability of being neither EVD ( + ) nor malaria ( + ) increased with age , indicating a wider range of differential diagnoses among older patients ( Fig 1E ) . Geographically , EVD and malaria prevalence was clustered in several locations across the catchment area of the GOAL-Mathaska ETC , where Kambia district had the highest percentage of EVD ( + ) cases among admissions ( Figs 1F and S1 ) . These variations could be related to the physical distance of the referring centre from the ETC , where the percentage of EVD ( + ) admissions increased by over 20% with increasing distance ( Fig 1G ) . According to the WHO guidelines [7] , pre-EVD-testing triage of suspect Ebola cases took place in 2 stages ( S2 Fig ) . Firstly , patients were identified for admission to the ETC after meeting the symptomatic criteria of the case-definition . As shown in Fig 1A , 72 . 5% of all patients were incorrectly selected for admission into the ETC ( i . e . later testing EVD ( - ) ) . The next stage of pre-EVD-test triage used clinical and epidemiological grounds to discriminate a higher risk “probable” group . While this process correctly identified 89% of all EVD ( + ) patients for allocation into the probable ward , 46% of selected patients in this high-risk ward later tested EVD ( - ) ( Fig 2A ) . Nevertheless , this process successfully reduced EVD ( + ) patients in the lower-risk “suspect” ward to 3% ( Fig 2A ) . Once patients were admitted to the ETC , discharge was conditional on two EVD-negative test results spaced 72 hours apart in addition to clinical approval ( S2 Fig ) . Among the EVD ( - ) patients admitted to the ETC , the average number of days spent awaiting discharge approval was 12 hours longer for those infected with malaria ( p = 0 . 045 ) ( Fig 2B ) . A recent report by Levine et al . described an elegant diagnostic score to improve pre-test triage accuracy by combining the weighted points for EVD contact ( +2 ) , diarrhoea ( +1 . 5 ) , anorexia ( +1 ) , myalgia ( +1 ) , dysphagia ( +1 ) and abdominal pain ( -1 ) [11] . Using this algorithm , we were able to externally validate the relevance of the score on our cohort , obtaining an area under the ROC curve of 76 . 8% ( Fig 2C ) ( almost identical to Levine et al . , who obtained 75% ) . However , even with this risk stratification , the “very high” risk category still included over 40% EVD ( - ) patients ( Fig 2D ) , which would have been only a marginal improvement ( <5% ) on the current WHO criteria used for admission to the “Probable” ward ( S2 Fig ) [7] . In an attempt to improve the accuracy of EVD ( + ) triage , we analysed the prevalence and diagnostic potential of the major clinical signs , symptoms and laboratory values among the EVD ( + ) and EVD ( - ) patients . Symptoms reported by over 50% of EVD ( + ) patients at triage were asthenia , myalgia , anorexia , vomiting , diarrhoea , pyrexia , and headache ( Fig 3A and Table 1 ) . The prevalence of several triage symptoms was notably different between EVD ( - ) and EVD ( + ) patients , as can be seen by comparing their ranking ( Fig 3A ) or their differential prevalence ( Fig 3B ) . As expected , a history of possible “sick contact” with an EVD ( + ) patient was approximately 50% more common among those later diagnosed as EVD ( + ) . Further , 20% more EVD ( + ) patients reported to the ETC within 4–9 days of their first symptom compared to their EVD ( - ) counterparts . The clinical features of conjunctivitis and diarrhoea , vomiting and pyrexia were over 10% more prevalent in EVD ( + ) patients at triage . Oppositely , malaria infection , dyspnoea and myalgia were over 10% more prevalent in EVD ( - ) patients ( Fig 3B ) . Univariate logistic regression revealed several signs and symptoms that were strongly predictive for the diagnosis of EVD and statistical significance was generally found among characteristics with the highest differential prevalence , such as sick contact , conjunctivitis , diarrhoea , referral time of 4–9 days , pyrexia , dysphagia , haemorrhage and hiccups ( p<0 . 05 for all ) ( Table 1 ) . Oppositely , we found the strongest predictors for not having EVD were myalgia , dyspnoea and malaria infection ( p<0 . 05 for all ) ( Table 1 ) . Indeed , malaria infection is a prevalent differential diagnosis of EVD manifesting with many of the same symptoms and may play a major role in reducing the level of triage accuracy [8] . In an attempt to better discriminate between the symptoms defining EVD and malaria , we analysed the differential prevalence and predictive potential of symptoms between EVD ( + ) /malaria ( - ) and EVD ( - ) /malaria ( + ) patient cohorts . Here , we identify several of the most predictive triage symptoms for malaria , such as dyspnoea , oedema , myalgia , and disorientation , which are thus poor indicators for EVD in a malaria endemic region ( Fig 3C and Table 2 ) . Univariate analysis on the predictive value of these symptoms identified conjunctivitis , diarrhoea , vomiting , pyrexia , hiccups and haemorrhage as the strongest differential indicators for EVD infection in a malaria-endemic region ( p<0 . 05 for each ) ( Table 2 ) . The number of days from symptom onset to admission at the ETC ( i . e . “referral time” ) was available for 87 . 3% of the EVD ( + ) cohort and 83 . 9% of the EVD ( - ) cohort . The mean number of days from symptom onset to admission did not differ significantly between EVD ( + ) and EVD ( - ) cohorts ( 4 . 2 days vs . 5 . 3 days respectively , p = 0 . 16 ) ( Fig 4A ) . However , EVD ( + ) patients were 2 . 1 fold more likely to report to an ETC 4–9 days from symptom onset ( p<0 . 0001 ) ( Fig 4B ) . Overall , gender and age were not significant factors in the time taken for a patient to present at an ETC . Referral time across age groups is shown in Fig 4C . We next investigated whether referral distance affected referral time . Comparing patients from the Port Loko and Kambia districts ( average distances from the ETC are 27 . 1 and 40 . 0 km respectively ) , we found no significant difference in referral times . Temporal analysis showed that referral sensitivity improved among the EVD ( + ) cohort as the epidemic progressed ( Fig 4D ) until June 2015 , when the last positive EVD case was admitted to the ETC ( albeit not significantly different from EVD ( - ) ) . Performing multivariate analysis , we selected the clinical characteristics most predictive for EVD infection when comparing EVD ( + ) vs . EVD ( - ) as well as when comparing EVD ( + ) /malaria ( - ) vs . EVD ( - ) /malaria ( + ) ( Tables 1 and 2 ) . By stepwise backwards elimination , and prioritizing the most prevalent symptoms , we identified 8 characteristics which yielded significant predictive values in both comparison groups . Characteristics that were statistically significant predictors of EVD infection were sick contact , conjunctivitis , diarrhoea , a referral time of 4–9 days , haemorrhage , dysphagia and pyrexia ( p<0 . 05 for all ) . Additionally , we selected myalgia , as a significant negative predictor of EVD infection . We then calculated weightings from their predictive coefficients with the aim to find a simplified scoring model using whole integers and calculations limited to subtraction or addition . Testing the sensitivity and specificity of these weightings for the prediction of EVD infection , we found that the characteristics yielded an area under the ROC curve ( AUC ) of approximately 90% ( 89 . 61% for the comparison between EVD ( + ) vs . EVD ( - ) ( CI95%: 86% , 93% ) ( Fig 5A ) and 88 . 80% for the comparison between EVD ( + ) only vs . malaria ( + ) only ( CI95%: 84% , 93% ) ( Fig 5C ) ) . The risk category cut-offs are illustrated in Fig 5B and each category contains at least 10% of the cohort . Fig 5D shows that the selected variables and cut-offs not only discriminate between EVD ( + ) only and double-negative patients but also between EVD ( + ) only and malaria ( + ) only patients . Further , our score predicts double-positive patients similarly to EVD ( + ) only patients ( Fig 5D ) . Examining the accuracy of the score on our cohort , we found that the “very high” classification was able reduce the EVD ( - ) patients in the high-risk group to less than 3% ( Fig 5E ) . Further , the “high” risk category contained 80% correctly classified EVD ( + ) patients ( >95% specificity ) ( Fig 5E ) . At the other end of the scale , the “very low” risk category contained over 95% EVD ( - ) patients ( Fig 5E ) and represented approximately 40% of the total cohort ( Fig 5F ) . A table listing the full details and intercept of the multivariate analysis is available in the supplement ( S1 Table ) . An internal validation of the score to discriminate EVD ( + ) from EVD ( - ) samples yielded a final discriminative power of 88 . 73% ( Table 3 ) . While a referral time of 4–9 days was significantly predictive of EVD diagnosis over the entire timeframe of the study ( Fig 4B ) , we tested the performance of our scoring system on patient populations arriving before and after this threshold and found minimal changes to sensitivity and specificity ( S3 Fig ) where our score maintained an AUC of over 85% . As our scoring system is designed to be sensitive to endemic malaria , another potential limitation is that it may not work well on co-infected EVD ( + ) /malaria ( + ) patients . However , testing the score on co-infected patients within our cohort , we maintain an AUC of 91% ( CI95%: 85 . 9% , 96 . 7% ) for discrimination of EVD infection ( i . e . no change ) ( Table 4 ) . An additional temporal concern would be malaria seasonality . However , testing scoring accuracy on the population presenting to the ETC during the low malaria transmission months ( November to April ) showed that the discriminative power remained within 3% of the overall value ( 85 . 56% AUC ) . As anticipated , this malaria-sensitive score was more powerful during the malaria season ( 98 . 55% AUC ) ( S4 Fig ) . Tweaking the period considered as “high malaria transmission” by a month in either direction had no statistical effect . A printable template of the scoring system is found in Fig 6 , including a probability curve on which to extrapolate the risk of EVD infection .
Various biases plague patient-reported data , where patients may deny EVD contact or misremember the date of symptom onset: all concerns raised previously by similar reports [9] . In our study , a referral time of 4–9 days was a highly significant discriminator between EVD ( + ) cases and both EVD ( - ) and malaria ( + ) only patients . Referral time may be particularly prone to socioeconomic nuance as it is inextricably linked to healthcare seeking behaviour . However , a systematic study on 4 , 437 cases of Ebola transmission in Liberia , showed no significant differences in referral time or hospitalisation access across socioeconomic strata [35] . Further , we show that there were no significant differences in referral time between genders or among different age groups ( Fig 4C ) . The referral time of 4–9 days was significantly predictive of EVD diagnosis over the entire timeframe of the study ( Fig 4B ) and testing the performance of our scoring system on patient populations arriving before and after this threshold resulted in minimal changes to sensitivity and specificity ( S3 Fig ) . Some differences in reporting behaviour do exist however . For example , it has been previously shown that adults in Sierra Leone have a significantly higher incidence of reporting possible EVD infection as compared to children [36] . In this study , we have a similar finding but show that the probability for receiving an EVD ( + ) test result was similar across all ages ( Fig 1E red line ) . This is explained by our observation that older patients were more likely to report to ETC with symptoms unrelated to EVD ( Fig 1E , blue line ) . As our scoring system is based on its malaria-sensitive discrimination of EVD ( + ) patients , a potential limitation is that it may not work well on co-infected EVD ( + ) /malaria ( + ) patients . However , testing the score on co-infected patients within our cohort , we maintain an AUC of 91% ( CI95%: 85 . 9% , 96 . 7% ) for discrimination of EVD infection ( i . e . no change ) ( Table 4 ) . An additional concern about a malaria-integrative score would be changing accuracy with malaria seasonality . However , testing scoring accuracy during the low malaria transmission months ( November to April ) also showed no significant difference in the discriminative power compared to the general population . As anticipated , this malaria-sensitive score performed better during the malarial transmission months of West Africa ( May to October ) . Here , the power to discriminate between groups increased by 9% compared to the overall population ( 99% vs . 90% ) : a welcome deviation , considering the potential confusion that malaria may cause to triage ( S4 Fig ) . Importantly , our ETC opened in December 2014 and the last EVD ( + ) patient was admitted to our facility at the end of June 2015 . Thus , the EVD ( + ) cohort is not fully represented across both seasons . Despite this high performance , the true accuracy of any scoring system can only be tested and improved by external validation on large independent cohorts , which pool statistics to fine-tune the weightings and ensure the most generalizable application . Indeed , as with any cohort study , the generalizability is often limited to the geographic and demographic profile of the selection criteria . In an effort to test the generalizability of this cohort , we externally validated the triage scoring system proposed by Levine et al . [11]: a scoring system developed for a rural cohort in Liberia . Here , our results differed by less than 2% , and served to validate the representational capacity of our cohort as well as display the robustness of using such scoring systems across geographically disparate areas with socioeconomic nuance and variable malaria prevalence .
As previously stressed , external validation and systematic meta-analyses are needed to fine-tune the statistical weightings of this score to further improve its accuracy and geographical relevance . However , as we may expect the symptoms and patient behaviour to evolve with each Ebola outbreak , it is becoming increasingly important to create machine-learning predictive tools , which are able to better adapt to the changing statistics of future outbreaks .
|
Four decades after the discovery of Ebola virus disease ( EVD ) , the sources , reservoirs and dynamics of infection are still largely unknown and thus the threat of re-emergence remains ever present . As EVD thrives on fragile healthcare systems in the developing world , it is essential that triage tools are low-cost and easy-to-use in order to best allocate limited resources and ensure sustainability of EVD surveillance . From a public health perspective , sensitivity is paramount when screening for highly contagious and fatal diseases such as Ebola . However , once these suspect patients arrive at the treatment centres , specificity becomes far more important in order to accurately allocate them to risk-appropriate wards and better distribute limited resources . Currently , pre-test triage to identify “suspect” Ebola patients consists of a binary evaluation of non-specific symptoms that are shared by the much more prevalent disease: Malaria . Using these guidelines , over 70% of patients selected for admission to the potentially contagious environment of an ETC did not have Ebola . Within the ETC , patients may be further triaged into a higher risk “probable” ward on the basis of a clinically subjective assessment known as the “Ebola look”: since proven to have comparable accuracy to flipping a coin . While compartmentalising risk by stratification is an essential component to infection prevention and control measures , patient triage should be sufficiently accurate to justify to its benefit . This study constructs an easy-to-use and highly accurate ( 90% ) triage scoring system that discriminates EVD infection risk in a malaria-sensitive manner: a strategy , which not only significantly improves the predictive accuracy for EVD but may also identify the ( more deadly ) infection of malaria .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"geographical",
"locations",
"ebola",
"hemorrhagic",
"fever",
"parasitic",
"diseases",
"diarrhea",
"signs",
"and",
"symptoms",
"gastroenterology",
"and",
"hepatology",
"neglected",
"tropical",
"diseases",
"africa",
"viral",
"hemorrhagic",
"fevers",
"infectious",
"diseases",
"conjunctivitis",
"sierra",
"leone",
"people",
"and",
"places",
"eye",
"infections",
"pain",
"management",
"diagnostic",
"medicine",
"hemorrhage",
"ophthalmology",
"myalgia",
"malaria",
"viral",
"diseases",
"vascular",
"medicine"
] |
2017
|
Predicting Ebola infection: A malaria-sensitive triage score for Ebola virus disease
|
Stability in a metabolic system may not be obtained if incorrect amounts of enzymes are used . Without stability , some metabolites may accumulate or deplete leading to the irreversible loss of the desired operating point . Even if initial enzyme amounts achieve a stable steady state , changes in enzyme amount due to stochastic variations or environmental changes may move the system to the unstable region and lose the steady-state or quasi-steady-state flux . This situation is distinct from the phenomenon characterized by typical sensitivity analysis , which focuses on the smooth change before loss of stability . Here we show that metabolic networks differ significantly in their intrinsic ability to attain stability due to the network structure and kinetic forms , and that after achieving stability , some enzymes are prone to cause instability upon changes in enzyme amounts . We use Ensemble Modelling for Robustness Analysis ( EMRA ) to analyze stability in four cell-free enzymatic systems when enzyme amounts are changed . Loss of stability in continuous systems can lead to lower production even when the system is tested experimentally in batch experiments . The predictions of instability by EMRA are supported by the lower productivity in batch experimental tests . The EMRA method incorporates properties of network structure , including stoichiometry and kinetic form , but does not require specific parameter values of the enzymes .
We use the following enzymatic systems as examples for our investigation . Three of these systems have been described previously and some experimental data are available to validate our predictions . The other system ( glucose to isoprene pathway ) has not been experimentally investigated . Methanol condensation cycle ( MCC ) ( Fig 2A ) is a metabolic pathway to convert methanol to higher alcohols with 100% theoretical carbon yield , in contrast to natural pathways like ribulose monophosphate ( RuMP ) which have a maximum of 67% theoretical carbon yield due to the decarboxylation of pyruvate [15] . The core of the pathway creates a C-C bond between two formaldehyde molecules derived from methanol for the generation of acetyl-phosphate , which can be enzymatically converted to acetate or ethanol . In this cycle , formaldehyde is incorporated into ribulose-5-phosphate ( Ru5P ) ( Fig 2A ) to generate hexulose-6-phosphate ( H6P ) by hexulose phosphate synthase ( Hps ) . H6P is then isomerized to fructose-6-phosphate ( F6P ) which can be cleaved by phosphoketolase . Erythrulose-4-phosphate ( E4P ) and F6P can then recombine via transaldolase , transketolase and isomerases ( Tal , Tkt , Rpe , Rpi ) to regenerate Ru5P . Alternately , xylulose-5-phosphate ( X5P ) can be cleaved by phosphoketolase , yielding G3P and acetyl-phosphate . G3P is then shuffled with F6P by transketolase to generate E4P and X5P , which can proceed to regenerate Ru5P via Tal , Tkt , Rpe , and Rpi . The X5P- and F6P-cleaving activities of phosphoketolase are referred to as Xpk and Fpk , respectively , and the pathway is investigated with different combinations of these activities . A molecular purge valve for the production of polyhydroxybutyrate from pyruvate in vitro was demonstrated by Opgenorth et al ( Fig 2B ) [16] . This system needs special attention to achieve redox balance , since pyruvate has a more reduced oxidation state than the product . To alleviate this cofactor imbalance , a method for dissipating excess reducing equivalents , termed a molecular purge valve , was designed for the conversion of pyruvate to downstream products like isoprene and poly ( hydroxybutyrate ) ( PHB ) . Two different pyruvate dehydrogenases ( PDH ) were used in the system—one with cofactor specificity for NADPH and one with specificity for NADH . The downstream pathway enzymes use NADPH to reduce metabolites and an NADH oxidase ( NoxE ) to dissipate the generated NADH . From two acetyl-CoA molecules , two enzymes are required to generate the final product PHB . A chimeric glycolysis system was demonstrated by Ye et al [17] ( Fig 2C ) . Canonical Embden-Meyerhof-Parnas ( EMP ) glycolysis generates a net of two ATP per glucose . In the chimeric system , a non-phosphorylating glyceraldehyde-3-phosphate dehydrogenase ( GAPN ) was used . This results in a system which is ATP balanced , making it more convenient for in vitro assays . Additionally , the system is NADH balanced since the final product was lactate , which has the same redox state as glucose . A system is considered for the conversion of glucose to isoprene ( Fig 2D ) . NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) and NADPH-dependent pyruvate dehydrogenase are used in the pathway . NADPH is used since the downstream reactions in isoprene synthesis use NADPH . The pathway converts three glucose to two isoprene molecules . Interestingly , this pathway is also ATP-balanced , with the ATP generated by the glycolytic pathway being used stoichiometrically downstream in the isoprene pathway reactions . However , to maintain redox balance , NADPH must be drained from the system , potentially via an oxidase or similar enzyme . This system is investigated both with and without a substrate-level regulation of glucokinase ( GK ) by G6P , implemented using an irreversible version of modular rate laws [18] proposed by Liebermeister . The kinetic form used is known as competitive inhibition , though many other kinetic forms are plausible . Inhibition of this step by G6P is well-known . For example , a human enzyme catalyzing this reaction is G6P-inhibited [19] . These equations show the effective kinetic forms of glucokinase used without and with regulation: NoRegulation:VGK=VmaxKm , Glc[Glc]+Km , ATP[ATP]+Km , ATPKm , Glc[ATP][Glc]+1 WithG6PInhibitionVGK=VmaxKm , Glc[Glc]+Km , ATP[ATP]+Km , ATPKm , Glc[ATP][Glc] ( 1+[G6P]Ki , eff , ATP ) +1
We used the four systems described in Fig 2 to examine the stability problem . In particular , we investigated how network structure affects the intrinsic possibility of reaching stability . Previous EMRA work starts from an ensemble of parameter sets that give the same reference steady state , and discards the parameter sets that generate a Jacobian matrix with a real part of an Eigenvalue greater than zero , which indicates instability . However , experimental systems are not guaranteed to be stable or reach a steady state . To place stability and steady state in a context which is more meaningful to experimental efforts , enzyme parameters were chosen completely at random , and the systems were then integrated in time domain to determine if a productive steady state was reached ( Fig 3A , dark blue bars ) . This method is more representative of experimental efforts which often have either little or indirect control over enzyme amount or activity ( in vivo ) , or don’t have rational methods for pathway balancing ( both in vivo and in vitro ) . Interestingly , the results show that pathways have very different likelihoods of resulting in a steady state ( Fig 3A , dark blue ) . The glucose to isoprene system had only 21% of randomly generated parameter sets reaching a non-trivial ( non-zero ) steady state . This could be because it is a relatively large system in terms of enzyme number and uses two different cofactors ( NADPH and ATP ) . A large system may be less likely to reach a steady state . If each enzyme has an acceptable range of values , then in a large system it is more likely that at least one of these values would be outside the range , resulting in system instability . However , when regulation of glucokinase was introduced via activation by ATP and inhibition by ADP , the likelihood of productivity jumped to 36% ( Fig 3A ) . Overall , these results show that intrinsic pathway structure and kinetic forms ( including regulation ) have a strong influence on possibility of reaching a productive steady state . The result is a varying , and sometimes low , likelihood of achieving stability and productivity . Thus , finding rational ways to balance pathways is an important goal which can improve and accelerate the pathway development process . If the enzyme parameters were first constrained to a fixed point by solving for parameter values which give reaction rates equal to the reference flux , then the probability of attaining stability is greatly increased . While this is not practical in experiments , the method proves useful in model construction . We found that ( Fig 3A , light blue bars ) the fraction of fixed points which were stable varied depending on the network structure . While most systems showed at least 99% of the parameter sets sampled to be at a stable steady state , the MCC ( Xpk-only ) system showed only 61% of parameter sets to be stable . Although for some systems the fraction of stable steady states is similar—5 systems which all show at least 99% stability by this measure—they have varying tendency to lose stability upon perturbation ( Fig 3B ) . Starting from the reference state , where parameter sets are chosen under the fixed point constraint , the region of instability could grow when enzyme parameter changes ( Fig 3B ) . Depending on the structure of the system , the instability region might grow in a different fashion upon perturbation , and eventually some might lose stability . This shows that stability of fixed points in metabolic systems is not guaranteed and that stability could be a critical factor in metabolic systems . EMRA uses continuous production models to simulate enzymatic systems . However , many experiments , including in vivo and in vitro , are conducted as batch processes . Thus , it’s not clear how a perturbation which causes bifurcation in a continuous system will inform the batch experiment . Where a pseudo-steady state may exist , the pseudo-steady state behavior can be predicted by the continuous model . In these systems , if a parameter set resides in a domain where no stable steady state exists in the continuous mode , then no stable pseudo-steady state exists in the batch mode . This can be justified by locally linearizing the input function to convert a pseudo-steady batch system to a continuous system . However , an experimentalist measuring only the product output at the end point would not detect the lack of stability . In this case , the product yield will gradually decrease even when the system has entered an instability region . To show how the existence of a continuous bifurcation could manifest itself in a batch system , we simulated a batch system in time domain . First , stable parameter sets were generated via EMRA in a continuous MCC system using Fpk/Xpk ratio as 1:3 . Then , the parameter sets were integrated using the continuation method to increase the phosphoketolase level until instability occurs , increasing Vmax for Fpk & Xpk at the same ratio . A representative parameter set is plotted in Fig 4A to show the effect of increased phosphoketolase on continuous steady state acetate flux up to the point of instability . As phosphoketolase increases , the flux towards product increases slightly before decreasing and finally becoming unstable . This parameter set was found to become unstable at a ~1 . 5-fold increase of phosphoketolase . Different amounts of phosphoketolase perturbation ( 1x , 1 . 1x , 1 . 7x , 1 . 8x , 2x , 10x –multiplier applied to both Fpk & Xpk Vmax values ) were chosen to show the dynamic response of the system in a batch simulation . All rate equations , parameter values and initial conditions were kept the same as in the continuous model ( i . e . all starting metabolite concentrations were normalized to unity ) , except that starting formaldehyde concentration was multiplied by 200 and the “in” and “out” reactions used in the continuous mode were eliminated to observe product accumulation in time domain simulation of the batch system . 200-fold increase in initial formaldehyde concentration was chosen arbitrarily to signify a batch reaction , in which the starting substrate was included as a single charge instead of being fed over time . It was found to adequately demonstrate the phenomena we were interested in , though other values could have worked as well . See Tables 1 and 2 . Interestingly , the final batch production observed for this system decreases gradually as phosphoketolase ( PK ) amount ( Fig 4B ) increases . In the continuous system , the underlying phenomenon is instability , a step change in the nature of the steady state . In the corresponding batch system , pseudo-steady state disappears because of instability . However , the product formation does not stop until key intermediates are depleted . Batch acetate production rate over time is plotted in Fig 4C . For the stable 1x and 1 . 1x conditions , a pseudo-steady state was achieved in which acetate production rate reached a constant level , only decreasing when the formaldehyde had been consumed . However , for the conditions which are past the instability point ( 1 . 7x – 10x ) , a steady rate of acetate production is never achieved . Instead , the rate decreases monotonically until it reaches zero . The productivity of the 10x condition falls the fastest , eventually resulting in the lowest production . This shows that a decrease in production , even gradually , in a batch system could be associated with an instability issue in an analogous continuous system . In Fig 4D , the acetate concentration over time is plotted to show how the system evolves over time . For systems that are stable , because the initial concentration of the starting substrate is much higher than the Km value of the uptake system , the rate of input holds largely constant until the substrate concentration approaches the Km value . During this time , the system is operating under a pseudo-steady state similar to a stable continuous system . This is seen in Fig 4 for 1x and 1 . 1x phosphoketolase concentrations . Thus , the property of continuous system simulation carries over , until substrate concentration approaches Km . Thus , the system is run almost the same as in a continuous system in the first 50 min time units or so ( Fig 4C ) , when most of the acetate is produced ( Fig 4D ) For systems that are unstable ( Fig 4 , 1 . 7x , 1 . 8x , 2x , 10x phosphoketolase concentrations ) , the output flux was not able to reach a steady-state ( Fig 4C ) , and it decreases rapidly from the start and approaches zero despite the presence of the initial substrate . The cumulative product formed ( acetate ) is the integral of flux over time ( Fig 4D ) , which decreases as the system moves further away from the bifurcation point . Additionally , we investigated the mechanism by which the bifurcation causes decreased production . In the 1x condition , F6P is maintained at a nonzero-level throughout production , while in the 2x condition , it is quickly depleted ( Fig 4E ) . R5P and X5P are also shown to deplete quickly in the 2x condition ( Fig 4F ) . Thus , it is the depletion of these cycle intermediates which causes cycle failure . A previous experimental effort ( [20] , data reproduced in Fig 4G ) showed that in in vitro enzymatic experiments , the batch production of acetate with from formaldehyde reached a local maximum with respect to phosphoketolase amount , supporting the EMRA analysis . In sum , EMRA could potentially have useful insights into experimental systems , by identifying enzymes which may be most sensitive to bifurcation , and how they affect the system in question . A purge valve system converting pyruvate to PHB was analyzed . Each enzyme is represented by a canonical Michaelis-Menten kinetic rate law , and the reference flux is fixed since there are no degrees of freedom . EMRA methodology was implemented in this system to show the effects of perturbation of each enzyme . High NADPH-dependent PDH ( PDHNADPH ) ( Fig 5B ) and low NADH-dependent PDH ( PDHNADH ) resulted in the most stability for the pathway . PDHNADH must be low to prevent too much pyruvate from taking this route which generates unusable NADH reducing power , while PDHNADPH must be high to ensure that enough NADPH is generated to allow for 100% yield from acetyl-CoA to PHB . The imbalance of these activities may cause system instability , according to EMRA . Indeed , the PHB pathway was experimentally demonstrated to have reduced production with a lower ratio of PDHNADPH:PDHNADH [16] ( Fig 5C ) , matching the results of EMRA . Another example of EMRA application is in a thermotolerant , cell-free glycolysis system which was demonstrated for the production of lactate from glucose by Ye et al ( Fig 6A ) [17] . Canonical Embden-Meyerhof-Parnas ( EMP ) glycolysis generates two net ATP per glucose , ( Fig 6A , Gap & Pgk enzymes ) . However , in the chimeric system , to prevent cofactor imbalance , a non-phosphorylating glyceraldehade-3-phosphate dehydrogenase ( GapN ) [21] was used—resulting in a net balance of ATP and NADH from glucose to lactate . Again , the system was modeled using EMRA methodology . The results show that increased glucokinase ( GK ) amount and glucose feed rate ( Fig 6B , GK , IN ) may cause instability . This system was experimentally tested for lactate production at different glucose feed rates ( Fig 6C , data from Ye et al [17] ) . It was found that beyond a certain point , increasing glucose feed rate reduced lactate production , even if the same total amount of glucose had been fed , matching the instability to feed rate ( IN ) predicted by EMRA . The instability apparently occurs by the depletion of ATP by glucokinase . ATP is required for both glucokinase and phosphofructokinase ( PFK ) . However , if glucose is fed too quickly , ATP may become depleted by glucokinase before it can be regenerated in lower glycolysis . A time domain simulation of this system was carried out using initial conditions similar to the experimental conditions reported [17] ( Fig 6D and 6E ) and a parameter set from EMRA which became unstable after increase of feed rate . The time domain simulation showed that at reference feed rate ( 1x ) , ATP level is maintained and lactate production continues . However , at 4x feed rate , the ATP is depleted and the lactate production stops . EMRA was also carried out on canonical EMP glycolysis converting glucose to lactate and similar instabilities were found ( S1 Fig ) . In both systems , reduction in PFK activity was shown to strongly increase chance of instability . This is because once a metabolite is past PFK , it may be used to regenerate ATP , so it’s important to ensure that the flux past PFK is sufficient to supply ATP for all of upper glycolysis . However , it’s more paradoxical that an increase in an enzyme would cause productivity and instability issues , particularly glucokinase , or even feed rate . To demonstrate the utility of EMRA in identifying potential points of instability , a not yet characterized pathway producing isoprene from glucose was investigated with EMRA ( full pathway stoichiometry in S7 Table ) . The pathway is ATP-balanced and maintains redox balance using an NADPH drain . EMRA identifies that the NADPH drain must be balanced , not too low or too high . GK & IN must not be too high , while all other enzymes must only not be too low ( Fig 7B ) . By introducing regulation of GK using modular rate laws [18] , the fraction of productive steady states increased ( Fig 3A ) . The stability to perturbation is also slightly improved for feed rate ( IN ) , GK and PFK ( Fig 7B ) . Interestingly , the NADPH drain is unstable to both decrease and increase . This could be because if the rate is too low , then NADP+ is not sufficiently available for GAPDH , and lower glycolysis is unable to regenerate ATP needed for earlier in glycolysis and later in the pathway–while if the rate is too high , NADPH will not be available for the biosynthetic steps of the isoprene pathway . This analysis shows that a longer pathway has many complex , interacting factors that can cause instability and that EMRA is able to identify some of these potential issues . Similar to the previously simulated glycolysis system , glucokinase , ( GK ) and phosphofructokinase ( PFK ) and feed rate ( IN ) showed some instability . Other enzymes show less sensitivity upon increase ( Fig 7B ) . Instability caused by decreasing enzyme concentration is common and seen in most if not all pathways . A rational experimental plan for this pathway would thus focus on having sufficiently high levels of most enzymes ( all except feed , glucokinase and drain ) , for example , by ensuring the total activity of each enzyme is significantly higher than the feed rate . Ensuring these enzymes are at a high level would ensure both stability , according to EMRA results , and the possibility of maximum productivity . For enzymes which become unstable at higher levels , more optimization is required . Levels of glucokinase , NADPH drain , and feed rate should be varied in order to avoid instability and to find the highest productivity condition . This significantly narrows the focus from 21 variables to just 3 .
The results show that EMRA has potential to be a valuable tool for investigating the propensity for stability of complex enzymatic pathways without a priori knowledge of specific enzyme parameter values . In three cases presented , ( MCC , molecular purge valve , chimeric glycolysis ) the experimental investigators were able to heuristically identify productivity issues ad hoc , but EMRA is able to unify all these results with a theoretical framework based on instability . Importantly , although some of the phenomena were experimentally determined , it was not necessarily known that instability of the system—causing a step change in the nature of the steady state , rather than a smooth change predictable by sensitivity analysis—could be an underlying reason . The success of the method with these systems presented here shows that it deserves consideration as a design tool in the invention of new pathways . The method has proven versatile enough to successfully predict features in three different pathways investigated in different laboratories and powerful enough to do so without a priori knowledge of specific enzyme parameter values . EMRA simulation of a longer and not-yet characterized pathway demonstrates the range of possibilities for potential applications of this technique . While the characterized pathways were optimized based on intuition , it’s possible that a longer pathway with more enzymes , such as the glucose-to-isoprene pathway , would be much more difficult to optimize without rational balancing methods like those presented here . The reduction of search space from 22 to 3 variables represents an exponentially more approachable experimental path towards productivity , resulting in 27 ( 33 ) experiments rather than about 10 million ( 321 ) if three different enzyme amounts are tested . Another insight provided by EMRA and follow-up analysis is the determination of failure modes for the pathways investigated . Using parameter and enzyme amount values in stable and unstable regions of the parameter continuation , time domain integration allows us to determine the failure modes for these pathways upon instability . In the MCC pathway , it is depletion of pathway intermediates—especially X5P , R5P and F6P—which causes productivity decline and eventual stopping . Although time domain simulations weren’t carried out in all systems , the demonstration of failure mechanism in the MCC system may lend credence to the other EMRA examples . In the chimeric glycolysis pathway , depletion of ATP eventually caused that pathway to stop when glucose feed rate was too high . Identifying these failure modes with EMRA is another potentially fruitful area of discovery . Glycolysis is a fundamental pathway of life and functions successfully in many organisms . However , our simulations and previous experiments ( Fig 6 , [17] ) have shown it can be unstable under high glucose feed conditions which apparently deplete ATP and accumulate hexose-monophosphates ( Fig 6D and 6E ) . Some hexokinase enzymes are product-inhibited by G6P , [22] however , the particular enzyme used in the experimental investigation ( from Thermus thermophilus ) was investigated and no G6P-inhibition was reported . [23] Interestingly , glycolysis has also been shown to be unstable to low levels of inorganic phosphate in yeast , a condition which prevents GAPDH from proceeding [24] . Glycolysis is a nearly universal pathway , but this evidence shows it to be unstable in some cases . This helps to explain the presence of elaborate regulations such as insulin and glucagon [25 , 26] in animals and the massively sophisticated regulation of phosphofructokinase [27 , 28] in many organisms . Rather than stability , alternate explanations such as chemical necessity [29] and thermodynamic efficiency [30 , 31] are more likely reasons for the universality of glycolysis . In these analyses , EMRA was used to successfully evaluate the stability of complex cell-free pathway assays . In vitro biocatalysis systems are a powerful alternative and complement to in vivo systems [32] . Importantly , however , this does not exclude the possibility of success with simulation of in vivo systems . Depending on growth mode ( exponential growth , stationary phase , fermentation etc . ) in vivo systems may have different reference fluxes , so more exploration is required to identify different possibilities . It is unsurprising that lower amounts of pathway enzymes or feed rate would hinder productivity . The powerful insight provided by these results is that for the pathways identified , increasing levels of certain enzymes or feed rates were shown to cause instability and consequently reduce production . A typical metabolic engineering approach may be to simply maximize the reaction rate of all pathway enzymes . However , we show here that for many enzymes , this will not always result in an optimal outcome . Additionally , we have shown that the intrinsic stability of pathways varies significantly depending on structure and kinetic forms . This highlights the importance of stability analysis in understanding metabolic systems . Additionally , it shows that many metabolic systems may be very difficult to balance without sufficient rational methods for analyzing which enzymes are most likely to contribute to pathway instability , and in which amounts . This shows the importance of EMRA and stability analysis in general in understanding pathways theoretically and exploiting them practically . The lack of requirement for a priori knowledge of specific enzyme parameter values could make EMRA particularly approachable for experimental researchers working with new pathways or unknown enzymes . This may be hampered somewhat by the need for sophisticated mathematical operations , though this obstacle could be overcome if an appropriate software suite is made available . We believe EMRA can significantly contribute to pathway development efforts and is an important contribution to the toolbox of metabolic engineering .
Production systems were described in Systems Description . Stoichiometric matrices for the pathways were constructed based on catalogued and reported reactions . Reversibilities of each enzyme were assigned based on if an enzyme i ) is an ATP-dependent kinase ii ) catalyzes a reaction with a highly negative ΔGo ( less than -20kJ ) , [33] or iii ) catalyzes a decarboxylation . The functional form used for each reaction used canonical Michaelis-Menten saturation kinetics . Affinity parameters ( Km ) are normalized to substrate concentration to reduce the number of parameters sampled for , and metabolite concentrations are normalized to unity . The normalized forms of the equations are as previously shown [7] . For the pathways chosen , exchange reactions were defined for inputs and outputs . Stoichiometric matrices , enzyme reversibilities and reference steady states for all simulations can be found in SI ( S1–S7 Tables ) . Ensemble modeling robustness analysis ( EMRA ) was performed on several enzymatic pathways that have been performed experimentally in the literature . EMRA is a technique which uses knowledge of an enzymatic pathway’s network structure to determine if its steady state is stable to changes in enzyme level . Michaelis-Menten style saturation rate equations are determined for each reaction based on stoichiometry and reversibility . Enzyme parameters are picked randomly to satisfy the reference steady state which is provided . For each enzyme , random values in a 100-fold range with a uniform distribution ( from 0 . 1 to 10 fold of the normalized metabolite concentration ) were selected for the scaled affinity parameters ( Km/[X]ss , r ) [7] . Vmax is solved for using the reference steady state flux value . To confirm the presence of a stable steady state , the system’s Jacobian is confirmed to have eigenvalues with only negative real parts [34] . Otherwise , the model is discarded . 1 , 000 parameter sets were generated this way for each system . For the completely random parameter values , a log-uniform 100-fold range of Vmax parameter was used . For the MCC ( Fpk/Xpk 1:3 ) system , the Fpk and Xpk Vmax were assigned ¼ and ¾ , respectively , of the value of a random number generated , to ensure that the overall phosphoketolase amount was controlled . Steady state determination was validated by decreasing integration time by 10-fold , and ensuring that the fraction of 1000 parameter sets which reach steady state was within the SD of the original condition ( n = 1000 x 3 ) . Vmax values were generated on a log-uniform scale , while Km was generated on a uniform scale . To test if method of generating random values was a significant factor , different methods were tested ( all values from a uniform distribution or log-uniform distribution ) but no major differences were seen ( S8 Table ) . To simulate batch systems from an analogous continuous system ( as in the simulation of the MCC system shown in Fig 4 ) , a minimum of changes were made . Functional forms for reaction flux equations were maintained . For the simulations in Fig 4 , an arbitrary parameter set was selected from those generated by ensemble modeling . For all rate law reactions used , reaction rate ( Vi ) was directly proportional to Vmax . Vmax is multiplied by a function of other parameters , ( λ¯ representing Km , Keq , etc . ) and normalized metabolite concentrations ( x¯ ) . To simulate a batch system , exchange reactions are removed by setting their Vmax to 0 . Normalized metabolite concentration values were used for both simulation conditions . To provide a supply of starting substrate , the initial amount of formaldehyde was multiplied by 200 . Normalized steady state ( continuous ) or initial ( batch ) metabolite concentration ( X¯i ) is shown in this table for continuous and batch simulations . Normalized metabolite amount , X¯i , is equal to the actual metabolite concentration , Xi , divided by the reference steady state concentration , Xi , SS . In combination , setting input and output Vmax to 0 and providing starting substrate formaldehyde allows the simulation to run as a batch system accumulating acetate while keeping the other rate equations and parameter values unchanged . For each parameter set in an ensemble , the steady state is perturbed using parameter-domain integration . This consists of numerically integrating using the continuation method . The mathematical justification for this method has been presented previously [7] . At each step in this integration , as an enzyme level is perturbed , the normalized metabolite concentration at the new steady state may be different , so it is necessary to re-compute the numerical values of the Jacobian matrix . If the new Jacobian has an eigenvalue with a positive real part , this indicates system instability [34] ( Fig 1 ) and the integration is halted . These computations allow for analysis at every level of enzyme perturbation . For each parameter set in an ensemble , the fraction which were stable ( i . e . all Jacobian matrix eigenvalues remain negative ) can be plotted versus the level of perturbation for each enzyme . Simulations were performed in MATLAB , and the code required to reproduce all figures is available at: https://github . com/theis188/theisen-plos-comp-bio .
|
A method of metabolic simulation called ensemble modelling for robustness analysis is used to predict the behavior intrinsic to the network structure ( stoichiometry and kinetic form ) of four enzymatic systems . Some network structures are shown to be prone to instability . Starting from a stable system , instability is also predicted to be caused by increasing amounts of certain enzymes . EMRA is a valuable tool for pathway design , particularly synthetic pathways which are uncontrolled and not stabilized through evolution .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"chemical",
"compounds",
"enzymes",
"metabolic",
"processes",
"ketones",
"enzymology",
"carbohydrates",
"organic",
"compounds",
"glucose",
"glycolysis",
"pyruvate",
"systems",
"science",
"mathematics",
"metabolites",
"enzyme",
"metabolism",
"enzyme",
"chemistry",
"computer",
"and",
"information",
"sciences",
"system",
"instability",
"system",
"stability",
"proteins",
"acids",
"chemistry",
"biochemistry",
"organic",
"chemistry",
"monosaccharides",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"metabolism"
] |
2016
|
Stability of Ensemble Models Predicts Productivity of Enzymatic Systems
|
The female gametophyte of flowering plants , the embryo sac , develops within the diploid ( sporophytic ) tissue of the ovule . While embryo sac–expressed genes are known to be required at multiple stages of the fertilization process , the set of embryo sac–expressed genes has remained poorly defined . In particular , the set of genes responsible for mediating intracellular communication between the embryo sac and the male gametophyte , the pollen grain , is unknown . We used high-throughput cDNA sequencing and whole-genome tiling arrays to compare gene expression in wild-type ovules to that in dif1 ovules , which entirely lack embryo sacs , and myb98 ovules , which are impaired in pollen tube attraction . We identified nearly 400 genes that are downregulated in dif1 ovules . Seventy-eight percent of these embryo sac–dependent genes were predicted to encode for secreted proteins , and 60% belonged to multigenic families . Our results define a large number of candidate extracellular signaling molecules that may act during embryo sac development or fertilization; less than half of these are represented on the widely used ATH1 expression array . In particular , we found that 37 out of 40 genes encoding Domain of Unknown Function 784 ( DUF784 ) domains require the synergid-specific transcription factor MYB98 for expression . Several DUF784 genes were transcribed in synergid cells of the embryo sac , implicating the DUF784 gene family in mediating late stages of embryo sac development or interactions with pollen tubes . The coexpression of highly similar proteins suggests a high degree of functional redundancy among embryo sac genes .
The life cycle of plants alternates between haploid gametophyte and diploid sporophyte generations . A central step in plant sexual reproduction is the transfer of sperm cells from the male gametophyte , the pollen grain , to the female gametophyte , the embryo sac , resulting in fertilization and the formation of a new sporophytic embryo . In flowering plants , each embryo sac develops within the sporophytic tissues of the ovule , which is itself located within the ovary of the flower . Embryo sac development is preceded by meiosis , and consists of precise series of mitotic divisions , nuclear migrations , cellularizations , and cell deaths ( reviewed in [1 , 2] ) . In Arabidopsis , the mature embryo sac consists of four cells: the egg cell , two synergid cells , and a large central cell . During fertilization , a pollen tube penetrates the sporophytic tissues of the ovule and terminates growth at one of the synergid cells of the embryo sac . Following the rupture of both the targeted synergid and the pollen tube cell , the two sperm cells fuse with the egg cell and the central cell to form the embryo and the endosperm . Genes expressed within the embryo sac are responsible for many aspects of embryo sac biology . Embryo sac–expressed genes control the developmental program of the embryo sac , as evidenced by the large number of female gametophytic mutants that result in the arrest of embryo sac development at various stages [3–9] . Embryo sac–expressed genes are also central to the fertilization process . Ovules that do not contain functional embryo sacs do not attract pollen tubes [9 , 10] , an observation which led to the suggestion that the embryo sac produces signals that guide pollen tube growth . Both Arabidopsis genetics [11] and in vitro studies with Torenia ovules [12 , 13] have identified the synergid cells as a source of embryo sac–derived pollen tube attractants . In addition , embryo sac–expressed genes are required for the reception of pollen tubes by the synergids [6 , 14 , 15] and for the coupling the initiation of seed development to fertilization [16–19] . There has been success in the identification of numerous female gametophytic mutants . However , the total set of genes expressed in the embryo sac is poorly defined due to the fact that the embryo sac is embedded within the sporophytic tissues of the ovule , making it difficult to directly isolate embryo sac tissue for gene expression analysis . Here , we used genetic subtraction to identify embryo sac–expressed genes by comparing gene expression in wild type ovules to that in determinate infertile1 ( dif1 ) and myb98 mutant ovules . DIF1 encodes a cohesin required for meiosis; sporophytic tissues of the ovule are unaffected in dif1 ovules , but gametogenesis is prevented by the failure of meiosis to produce functional megaspores [20 , 21] . dif1 ovules therefore represent a clean , genetic ablation of the entire embryo sac . MYB98 encodes a transcription factor expressed specifically within the synergid cells of the embryo sac [11] . The early stages of embryo sac development in myb98 ovules resemble wild type , and the only observable morphological differences of mature myb98 embryo sacs compared to wild type are abnormalities within the subcellular structures of the synergid cells [11] . In addition , myb98 ovules have an incompletely penetrant pollen tube guidance defect [11] . myb98 mutations therefore impact the last stages of embryo sac development and specifically impact interaction with pollen tubes . We used high-throughput ( 454 ) cDNA sequencing and whole-genome tiling arrays to compare gene expression in wild-type ovules to that in dif1 and myb98 ovules . Importantly , these two techniques allow for genome-wide measurement of gene expression that is unbiased toward annotated genes . We identified 382 genes that were downregulated in dif1 ovules and 77 genes that were downregulated in myb98 ovules . The majority of genes downregulated in each mutant belonged to families of small , potentially secreted proteins . Because most embryo sac–dependent genes were unannotated or recently annotated , only 31% were by reported by recent studies of embryo sac gene expression using the annotation-based ATH1 microarray [22 , 23] . Our results identify a surprisingly large number of embryo sac–expressed , secreted proteins as candidate extracellular signaling molecules during embryo sac development and fertilization , and in particular implicate the poorly understood DUF784 gene family as potential mediators of the last stages of embryo sac development or signaling interactions between the embryo sac and the pollen tube .
To obtain an unbiased , genome-wide survey of ovule gene expression , we sequenced stage 14 ovule cDNAs from male sterile ms-1 plants ( Landsberg ecotype ) using the high-throughput 454 sequencing method [24] . We obtained 249 , 440 cDNA reads , comprising a total of 26 . 5 million bp , with the reads having a median length of 106 bp ( Table 1 ) . 225 , 499 reads ( 90% ) could be confidently aligned to the Arabidopsis genome [25] using blat or blastn ( Dataset S1 ) , with the majority of unalignable reads consisting primarily of simple sequence repeats and/or PCR primer sequence . 28 , 732 reads matched equally well to more than one genomic location and thus represent transcripts from recent duplications . Eighty-five percent of alignable reads mapped to annotated exons , covering 12 . 5% of all annotated exonic sequence . In total , 15 , 312 annotated genes were matched by at least one cDNA read . The number of reads per gene ranged from 0 to 3 , 099 ( to AtMg00020 , a ribosomal protein encoded by the mitochondrial genome ) , with most genes having 0–5 reads , 15% of genes having ten or more reads , and 5% of genes having 25 or more reads ( Figure S1 ) . The aligned cDNA reads were 97 . 5% identical to the published genome sequence . The 2 . 5% of bases not matching genomic sequence were likely the result of cDNA sequencing errors as well as ecotype-specific polymorphisms due to the alignment of Landsberg cDNAs to the Columbia genome . We also searched for embryo sac–dependent transcripts across the entire Arabidopsis genome by using whole-genome tiling microarrays to compare gene expression in wild-type ( Columbia ecotype ) ovules to that in dif1 and myb98 ovules . We dissected ovules from mature ( stage 14 ) flowers , collecting sufficient material to yield at least 2 μg of total RNA for each of four biological replicates for each genotype ( wild type , dif1 , and myb98 ) , resulting in a total of 12 samples . After reverse transcription , second-strand cDNA synthesis , and double-stranded random labeling , samples were hybridized to Genechip Arabidopsis Tiling 1 . 0F arrays ( Affymetrix ) , which contain over 3 , 000 , 000 25mer perfect match probes spread across the Arabidopsis genome with a median gap of 10 bp between probes . The log2 transformed hybridization signals to the perfect match probes were quantile normalized across the twelve arrays; pairwise correlations ranged from 0 . 952 to 0 . 966 ( Table S1 ) . Mutations in DIF1 result in the ovules that entirely lack embryo sacs [20] . To identify embryo sac–dependent transcripts de novo , without bias towards existing gene models , we used a simple algorithm to identify genomic regions differentially expressed between dif1 and wild-type ovules . In brief , a Welch's t-test was performed for each probe comparing the log2 scale expression values of the four wild-type replicates to the four mutant replicates . An arbitrary threshold was applied to define probes that correspond to differentially expressed messages ( p ≤ 0 . 05 , log2 fold change ≥ 1 ) . Neighboring probe matches within 80 bp of each other that met this threshold were joined to define differentially expressed intervals , with the requirement that at least three differentially expressed probe matches were required to define an interval . Using this algorithm , we identified 1 , 099 genomic intervals that were downregulated in dif1 ovules compared to wild type . Of these intervals , 969 mapped to an annotated gene . The remaining 130 seemingly intergenic intervals were compared to the genomic alignments of the ovule cDNAs . After joining adjacent intervals ( those separated by less than 200 bp ) , 27 dif1 downregulated intervals that overlapped with ovule cDNA alignments were considered as putative unannotated genes . Using cDNA and EST sequences as guides , open reading frames ( ORFs ) were found for 22 of these putative genes ( Tables 2 and S3 ) . Sixteen of the newly identified ORFs were matched by at least three ovule cDNA reads with unique matches to the genome ( Table 2 ) . As an example of the array and cDNA data supporting one of the newly annotated genes , a region between the annotated genes At1g01300 and At1g01310 with substantially more expression in wild-type ovules than in dif1 ovules was detected as a differentially expressed interval , whereas surrounding genic and intergenic probes had highly similar expression values between all three genotypes ( Figure 1A ) . In this case , the same interval was also differentially expressed between wild-type and myb98 ovules , and overlapped with the genomic alignments of 11 ovule cDNAs ( Table 1; Figure 1A ) . The differentially expressed interval also overlapped well with a putative ORF and was assigned the name At1g01305 ( Figure 1A ) . We used the tiling array data to quantify changes in transcript levels between wild-type and dif1 ovules for the entire set of Arabidopsis genes , including both the 22 genes we identified as well as previously annotated genes ( The Arabidopsis Information Resource [TAIR] release 7 annotations , containing 27 , 029 protein coding genes , 3 , 889 pseudogenes , and 1 , 123 noncoding RNA genes [http://www . arabidopsis . org] ) . For each gene , a t-test comparing wild-type and dif1 signal intensities was performed across all probes matching that gene ( Table S4 ) . We defined genes as having significantly different expression in dif1 compared to wild type by setting a p-value threshold of 0 . 001 and a log2 fold change threshold of 1 ( which corresponds to a 2-fold change ) . At these cutoffs , we found 382 protein-coding genes that were expressed at lower levels , and 35 genes that were expressed at higher levels in dif1 ovules ( Tables 3 , S5 , and S6 ) . To empirically assess the extent to which sampling error contributed to the observed differential expression , we estimated the false discovery rate ( FDR ) by shuffling the eight arrays ( four wild type and four dif1 ) into two permuted groups of four and reanalyzed the data to identify the number of genes with seemingly differential expression between these arbitrarily grouped sets or arrays . To control for differences in gene expression relating to the dif1 phenotype , we considered only the 18 balanced permutations in which the two groups of arrays being compared each contained two wild-type arrays and two dif1 arrays . On average , 3 . 6 genes had the p-values less then 0 . 001 and log2 changes in expression greater than 1 for the balanced permutations of the wild-type and dif1 datasets . Therefore , we estimate the FDR to be approximately 1% for the 417 genes differentially regulated between dif1 and wild type at this threshold . Estimates of the FDR at more relaxed thresholds suggest that several hundred additional genes are differentially expressed in dif1 ovules with changes in expression less than 2-fold ( Table S2 ) . Because the dif1 mutation only affects cells that undergo meiosis , the simplest interpretation of these data is that the 382 dif1 downregulated genes are expressed preferentially within the embryo sac as compared the sporophytic ovule . It is also possible that some of these genes require the presence of the embryo sac for expression within the sporophytic ovule . To characterize the set of DIF1-dependent genes , we analyzed the abundance of protein domains in the sets of differentially expressed genes as compared to the total set of protein coding genes . Ten gene families were significantly overrepresented in the set of dif1 downregulated genes ( Table 3 ) ; 241 of the 382 dif1-downregulated genes belonged to one of these ten gene families . Several of these families , such as Domain of Unknown Function 784 ( DUF784 ) , DUF1278 , and DUF239 , lack homology to any protein with a known function . Two families , the Defensin-Like ( DEFL ) genes and the thionin-like genes , have homology to small , secreted antipathogenic peptides , whereas the Papaver Self-Incompatibility-Like ( PSIL ) genes have homology the pistil-secreted S1 protein of Papaver . The functions of these six families within the context of the Arabidopsis ovule are unknown . The remaining overrepresented families encode proteins with presumed functions as catalytic enzymes ( peptidases , lipases , and polygalacturonases ) or as enzyme inhibitors ( pectinmethylesterase inhibitors [PMEIs] ) . Many members of these gene families are encoded by tandemly arrayed , recently duplicated genes ( Figures 1B , 1C , S2 , and S3 ) . In addition to the overrepresentation of certain protein domains , the set of dif1 downregulated genes was highly enriched for genes encoding small proteins that contain putative signal peptides ( Table 3 ) . Seventy-eight percent of dif1 downregulated genes were predicted to encode for a signal peptide , as compared to 18% among all protein-coding genes , and 66% of dif1 downregulated genes were predicted to encode proteins that weigh less than 20 kilodaltons , as compared to 20% among the total set of annotated proteins ( Table 3 ) . This bias towards small proteins with signal peptides was related to the bias towards certain protein families; 91% of the 215 DUF784 , DUF1278 , PSIL , DEFL , PMEI , and thionin-like genes downregulated in dif1 encode proteins that contain putative signal peptides and that weigh less than 20 kD . The presence of a signal peptide can target a protein for one of several fates , such as localization to a membrane-bound organelle , localization to the cell membrane , or secretion from the cell . However , the abundance of putative signal peptides amongst DIF1-dependent proteins , as well as the fact that numerous DIF1-dependent genes have homology to proteins that are known to be secreted in other organisms or tissues ( e . g . , DEFL , thionin-like , and PSIL genes ) [26–28] , suggests that many embryo sac–dependent proteins have the potential to act outside of their cells of origin . The 140 dif1 downregulated genes that did not belong to the ten gene families listed in Table 2 represented a broad range of functionalities ( Table S5 ) . Several DIF1-dependent genes have known roles in embryo sac biology , including the synergid-expressed transcription factor MYB98 as well as two genes , FERTILIZATION INDEPENDENT SEED2 and MEDEA , that regulate the development of the central cell and endosperm [16 , 17] . While 382 genes were downregulated at least 2-fold in dif1 ovules , some genes were more highly downregulated . Most of the genes with large differences in expression levels between dif1 and wild-type ovules belonged to multigenic families encoding small , potentially secreted proteins ( Figure 2 ) . For example , 83% of the 189 genes downregulated at least 8-fold in dif1 ovules belonged to the DUF784 , DUF1278 , DEFL , PSIL , PMEI , or thionin-like gene families ( Figure 2 ) . All 23 genes that were downregulated at least 64-fold in dif1 ovules belonged to the DUF784 or DUF1278 families ( Figure 2 ) . We analyzed the expression patterns of 59 dif1 downregulated genes by reverse transcriptase PCR ( RT-PCR ) , with a focus on members of the DUF784 , DUF1278 , PSIL , and DEFL gene families ( Figure 3 ) . In all 59 cases , including eight previously unannotated genes , expression was lower in dif1 ovaries than in wild type ( Figure 3 ) . In addition to 382 protein-coding genes , 26 annotated pseudogenes were also significantly downregulated in dif1 ovules ( Table S9 ) . Most of these pseudogenes had a high degree of homology to adjacent , tandemly arrayed protein coding genes also downregulated in dif1 ovules ( e . g . , DUF784 pseudogenes ) . While some of the observed expression of these pseudogenes may have been due to cross-hybridization to transcripts from homologous protein-coding genes , the fact that 13 were uniquely matched by ovule cDNA reads indicates that many are in fact transcribed ( Table S9 ) . It seems that some recently duplicated , embryo sac–dependent genes have retained regulated , functional promoters despite having acquired frame shift or nonsense mutations within their ORFs . It is unclear as to what , if any , functional roles these expressed pseudogenes might play . Many embryo sac–dependent genes have similarity to each other at the nucleotide sequence level , suggesting a common origin and function . The ORFs of 109 dif1 downregulated genes were >90% identical to another dif1 downregulated gene . In most cases , highly similar genes were present in tandem arrays of apparently recently duplicated genes . Seventy-five of the DIF1-dependent genes could be grouped into 17 clusters of highly similar genes that shared at least 50% of their tiling array probes with another gene . Twenty-six of these partially ambiguous genes could still be identified as significantly downregulated in dif1 ovules based solely on the expression values of probes with unique matches in the Arabidopsis genome . Another 14 were uniquely matched by ovule cDNA fragments , providing evidence that they are expressed in the ovule . Nonetheless , for approximately 50 genes , it is difficult to be certain that the differential expression observed on the tiling array truly reflected the expression of each individual gene or if only a subset of genes were differentially expressed . The most extreme example of closely related embryo sac–dependent genes is that of 30 DUF1278 genes ( as well as three DUF1278 pseudogenes ) that are >95% identical to each other . Most probes on the tiling array that correspond to this cluster perfectly match multiple genes; moreover , it was not possible to design RT-PCR primers specific to any particular gene from this cluster . Fifty-six of the 60 cDNA reads matching genes from this cluster matched more than one gene equally well . It is therefore difficult to be certain that the expression observed for genes of this cluster corresponds to all 30 genes or to a subset of the 30 genes . However , the high number of cDNA reads mapping to this genes in this region , together with the high degree ( 40- to 70-fold ) of DIF1 dependence detected for this region , make it clear that as a unit , this region is expressed in an embryo sac–dependent manner . Moreover , the RT-PCR primers to this region ( i . e . , to gene At5g36350 ) failed to detect any expression from this cluster in dif1 ovules despite being perfectly complementary to most of the 30 genes ( Figure 3 ) , further demonstrating that no gene from this cluster is highly expressed in ovules that lack embryo sacs . In contrast to the set of dif1 downregulated genes , the 35 genes upregulated in dif1 ovules were not significantly enriched for any protein domains , nor for genes predicted to encode proteins with signal peptides or weighing less than 20 kD ( Tables 3 and S6 ) . Furthermore , the magnitude of upregulation was modest compared to changes in expression levels observed for dif1 downregulated genes . Only one gene was upregulated more than 8-fold in dif1 ovules ( Figure 2 ) . Whole-genome tiling arrays allow for the comprehensive , genome-wide measurement of gene expression . However , because the Tiling 1 . 0F array has been developed only recently , few studies that use it to measure gene expression have been published . In contrast , the Genechip Arabidopsis ATH1 Genome array ( Affymetrix ) , containing 22 , 500 probe sets that match to 23 , 688 genes , is a widely used tool to measure gene expression in Arabidopsis . Two recent studies used the ATH1 microarray to identify genes that are downregulated in ovules that lack embryo sacs: Yu et al . identified 249 genes ( representing 225 probe sets ) downregulated in sporocyteless/nozzle ( spl/nzz ) ovules [23] , and Steffen et al . identified 104 genes ( representing 86 probe sets ) downregulated in dif1 ovules [22] . A comparison to these datasets illustrates the utility of using genome-wide expression measures to profile gene expression and also validates the sensitivity of the whole genome tiling array as a means of quantifying gene expression . Only 31% of the dif1 downregulated genes identified by the whole genome tiling array analysis were reported as embryo sac–dependent by one or both studies using the ATH1 array ( Figure S4 ) . The large number of DIF1-dependent genes uniquely discovered by the tiling array is primarily due to the fact that a surprisingly large number of embryo sac–dependent genes were not measured by the ATH1 array . While 84% of all currently annotated Arabidopsis protein-coding genes had a corresponding probe set on the ATH1 array ( at least six of 11 probes perfectly matching ) , a significantly smaller percentage ( 41% ) of DIF1-dependent genes were represented by ATH1 probe sets ( Table 4 ) . In total , 224 dif1 downregulated genes did not have ATH1 probe sets ( Table 4 ) . In addition to failing to detect the majority of embryo sac–dependent genes , the ATH1 array is specifically biased against certain gene families . Whereas over 90% of DIF1-dependent genes encoding lipases , subtilisins , or polygalacturonases were represented in the ATH1 array , less than 20% of the DIF1-dependent genes belonging to families encoding small , functionally uncharacterized proteins ( DUF784 , DUF1278 , DEFL , PSIL , and thionin-like genes ) had corresponding ATH1 probe sets ( Table 4; Figure S4 ) . The poor representation of these families can be attributed to the fact that many members of these families were annotated after the design of the ATH1 array [29–31] . For example , 65 DIF1-dependent DUF784 , DUF1278 , and thionin-like genes were unannotated prior to the TAIR7 annotation release of April , 2007 [30] . The combined analysis of tiling array data and high-throughput cDNA sequencing led to the finding that large families of poorly understood , potentially secreted proteins are embryo sac dependent , a finding that was not evident from the more limited and biased sets of embryo sac–dependent genes detected by the ATH1 array . Considering only those genes with ATH1 probe sets , there was considerable overlap between the tiling array and ATH1 data; 76% of the 158 dif1 downregulated genes with ATH1 probe sets were reported as downregulated in at least one of the ATH1 analyses ( Figure S4 ) . Sixty-five genes were reported as downregulated in all three studies , and another 55 genes were reported as downregulated in both our analysis and in one of the other studies ( Figure S4 ) . Thus , the reproducibility and accuracy of gene expression quantification by the whole genome tiling array was at least roughly comparable to that of the more commonly used expression array . 114 spl/nzz downregulated genes were not identified as downregulated in either study using dif1 ovules ( Figure S4 . ) Some of these genes were downregulated in dif1 ovules , but at levels below our statistical thresholds . However , 69 SPL/NZZ-dependent genes were expressed at similar levels in dif1 and wild-type ovules ( p > 0 . 25 ) and seven were actually upregulated in dif1 ovules ( p < 0 . 05 ) . The fact that a large number of SPL/NZZ-dependent genes are not downregulated in dif1 ovules is most likely due to the different stages of ovule development at which SPL/NZZ and DIF1 act . SPL/NZZ is known to be required for the proper expression of several key genes during development of the somatic ovule , and spl/nzz ovules never initiate meiosis [32–35] . In contrast , the somatic development of dif1 ovules appears to be entirely wild type , and dif1 ovules initiate meiosis properly [20] . Therefore , it seems that approximately one-fourth of SPL/NZZ-dependent genes are not embryo sac–dependent but rather require SPL/NZZ for expression in the somatic ovule . Although many embryo sac genes are not detected , the ATH1 array is capable of allowing quantitative comparisons to published studies using the same platform . We measured gene expression in three biological replicates from functionally wild-type ovules ( from the male sterile ms-1 mutant ) on the ATH1 array . Normalized probe set expression values were calculated via the RMA method [36] from probe level data from the ovule arrays , together with probe level data from 41 sets of Affymetrix gene chip experiments from various wild-type tissues and developmental stages that did not contain stage 12 or later ovules [37 , 38] . We analyzed the data to identify genes for which ( 1 ) ovule expression was at least two times higher than that of any other tissue and ( 2 ) ovule expression was at least three standard deviations above the mean expression level in non-ovule tissues , resulting in 155 ovule-enriched genes ( Table S10 ) . Of the 158 DIF1-dependent genes with ATH1 probe sets , 55 were identified as ovule enriched ( Table 4 ) . Certain gene families were highly represented amongst the set of ovule-enriched genes , including all DIF1-dependent DUF784 , DUF1278 , and thionin-like genes with ATH1 probe sets , as well as the majority of DIF1-dependent DEFL and PSIL genes with ATH1 probe sets ( Table 4 ) . In combination , the ATH1 array data and RT-PCR data ( Figure 3 ) show that numerous members of the DUF784 , DUF1278 , DEFL , and PSIL gene families are expressed primarily within the ovule . In contrast to the total ablation of embryo sac tissue in dif1 ovules , myb98 embryo sacs are morphologically similar to wild type with exception of the subcellular structure of the synergid cells [11] . myb98 embryo sacs are also impaired in mediating pollen tube guidance [11] . Genes with reduced expression levels in myb98 ovules are therefore likely to represent genes active during the final stages of embryo sac development and during the initial steps of the fertilization process . Using the same significance thresholds as in the dif1 versus wild type comparison ( p ≤ 0 . 001 , log2 fold change > 1 ) , we found that 77 genes were downregulated in myb98 mutants compared to wild type , whereas 40 were upregulated ( Tables 3 , S7 , and S8 ) . As would be expected from the more severe dif1 phenotype and the fact that DIF1 is required for MYB98 expression , the set of MYB98-dependent genes is largely a subset of the DIF1-dependent genes; 76 of the 77 myb98 downregulated genes were also downregulated in dif1 ovules ( Figure 4 ) . The set of myb98 downregulated genes was even more highly enriched for genes encoding potentially secreted proteins ( 92% ) and proteins weighing less than 20 kD ( 84% ) than was the set of dif1 downregulated genes ( Table 3 ) . Thirty-seven of the 40 DUF784 genes encoded in the Arabidopsis genome were downregulated in myb98 ovules ( Table 3 ) . In total , DUF784 genes comprised nearly 50% of the myb98 downregulated genes , while other gene families overrepresented among dif1 downregulated genes were represented to varying degrees among the myb98 downregulated genes ( Table 3 ) . The magnitude of gene downregulation in myb98 ovules was less than that in dif1 ovules ( Figure 2 ) . Whereas all 40 DUF784 genes were downregulated at least 8-fold in dif1 ovules as compared to wild type , only 23 were downregulated 8-fold in myb98 ovules ( Figure 2 ) . No genes were downregulated 64-fold in myb98 ovules . RT-PCR analysis confirmed the degree of myb98 dependence among the different gene families . Consistent with the microarray results , all of the 16 DUF784 genes tested were expressed at lower levels in myb98 ovaries than in wild type , but many were detected at higher levels than in dif1 ovaries ( Figure 4 ) . Also consistent with the array results , only a fraction of the DUF1278 , PSIL , and DEFL genes tested were expressed at lower levels in myb98 ovaries ( Figure 3 ) . Of the 40 genes significantly upregulated in myb98 ovules compared to wild type ( Table 3 ) , 27 were downregulated in dif1 ovules ( Figure 3 ) . These appear to be genes that are expressed within the embryo sac , perhaps during the early stages of embryo sac development , and that go down in expression in response to MYB98 . However , the fold increase amongst myb98 upregulated genes was modest; only one was upregulated more than 4-fold relative to wild type ( Figure 2 ) . To localize the expression of DIF1-dependent and MYB98-dependent genes within the embryo sac , we constructed 11 transgenic lines expressing a glucouronisidase ( GUS ) reporter gene under the control of a putative promoter sequence corresponding to the genomic region upstream of an embryo sac–dependent gene . The four DUF784 promoters tested ( At5g35405 , At4g08025 , At5g34885 , and At2g21727 ) corresponded to genes that had large numbers of ovule cDNA reads ( Table S4 ) and that represent different subfamilies of the DUF784 phylogenetic tree ( Figure S2 ) . In T1 plants of all four DUF784::GUS lines , ∼50% of ovules had a single , punctate spot of GUS expression located at the extreme micropylar end of the embryo sac ( Figure 5A–5E ) . This localization of GUS expression is most consistent with transcription in the synergid cells , although in some cases GUS expression appeared to extend into the egg cell . GUS staining was observed in ovules of stage 12c flowers of DUF784::GUS plants , but not in ovules from stages 12b or earlier ( unpublished data ) . The four DUF1278 promoter::GUS lines analyzed also all resulted in GUS expression in synergid cells ( Figure 5F–5I ) . Unlike the DUF784 promoters , all of which corresponded to MYB98-dependent genes , only two of the DUF1278 promoters tested ( At5g54062 and At5g42895 ) corresponded to myb98 downregulated genes as determined by array analysis or RT-PCR . The synergid specific expression of At5g36340::GUS and At2g24205::GUS , neither of which correspond to MYB98-dependent genes , demonstrates that some synergid specific markers are MYB98 independent . In contrast to the expression in synergid cells observed for DUF784 and DUF1278 promoters , the DEFL promoter ( Figure 5J ) and PSIL promoters ( Figure 5K and 5L ) that were tested drove GUS expression within the central cell of the embryo sac . Of the 382 DIF1-dependent genes , 241 ( 63% ) belonged to one of ten gene families ( Table 3 ) . The subset of embryo sac–dependent genes that required the synergid-specific transcription factor MYB98 was even more enriched ( 80% ) for these same families ( Table 3 ) . While many of these gene families are similar in that they encode for small , potentially secreted proteins , each has a unique sequence profile , evolutionary history , and , presumably , role in embryo sac biology . More than 20% of the embryo sac–dependent genes encoded for defensin-like or thionin-like proteins ( Table 3 ) , two classes of small , secreted proteins with disulfide-linked cysteines . Members of both classes are known to have antimicrobial or antifungal properties [27 , 28] . Not counting pseudogenes , there are approximately 286 defensin-like genes and 62 thionin-like genes in the Arabidopsis genome . For the vast majority of these genes , no functional or biochemical data exists [31] . We found that 32% of all Arabidopsis DEFL genes and 19% of all Arabidopsis thionin-like genes are embryo sac–dependent ( Table 3 ) . While the embryo sac–expressed genes of these families have homology to antipathogenic peptides , it is unknown whether the role of these families in the embryo sac is related to defense against pathogens or whether they serve other roles as small secreted proteins . 22 embryo sac–dependent genes , including seven annotated in this work , have homology to the S1 self-incompatibility protein of the genus Papaver . In Papaver , pistil-secreted S1 proteins inhibit growth and trigger cell death of incompatible pollen [39] . While Arabidopsis thaliana is self-fertile , other Brassicaceae , including the near relative Arabidopsis lyrata , exhibit self-incompatibility , albeit through incompatibility factors that do not resemble those in the Papaver stigma [26] . We found that 40% of PSIL genes encoded in the Arabidopsis genome are embryo sac–dependent . Data from the AtGenExpress expression atlas [37] show that most PSIL genes that are not embryo sac–dependent are expressed most highly in anthers or pollen , suggesting that PSIL genes also play a role in the male gametophyte . The DUF784 and DUF1278 gene families are unique in that the majority of genes belonging to these two families are embryo sac–dependent ( Table 3 ) . In the case of DUF784 , all 40 genes encoded in the genome are downregulated in dif1 ovules , and 37 out of 40 are downregulated in myb98 ovules ( Table 3 ) . Neither family has apparent homology to any protein with a known molecular or biological function , although the DUF1278 genes are related to the EARLY CULTURE ABUNDANT1 gene identified in barley microspores [40] and to the EC1 gene that is expressed in wheat egg cells [41] . Both families are defined by the presence of six highly conserved cysteines that are present in almost all family members ( Figures S2 and S3 ) . While the pattern of conserved cysteines amongst DUF1278 proteins appears to be similar to that of DUF784 , sequence homology between members of these two families was not detected by blastp , nor did HMMer searches using a HMM from one family find significant homology to members of the other family . While several gene families were overrepresented among the set embryo sac–expressed genes , the extent of embryo sac specificity amongst DUF784 genes is particularly striking . All 40 genes belonging to this family were down regulated at least 8-fold in dif1 ovules , and no DUF784 gene detected by the ATH1 array or tested by RT-PCR showed high levels of expression in any tissue outside of the ovule . Furthermore , all four of the DUF784 genes tested were transcribed within the synergid , and most DUF784 genes are significantly downregulated in myb98 ovules . In total , DUF784 genes accounted for ∼50% of the myb98 downregulated genes . The fact that the DUF784 family is both synergid expressed and MYB98 dependent suggests that the pollen tube guidance defect in myb98 ovules may be due to the downregulation of DUF784 genes and implicates the DUF784 family as being important for the development of pollen tube–attraction competence in synergid cells or as potential signaling molecules that are perceived directly by pollen tubes . Despite decades of research in plant sexual reproduction , the genetic mechanisms that underlie the development of the embryo sac and the interactions between the embryo sac and the pollen tube have remained poorly characterized . Through the use of truly genome-wide measures of gene expression that are capable of detecting unannotated genes and recently annotated genes , our analysis uncovered the embryo sac–dependent expression of hundreds of genes not analyzed by recent studies using annotation-based microarrays [22 , 23] . The finding that the majority of embryo sac–dependent genes are functionally uncharacterized underscores the limited extent of our understanding of embryo sac molecular biology . The finding that hundreds of embryo sac–dependent proteins are potentially secreted suggests that the number and complexity of intracellular communications , cell well modifications , and other extracellular events that take place during embryo sac development , fertilization , and the initiation of seed development may be even greater than previously realized . We find that hundreds of related genes , comprising entire families and subfamilies of genes with unknown function , require the mature embryo sac for expression in ovules . The fact that so many paralagous genes have overlapping domains of expression in the embryo sac suggests that there is a high degree of functional redundancy between embryo sac genes . The coexpression of functionally redundant paralogs may explain why genes from these families have not been identified in forward genetic screens for female gametophytic mutants . Furthermore , many of these embryo sac–dependent genes are not expressed at high levels in tissues other than ovules , suggesting that they may be specialized for roles in female reproductive development and function . Future experiments to discover the potentially overlapping functions of embryo sac–dependent gene families will likely be crucial to building a more complete understanding of the genetic mechanisms that underlie plant sexual reproduction .
Seeds for ms-1 ( CS75 , Landsberg background ) , dif1 ( SALK_091193 , Columbia background ) , and myb98–1 ( SALK_020263 , Columbia background ) were obtained from the Arabidopsis thaliana Biological Resources Center . Plants were grown in a growth chamber under long day ( 16 h light/8 h dark ) conditions at 22 °C . Total RNA ( 0 . 5 μg ) from stage 14 ms-1 ovules was reverse transcribed and PCR amplified for 15 cycles using the BD SMART cDNA synthesis kit ( Clontech ) as per the manufacturer's instructions . The cDNA was fragmented and subjected to high-throughput 454 sequencing ( 454 Life Sciences ) [24] . Primer sequence in the 454 reads was masked with Crossmatch ( http://www . phrap . org ) , and each read was aligned to the Arabidopsis genome ( January 2004 release ) using blat with default settings [42] . Reads that had no blat hits were aligned to the genome with blastn ( http://blast . wustl . edu ) ( parameters S = 100 , S2 = 5 , gapS2 = 200 , X = 26 , gapX = 55 , W = 12 , gapW = 18 , gapall , Q = 11 , R = 11 , M = 5 , N = −11 , Z = 3e9 , Y = 3e9 , V = 1e6 , B = 1e6 , hspmax = 1000 , hspsepqmax = 2e5 , topcomboN = 200 , wordmask = seg , maskextra = 10 , hspsepsmax = 2000 ) . For each match found by blat or blastn , more precise exon–exon boundaries were defined by running exalin [43] on the genomic region found by blat or blastn , with an additional 200 nucleotides of flanking sequence on each side . For each read , matches with submaximal exalin scores were discarded , as were matches which contained less than 20 aligning nucleotides , were composed of primarily ( >75% ) of a single nucleotide ( usually A or T ) , or for which the read had less than 80% identity when compared to genomic sequence . Matches with overlapping genomic coordinates and which were not transcribed from opposite strands were grouped together to build consensus “contigs . ” The genomic coordinates of matches to cDNA reads were compared to those of annotated genes ( TAIR release 7 ) . Each gene was assigned a normalized number of reads , where each match to a read was weighted relative to the number of genomic matches that read had ( i . e . , a match to read with a unique genomic match was given a weight of 1 , whereas each match to a read with four equally good genomic matches was give a weight of 0 . 25 ) . Ovules were dissected from approximately one-month-old wild-type ( Columbia ecotype ) , dif1 , and myb98 plants . To obtain mature , unpollinated ovules , stage 12a flowers were emasculated 24 h before collection of ovules . RNA was purified by RNAqueous-micro spin columns ( Ambion ) . Sufficient material was collected to yield at least 2 μg of total RNA ( requiring ∼1600 ovules from ∼40 ovaries ) for each of four biological replicate for each genotype ( i . e . , 12 samples total ) . Preparation of samples for hybridization to tiling arrays was carried out as previously described [44] . PolyA RNA was purified with Oligotex beads ( Qiagen ) , and random hexamer-primed first-strand cDNA was reverse transcribed with Superscript III reverse transcriptase ( Invitrogen ) at 42 °C for 1 h . Second-strand cDNA was synthesized in second-strand reaction buffer ( Invitrogen ) with 40 U of E . coli DNA polymerase I ( New England Biolabs ) , 10 U of E . coli DNA ligase ( New England Biolabs ) , and 2 U of E . coli RNase H ( Epicentre ) at 16 °C for 2 h . cDNA samples were incubated with 10 U of RNase H , 0 . 5 U of RNaseA , and 20 U of RNaseT1 at 37 °C for 20 min and then purified on Qiaquick spin columns ( Qiagen ) . Samples were biotin labeled using the Bioprime system ( Invitrogen ) and concentrated by ethanol precipitation . Samples were hybridized to Genechip Arabidopsis Tiling 1 . 0F arrays ( Affymetrix ) and probe intensities were scanned at the University of Chicago Functional Genomics Center . Expression data was analyzed using the R Project for Statistical Computing and the affy package [45] . Probe intensities were corrected for spatial abnormalities [46] , the perfect match intensities were background corrected with the bg . adjust function of the affy package , and the log2 transformed perfect match intensities were quantile normalized across the 12 arrays . All 25mer probes on the Arabidopsis Tiling 1 . 0F array were matched to the Arabidopsis genome using blastn . Probes perfectly matching the genome more than 30 times were removed from the analysis . For each probe match , a p-value and mean log2 fold change were calculated by a t-test comparing the expression values of the four replicate wild-type expression values against the expression values of the four dif1 replicates . Probes with log2 fold changes less than 1 or p-values greater than 0 . 05 were removed from the analysis . Probe matches that passed these thresholds and that were located within 85 bp of each other were iteratively grouped together to define differentially expressed intervals . Intervals containing less than three passing probe matches were removed from the analysis . The set of dif1 downregulated genomic intervals was compared to genomic locations of existing gene annotations ( TAIR release 7 ) to identify those not mapping to within 50 bp of an annotated gene . Intergenic intervals that overlapped the genomic alignments of ovule 454 cDNA contigs were considered as potential gene fragments . Adjacent contigs were considered to be part of the same gene if they were within 200 bp or overlapped the same cDNA contig . ORFs were predicted by extending the longest ORF within the interval into flanking genomic sequence . In cases where cDNA reads suggested splicing , the splice sites of the cDNA reads were used to guide ORF annotation . The positions of probe matches were compared to the positions of annotated exons ( TAIR7 release ) and to the positions of newly identified protein coding genes . Probe matches that overlapped with more than one gene ( i . e . , regions for which both strands are transcribed ) were removed from the analysis , as were genes having less than three probe matches . For each gene , the mean log2 fold change and corresponding p-value was calculated from a t-test of the wild-type expression values against the mutant expression values ( dif1 or myb98 ) across all probes matching that gene . Genes were considered to be differentially regulated if the p-value was less than 0 . 001 and the log2 change in expression was greater than 1 . The FDR at this threshold was estimated by permuting the groupings of the four wild-type and four dif1 arrays . The set of eight arrays was partitioned into two “balanced” groups of four such that each group of four contained two wild-type arrays and two dif1 arrays . The expression data was reanalyzed for each of 18 possible “balanced” permutations of array groupings , and the FDR estimated as the average number of genes passing the statistical thresholds for the permuted groupings as a percentage of the number passing the thresholds for the actual grouping of arrays . The decision to base FDR estimate on the 18 balanced permutations , rather than on all 35 possible permutation ( including the actual permutation ) , was based on the observation that the large number of genes highly downregulated in dif1 ovules resulted in an unreasonably strict threshold when the nonbalanced permutations were included in the FDR analysis . The list of PFAM domains in TAIR7 proteins was downloaded the TAIR website . For the DEFL , DUF784 , DUF1278 , and thionin-like families , the existing PFAM annotations were found to omit family members . The set of nonpseudogene DEFL genes was taken from Silverstein et al . [31] . For the DUF784 and DUF1278 families , the HMMs downloaded from the PFAM website were used to iteratively search the annotated peptides using HMMer ( version 3 . 2 , http://hmmer . janelia . org/ ) to identify additional family members . Most genes annotated as “plant thionins” were found not to correspond to the “plant thionin” ( PF00321 ) PFAM domain . An HMM was built using all proteins annotated as encoding a “thionin” or “thionin-like” protein , which was then used to search for additional family members . The genes found by this HMMer search are referred to as “thionin-like” in this paper . Gene families were considered to be overrepresented in the set of dif1 downregulated genes if at least five dif1 downregulated genes encoded that domain and the number of down regulated gene encoding that domain was significantly greater than the expected number ( based on the frequency of that domain amongst all proteins ) as determined by a chi-squared test . The presence of putative signal peptides was predicted with SignalP [47] . To allow for the comparison of previously published gene sets from studies using the ATH1 array , we mapped ATH1 probe sets to the most recent gene annotations ( TAIR7 ) by blasting the probe sequences against the annotated cDNA sequences . A probe set was considered to match a gene if at least six of the 11 probes perfectly matched that gene . Previously published sets of SPL/NZZ-dependent genes [23] and DIF1-dependent genes [22] were retabulated based on the mappings of the published ATH1 probe sets to the TAIR7 annotations . In cases for which a single ATH1 probe set mapped to multiple genes , all matched genes were considered as downregulated for purposes of comparison to the whole-genome tiling array data . Ovules were microdissected out of mature ( ∼stage 14 ) ovaries from ms1/ms1 homozygotes ( Landsberg ecotype ) , and RNA was purified using RNeasy columns ( Qiagen ) as per the manufacturer's instructions . Three replicate samples were collected , each of which yielded sufficient RNA ( >7 μg ) to allow unamplified preparation of cRNA for Affymetrix analysis . Preparation of labeled cRNA , hybridizations to ATH1 Genechip arrays ( Affymetrix ) , and scanning of probe-level scores were carried out at the Keck Foundation Biotechnology Resource Laboratory ( Yale University , New Haven , Connecticut ) . Probe level data from the three ovule arrays , along with probe level data from multiple tissues and developmental stages contained in AtGenExpress ( samples ATGE_1 , ATGE_10 , ATGE_12 , ATGE_13 , ATGE_14 , ATGE_15 , ATGE_16 , ATGE_17 , ATGE_2 , ATGE_24 , ATGE_25 , ATGE_26 , ATGE_27 , ATGE_28 , ATGE_29 , ATGE_3 , ATGE_31 , ATGE_32 , ATGE_34 , ATGE_35 , ATGE_36 , ATGE_4 , ATGE_40 , ATGE_41 , ATGE_42 , ATGE_43 , ATGE_5 , ATGE_6 , ATGE_73 , ATGE_76 , ATGE_77 , ATGE_78 , ATGE_79 , ATGE_8 , ATGE_81 , ATGE_82 , ATGE_83 , ATGE_84 , and ATGE_9 ) [37] , as well as seedling and stigma data from Swanson et al . [38] , were normalized using the RMA method [36] as implemented in the affy R package [45] , and the mean expression value was calculated for each probe set in each tissue type . The “ovule-specificity factor” ( Ovsp ) was calculated for each probe set as Ovsp = ( Expovule − Expmean ) /SD , where Expovule is the mean expression in stage 14 ovules , Expmean is the mean log2 expression value across all tissue types ( excluding those that contain stage 12 to stage 14 ovules ) , and SD is the standard deviation of mean expression values across all tissue types ( again excluding tissues containing stage12 or stage 14 ovules ) . ATH1 probe sets were mapped to the most recent gene annotations ( TAIR7 ) by blasting the probe sequences against the annotated cDNA sequences . A probe set was considered to match a gene if at least six of the 11 probes perfectly matched that gene . Samples from different tissues and developmental stages were collected from 1-month-old ms-1 plants , except for anther samples , which were collected from wild-type Landsberg plants . For wild-type Columbia , dif1 , and myb98 ovary samples , stage 12a flowers were emasculated 24 h before ovary tissue was collected . RNA was purified by RNeasy spin columns ( Qiagen ) . For each tissue , 0 . 5 μg of RNA was treated with DNase before first-strand cDNA was reverse transcribed using Protoscript II RT-PCR kit ( New England Biolabs ) and diluted to 100 μl . Each RT-PCR reaction used 2 μl of first-strand cDNA ( or 2 μl of no reverse transcriptase control ) as template in a 30 μl reaction with 0 . 33 μM of each primer . Primer sequences are contained in Table S11 . Genomic sequences upstream of embryo sac–dependent genes were PCR amplified and cloned into pCAMBIA-1381Z upstream of the GUS ORF ( Table S12 ) . Arabidopsis thaliana plants ( Columbia ecotype ) were transformed by the floral dip method [48] . Seeds were grown on MS media containing 25 mg/ml hygromycin for 12 d before seedlings with true leaves were transferred to soil . Stage 12c flowers were emasculated 24 h before ovules were dissected into GUS staining solution ( 50 mM sodium phosphate buffer pH 7 . 0 , 10 mM EDTA , 0 . 1% Triton X-100 , 2 mM potassium ferrocyanide , 2 mM potassium ferricyanide , and 1 mg/ml X-Gluc ) on ice before incubation at 37 °C for 45 min ( 12 h for At5g34885::GUS and At4g24974::GUS ) . Samples were cleared in 20% methanol/4% concentrated HCl at 55 °C for 15 min followed by 60% ethanol/1 . 8 M NaOH at 25 °C for 10 min . Samples were washed with 30% ethanol and 10% ethanol and transferred to 50% glycerol for mounting on slides . Samples were viewed on a Zeiss Axioscope using DIC optics , and images were captured on a Zeiss AxioCam HRc digital camera .
Raw and processed microarray data is available at the National Center for Biotechnology Information ( NCBI ) Gene Expression Omnibus ( http://www . ncbi . nlm . nih . gov/geo/ ) , with the series identifier GSE8392 . Sequences of ovule cDNA reads are available at NCBI dbEST ( http://www . ncbi . nlm . nih . gov/dbEST/ ) , with the identifier numbers 45453167–45702604 .
|
During the sexual reproduction of flowering plants , a pollen tube delivers sperm cells to a specialized group of cells known as the embryo sac , which contains the egg cell . It is known that embryo sacs are active participants in guiding the growth of pollen tubes , in facilitating fertilization , and in initiating seed development . However , the genes responsible for the complex biology of embryo sacs are poorly understood . The authors use two recently developed technologies , whole-genome tiling microarrays and high-throughput cDNA sequencing , to identify hundreds of genes expressed in embryo sacs of Arabidopsis thaliana . Most embryo sac–dependent genes have no known function , and include entire families of related genes that are only expressed in embryo sacs . Furthermore , most embryo sac–dependent genes encode small proteins that are potentially secreted from their cells of origin , suggesting that they may act as intracellular signals or to modify the extracellular matrix during fertilization or embryo sac development . These results illustrate the extent to which our understanding of plant sexual reproduction is limited and identifies hundreds of candidate genes for future studies investigating the molecular biology of the embryo sac .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"plants",
"genetics",
"and",
"genomics",
"arabidopsis",
"(thale",
"cress)"
] |
2007
|
Genome-Wide Expression Profiling of the Arabidopsis Female Gametophyte Identifies Families of Small, Secreted Proteins
|
RNA polymerase II synthesizes a diverse set of transcripts including both protein-coding and non-coding RNAs . One major difference between these two classes of transcripts is the mechanism of termination . Messenger RNA transcripts terminate downstream of the coding region in a process that is coupled to cleavage and polyadenylation reactions . Non-coding transcripts like Saccharomyces cerevisiae snoRNAs terminate in a process that requires the RNA–binding proteins Nrd1 , Nab3 , and Sen1 . We report here the transcriptome-wide distribution of these termination factors . These data sets derived from in vivo protein–RNA cross-linking provide high-resolution definition of non-poly ( A ) terminators , identify novel genes regulated by attenuation of nascent transcripts close to the promoter , and demonstrate the widespread occurrence of Nrd1-bound 3′ antisense transcripts on genes that are poorly expressed . In addition , we show that Sen1 does not cross-link efficiently to many expected non-coding RNAs but does cross-link to the 3′ end of most pre–mRNA transcripts , suggesting an extensive role in mRNA 3′ end formation and/or termination .
Early in each transcription cycle RNA polymerase II ( Pol II ) can follow one of two paths; terminate early through the Nrd1-Nab3-Sen1 pathway or continue on to form longer , potentially coding transcripts [1] , [2] . Yeast Nrd1 , Nab3 and Sen1 are part of a complex [3] that interacts both with the phosphorylated Pol II C-terminal domain ( CTD ) [4]–[6] , and with specific sequences in the nascent transcript [7] , [8] . If this set of protein-protein and protein-RNA contacts is sufficient , the Nrd1-Nab3-Sen1 complex directs Pol II termination and couples this to processing of the nascent transcript by TRAMP and the nuclear exosome [9] , [10] . In addition to directing termination of snoRNAs and cryptic unstable transcripts ( CUTs ) the Nrd1-Nab3-Sen1 pathway directs premature termination of several pre-mRNA transcripts including NRD1 , IMD2 , URA2 , URA8 , and ADE12 [11]–[15] . The mechanism by which the Nrd1-Nab3-Sen1 complex leads to Pol II termination is unknown but involves recognition of specific terminator sequences by Nrd1 and Nab3 [7] , [8] , [16] and the putative helicase activity of Sen1 [16] , [17] . In this study we have used a recently developed in vivo cross-linking approach [18] to derive high-resolution transcriptome-wide maps of binding sites for Nrd1 , Nab3 , Sen1 , and the Pol II subunit Rpb2 . This approach yields a more precise picture of known Nrd1 and Nab3 binding sites on snoRNA , CUT and mRNA targets and reveals a set of previously unknown Nrd1 binding sites both on the 5′ ends of mRNAs and on 3′ antisense transcripts . Surprisingly , Sen1 does not cross-link at many of these Nrd1-Nab3 binding sites . Instead , we observe Sen1 cross-linking on mRNA transcripts , particularly at the 3′ end suggesting a potential role for Sen1 in mRNA 3′ end formation through the cleavage and polyadenylation pathway .
Previous studies have shown that Nrd1 and Nab3 are present on chromatin and localized to genes that are regulated by Nrd1 and Nab3 [4] , [6] , [9] , [11] , [23] . Because the Nrd1 complex interacts both with the nascent transcript and with Pol II , it is not clear which interaction is primarily responsible for chromatin binding . We have analyzed Nrd1 and Nab3 interactions with both chromatin and RNA . Figure 1 shows a comparison of ChIP-chip of Nrd1 compared to Nrd1 and Nab3 PAR-CLIP . Two tracks displaying Pol II subunits Rpb2 and Rpb3 are shown for reference , as they are both part of the catalytic core and thus are expected to co-localize . All five data sets were obtained from cells growing in log phase . Nrd1 and Pol II co-localize by ChIP at both the RPS5 and SNR3 genes ( Figure 1 ) . Nab3 ChIPs similarly to Nrd1 ( not shown ) . In contrast , Nrd1 and Nab3 cross-link to RNA in a subset of the major Nrd1 ChIP peaks . For example , in Figure 1 we observe efficient cross-linking of Nrd1 and Nab3 to SNR3 transcripts but not to RPS5 transcripts . Genomewide , the Nab3 RNA cross-linking pattern is nearly identical to Nrd1 ( Wilcoxon rank sum p<10−8 ) . Similar specificity of Nrd1 RNA cross-linking is seen at other highly expressed genes . Among the top 100 Nrd1 ChIP peaks determined using CisGenome [24] are ten other ribosomal protein genes , none of which are among the top 100 Nrd1 PAR-CLIP peaks ( not shown ) . We conclude that Nrd1 , Nab3 , and presumably Sen1 are able to enter the early Pol II elongation complex independent of RNA binding and monitor the nascent RNA for appropriate binding sites . Figure 2A shows the logos [25] of motifs in a representative MEME [26] trial using sequences of 40 nt centered on the most prominent RNA cross-link ( T to C ) sites . Nrd1 cross-links reveal a consensus sequence UGUAG with an E-value of 1 . 8e−25 where the underlined U is the most frequently cross-linked residue . The top-scoring motif in the Nab3 pool is a consensus sequence GNUCUUGU . The E-value under the same conditions as the Nrd1 analysis is much higher ( 6 . 7e3 ) indicating that this motif is not nearly as constrained . These motifs are nearly identical to the GUA[A/G] and UCUU sequences that we previously identified using genetic and biochemical approaches [7] , [8] and contain many of the over-represented 4 mer sequences in sequences that cross-link to Nrd1 and Nab3 by the CRAC protocol with 254 nm UV irradiation [20] . Our Nab3 motif is very similar to the UUCUUGUW motif identified by microarray analysis of RNA co-purified with Nrd1 in the absence of cross-linking and under non-denaturing conditions [27] . No conserved motifs were observed for Sen1 and Rpb2 , consistent with their presumed roles as enzymes that interact non-specifically with RNA . About 50% of the Nrd1 and Nab3 cross-linked regions overlap . This is not surprising given that Nrd1 and Nab3 are known to dimerize in vivo and in vitro [7] , [28] . The second most significant motif , as determined in MEME , in the Nrd1 set corresponded to the top rated motif in the Nab3 analysis and vice versa , again suggesting that Nrd1 and Nab3 binding sites are located close to each other . In Figure 2B we show that Nrd1 cross-linking sites are clustered . The 50 Nrd1 cross-linked sites with the most reads ( as determined by MochiView [29] ) were used as an anchor in this plot . The number of cross-links observed as a function of the distance from the top 50 cross-linked sites is increased in a window approximately 30 nt upstream and downstream of the central Nrd1 cross-link . No similar clustering of Nab3 cross-linking sites was observed indicating that most Nab3 binding regions do not contain multiple motifs . Together , these results suggest that these strong Nrd1 binding sites have been selected to bind multiple Nrd1 proteins , a result consistent with our in vitro binding experiments suggesting that the Nrd1-Nab3 complex binds cooperatively to RNA targets [7] . For both Nrd1 and Nab3 data sets binding sites on snoRNA transcripts were among the most extensively cross-linked ( Tables S2 and S3 ) . At most snoRNAs we observed a peak of Nrd1 binding downstream of the mature 3′ end of the RNA consistent with the previously demonstrated role in termination [16] . Figure 3A and 3B shows the distribution of cross-linked sequences downstream of SNR3 and SNR13 . Nrd1 binds predominantly downstream of the mature RNA , while Rpb2 cross-links across the transcript . Surprisingly , Sen1 cross-links efficiently to some snoRNA transcripts like snR3 , but not to others , like snR13 . This is an unexpected result considering previous experiments showing that at both of these snoRNA genes Pol II reads through the terminator in a sen1 mutant at the non-permissive temperature [16] and both genes have Sen1 present on chromatin as determined by ChIP [30] , [31] . We have previously shown that Nrd1 regulates its own expression by binding to sites in the 5′ end of its mRNA and directing premature termination . This attenuation mechanism requires the function of Nrd1 , Nab3 and Sen1 [16] . The Nrd1-Nab3-Sen1 pathway is also required for the formation of attenuated transcripts from the IMD2 promoter [12] . We were therefore surprised that Sen1 does not crosslink as efficiently to NRD1 or IMD2 mRNA as it does to other mRNAs , for example CCW12 ( Figure 3C and 3D and Figure S2 ) . Of the top 300 Nrd1 and Nab3 peaks , 93 and 54 , respectively , overlap with CUTs ( out of 925 annotated CUTS [32] , [33] ) . By contrast , only 23 ( Nrd1 ) and 8 ( Nab3 ) peaks overlap with SUTs ( stable unannotated transcripts , out of 847 annotated SUTs [32] , [33] ) . This preference for binding unstable transcripts is consistent with Nrd1-Nab3 binding being a key feature that distinguishes between CUTs and SUTs [20] . In Figure S3 we show that Nrd1 binds to several known CUTs . The Nrd1 cross-linking sites on CUT transcripts tend to be located toward the 5′ end , a position in which Nrd1 and Nab3 may direct the observed range of termination sites downstream of these binding sites [9] , [10] , [34] . The observation that some CUTs are not bound by Nrd1 is somewhat surprising . In Figure S3 we observe that some CUTs and SUTs are not expressed under our experimental conditions as shown by the lack of cross-linking to Rpb2 . Nrd1 , Nab3 , and Sen1 also bind to RNA sequences derived from protein-coding genes . Cross-linked sequence reads were sorted into ten bins covering each coding region in order to display genes of different length on the same scale . Figure 4A–4C shows the distribution of Nrd1 plus-strand and minus-strand reads on genes ranked by expression level . Figure 4D–4F shows a similar distribution of Sen1 reads . Highly expressed genes have a peak of Nrd1 sense-strand reads derived from the 5′ end . Genes with the lowest level of expression show a peak of Nrd1 cross-linked antisense reads at the 3′ end . For Sen1 we observe the largest number of reads on the 3′ end of the most abundantly transcribed genes . Nrd1 has been implicated in several regulatory mechanisms involving binding to sequences in the 5′ end of transcripts . Autoregulation of Nrd1 expression by attenuation is one example . Another form of regulation involving Nrd1 and Nab3 binding near the 5′ end of transcripts is alternative transcription start site ( TSS ) selection on genes involved in nucleotide biosynthesis . IMD2 , URA2 , URA8 and ADE12 have been shown to use alternative transcription start sites ( TSSs ) in response to nucleotide availability . For IMD2 and URA2 the upstream starts used in the presence of sufficient NTP result in the elongation complex passing through a Nrd1-Nab3 terminator thus reducing transcription of mRNA [12] , [13] , [15] . We observe peaks of Nrd1 binding on all of these transcripts ( not shown ) . In addition , we see similar binding on other genes involved in nucleotide metabolism including HPT1 , GUA1 and ADE17 ( Figure S4 ) . For each of these genes multiple TSSs have been reported and are located upstream and downstream of the Nrd1-Nab3 binding region , consistent with regulation by TSS selection . Figure 5 shows two additional genes , PCF11 and RPB10 that have 5′ Nrd1 binding peaks . In each case the peak of binding is downstream of at least one mRNA 5′ end suggesting that the gene may be regulated by premature termination . Increased levels of RNA from cells in which Nrd1 levels were depleted confirm that PCF11 and RPB10 are negatively regulated by Nrd1 ( Figure 5C and 5D ) . In the case of PCF11 this regulation is particularly interesting because PCF11 encodes a protein with a similar termination function to Nrd1 . While Pcf11 has primarily been associated with termination of mRNAs it also plays a role in termination of non-poly ( A ) transcripts 30 , 35 . Nrd1 and Pcf11 have been proposed to compete for recruitment to the transcription complex [36] , [37] and the ability of Nrd1 to regulate Pcf11 expression may play a role in balancing this competition . For example , conditions that reduce Nrd1 protein levels would be expected to lead to compensatory increases in both NRD1 and PCF11 expression . Nrd1 binding also regulates expression of RPB10 , a gene encoding an RNA polymerase subunit that is common to all three nuclear RNA polymerases . This observation suggests a possible role for the Nrd1 pathway in regulating expression of the transcription machinery . In addition to sense-strand binding , we observe a large number of Nrd1 reads derived from RNAs that are anti-sense to annotated genes . Figure 4 shows that these reads are concentrated at the 3′ end of genes that are not heavily transcribed in the sense direction . Figure 6 shows Nrd1 and Rpb2 binding to antisense transcripts at the 3′ ends of the YKL151C and USA1 genes . We do not observe abundant Sen1 cross-linking to the antisense transcript despite the fact that it cross-links efficiently to the sense transcript of the downstream gene . In Figure S5 we show that YKL151C encodes a 3′ antisense CUT . We think it is quite likely that these antisense transcripts originate from divergent transcription from the downstream promoters [32] , [33] and may be involved in suppressing transcription in the sense direction of YKL151C and USA1 . Our cross-linking experiments show that Sen1 cross-links to abundantly transcribed mRNAs . In Figure 7E we show examples of Sen1 binding to transcripts derived from the heavily transcribed RPL28 , RPS13 , RPL30 and PMA1 genes . Several arguments suggest that this Sen1 cross-linking occurs on nascent transcripts . First , Sen1 and Rpb2 both cross-link in clusters that are spaced along the transcript in coincident peaks . A similar clustering of Pol II-associated RNA has recently been attributed to pausing of Pol II as nucleosomes are removed from the path of the elongating polymerase [38] . Such clustering would not be expected on mature transcripts . Second , although upstream cross-links are less abundant on ribosomal gene transcripts , there is some cross-linking to intron sequences that are enriched on nascent transcripts . Taken together , these results suggest preferential binding to nascent RNA but we cannot rule out binding to unprocessed precursors that have been terminated and released from the template . We note that Sen1 peaks are stronger toward the 3′ end , especially on the ribosomal protein transcripts . This is confirmed in Figure 7D by plotting all Sen1 and Rpb2 reads with respect to the 3′ end of transcripts derived from the most heavily transcribed genes [39] . Interestingly , the peak of Sen1 is broadly distributed over the 75 nt before the polyadenylation site ( pA ) with few if any reads extending beyond this cleavage site . In Figure 8 we show that this downstream Sen1 cross-linking peak does not correspond to Nrd1 or Nab3 cross-linking sites . A distinct peak of Pol II is observed between 25 and 50 nt downstream of the pA site . Clearly , Pol II continues to transcribe beyond the pA site but we see little evidence for Pol II more than a few hundred bases further downstream ( Figure 7E ) . In Figure 8 we have also compared our Rpb2 cross-linking data set with the Net-seq data set derived from RNA non-covalently associated with affinity purified Pol II [38] . While we see a very similar pattern of reads within coding regions , we notice a difference in the pattern of Pol II downstream of the pA site . The downstream peak of Pol II that we observe in our PAR-CLIP data is often missing in the Net-seq data ( Figure 8 ) .
Another unexpected observation is that many poorly expressed protein-coding genes express Nrd1-bound 3′ antisense sequences . Previous studies in yeast have identified regulatory antisense transcripts for IME4 , PHO5 , PHO84 and the Ty1 retrotransposon [45]–[49] . In the case of PHO84 and Ty1 , antisense transcripts appear to act in trans [47] , [49] , although in the case of PHO84 cis-suppression is also observed [48] . In Figure 6 we show several genes with Nrd1 cross-linked anti-sense RNA peaks localized to the 3′ coding region . In each case the gene exhibiting 3′ antisense transcripts is poorly transcribed in the sense direction as determined by RNA sequencing [39] and the downstream gene is highly expressed in glucose-containing media [50] . These antisense transcripts likely result from bi-directional transcription from the downstream promoter [32] , [33] . The Rpd3s histone deacetylase complex has been shown to repress antisense initiation at many promoters [38] and it is possible that the Nrd1-Nab3 non-poly ( A ) termination pathway prevents elongation of those antisense transcripts that lack or escape Rpd3s control . The question of whether these antisense transcripts are regulatory remains to be answered , but we note that each of these Nrd1 antisense peaks correlates with Rpb2 cross-linking sites but does not show efficient Sen1 binding . We propose that Nrd1 pauses Pol II at these sites , preventing sense strand transcription either through transcription interference or by establishment of chromatin marks that are inappropriate for transcription of the sense strand . The distribution of Sen1 is surprising on two counts . First , we failed to observe efficient cross-linking on some transcripts that had previously been shown to depend on Sen1 function for proper termination . SNR13 transcripts normally terminate just downstream of the Nrd1 binding site but in a sen1 mutant Pol II reads through into the downstream TRS31 gene [16] and Sen1 ChIPs to this downstream region [31] . Similarly , Nrd1 autoregulation is disrupted in a sen1 mutant with steady-state levels of Nrd1 mRNA increasing about 10-fold [11] , [16] . The failure of Sen1 to efficiently crosslink to these transcripts can be explained in several ways . First , the RNA at these sites may interact with Sen1 in a manner that prevents close apposition of Sen1 amino acid side chains with 4SU residues in the bound RNA . A second possibility is that Sen1 may play a structural role , stabilizing the Nrd1-Nab3-Sen1 complex in a manner that does not require the helicase activity . In this model the sen1 mutation may alter the structure of the complex leading to disruption of Nrd1 and/or Nab3 termination function . A third possibility is that Sen1 may be required for expression of another factor that is required for termination of non-poly ( A ) transcripts . Finally , the low amount of Sen1 cross-linking may indicate that low levels of Sen1 are sufficient for proper termination at some genes . Future experiments must be directed at understanding the role of Sen1 in termination of snoRNA and attenuated Pol II transcipts . A second unexpected result from the Sen1 cross-linking data set is the widespread distribution of Sen1 on mRNA transcripts . A role for Sen1 in mRNA 3′ end formation has been suggested by previous experiments . Sen1 interacts with Glc7 , a protein phosphatase that is part of the CPF complex [51] , [52] . In addition , sen1 mutants display a weak read through phenotype on some pA terminators [53] , [54] , [55] but not others [30] . Although a genome-wide survey of Pol II distribution in a sen1-E1597K mutant did not detect widespread read through of pA terminators [31] some read through was observed and our data shows that several of those genes including RPL43B , RPS28A , RPL36B , and SOD1 display prominent Sen1 binding near their pA site ( not shown ) . Our data shows that Sen1 cross-linking occurs all along mRNA transcripts , peaking at the 3′ end ( Figure 7 ) . Sen1 interacts with the Pol II CTD [5] but whether this interaction is direct is not known . Based on the increased cross-linking toward the 3′ end of genes it is possible that Sen1 interacts with the Ser2 phosphorylated form of Pol II or another protein that binds to this phosphorylated form of the CTD . The failure of Sen1 to cross-link downstream of the pA site would seem to rule out a rho-like termination model in which Sen1 translocates along the transcript facilitating termination upon reaching the paused downstream Pol II . Recently Mischo et al . [54] have shown that Sen1 helicase activity is required to remove R loops that form at the 3′ end of some transcripts . This proposal is based in part on the identification of DNA∶RNA hybrids downstream of the pA site [54] . Sen1 helicase activity could act to remove RNA from the DNA∶RNA hybrid and expose RNA downstream of the pA site for degradation by the 5′-3′ exonuclease Rat1 . Our data argue , however , that Sen1 acts upstream but not downstream of the pA site . We can clearly observe cross-linked Pol II downstream of the pA site but there is no corresponding Sen1 cross-linking in this region . Sen1 could act upstream of the cleavage site to remove RNA from R loops and allow access of the cleavage/polyadenylation machinery and subsequently the 5′-3′ exonuclease Rat1 . This model is consistent with previous experiments showing a synergistic effect of sen1 and rat1 mutations [53] , [55] . Our data also suggest that the downstream peak of Pol II represents the Pol II termination complex . Kinetic modeling of yeast Pol II transcription suggests a termination time of about one minute [56] . This is greater than the amount of time needed to transcribe many yeast genes , suggesting that termination is the rate-limiting step for formation of some mRNAs . We suggest that the peak of Rpb2 cross-linking we observe downstream of the pA site of heavily transcribed genes ( Figure 7 ) represents this rate-limiting Pol II elongation complex that is in the act of terminating . We note that this downstream peak of Pol II is not as prominent in the NET-seq data ( Figure 8; [38] ) . The NET-seq data is obtained by affinity purifying Pol II elongation complexes from chromatin after digestion with DNase I . We propose that the downstream termination complex is sensitive to DNase I digestion because of an allosteric change that takes place as the elongation complex passes through the pA site [57] . Thus , the RNA is lost from these complexes and is under-represented in the NET-seq data .
The genomic NRD1 , NAB3 , SEN1 and RPB2 genes in BY4733 were tagged with both 6His and biotin tags ( HTB ) [19] . The resulting strains expressed proteins with a slightly higher molecular weight that could be enriched on streptavidin beads ( Figure S6 ) . These strains displayed no abnormal growth phenotypes . A TAP-tagged NRD1 yeast strain was used for chromatin immunoprecipitation . For the first two data sets , yeast cells expressing HTB tagged Nrd1 or Nab3 were grown at 30°C in 2 L of synthetic complete ( SC-URA ) medium supplemented with 2% dextrose , 120 µM Uracil , 0 . 01 µM Biotin from OD600∼0 . 1 to mid-exponential phase ( OD600∼0 . 5 ) . 4SU was added to a final concentration of 300 µM and growth continued at 30°C until OD600∼1 . 5 . Addition of 4SU had no effect on the growth rate of yeast during the time course of the experiment . Cells were harvested by centrifugation and the cell pellet was re-suspended in 60 ml of ice-cold water , separated into two 30 ml aliquots and placed in 145×20 mm sterile tissue culture dishes kept on ice . Cells were irradiated on ice with 365 nm UV light ( 0 . 15 J/cm2 ) in a Stratalinker 2400 ( Stratagene ) , three times for 10 minutes each with shaking between irradiations . Cells were pooled , centrifuged , and the cell pellet was re-suspended in 5 ml of buffer-1 ( 8 M urea , 300 mM NaCl , 0 . 5% Nonidet P-40 , 50 mM sodium phosphate , 50 mM Tris-HCl , pH 8 . 0 , and EDTA-free protease inhibitor mix for His-Tag sequences ( RPI ) ) then frozen in droplets in liquid nitrogen . Cell droplets were kept in −80°C until processed as described below . In the analysis of the initial Nrd1 and Nab3 data sets we observed binding to a number of RNAs derived from stress-induced genes . The RNA in these early experiments was obtained from cells that were irradiated after centrifugation and re-suspension in ice-cold water [21] , a procedure that is likely to induce a stress response . To eliminate this possibility we developed a second technique to irradiate cells growing in liquid media at 30°C . Two liters of growing cells were placed in a two-liter beaker on a magnetic stirrer and irradiated from a distance of 10 cm for 10 min with a UV Power–Shot 1100 Lamp . This lamp delivers 1 W/cm2 primarily in the in the 300–400 nM range with a peak at 365 nm . To eliminate shorter wavelength light the beaker was covered with a Pyrex baking dish . Cells irradiated in this manner were processed as described below . Protein purification was based on a previously published protocol [58] . Cell droplets were lysed in liquid nitrogen using a Spex SamplePrep 6870 freezer mill with 10 cycles of one minute of breakage and two minutes of cooling , at a frequency setting of 15 cps . Lysates were thawed at room temperature , resuspended in 5 ml of buffer-1 then sonicated using a 1/8″ microprobe tip of Branson sonifer cell disruptor Model 250/450 . Sonication was performed three times at 50% power for 5 seconds with 30 second intervals at room temperature . Cell lysates were cleared by centrifugation at 40 , 000 rpm in Beckman L-80 ultracentrifuge at room temperature for 30 minutes using Ti 70 . 1 rotor . Cleared lysates were incubated with Ni-NTA agarose ( QIAGEN , 500 µl slurry pre-equilibrated in buffer-1 ) for 3 hours at room temperature . Ni-NTA agarose was then washed in 5 ml of buffer-1; 5 ml of buffer-1 , pH 6 . 3; and 5 ml of buffer-1 , pH 6 . 3 , +10 mM imidazole . Proteins were eluted in 8 ml of buffer-2 ( 8 M urea , 200 mM NaCl , 2% SDS , 50 mM sodium phosphate , 10 mM EDTA , 100 mM Tris-HCl , pH 4 . 3 , and EDTA-free protease inhibitor mix for His-Tag sequences ( RPI ) ) . The pH of the eluate was neutralized and loaded onto streptavidin magnetic beads ( New England Biolabs ) . A 200 µl slurry of beads was pre-equilibrated in buffer-3 ( 8 M urea , 200 mM NaCl , 0 . 2% SDS , 100 mM Tris-HCl , pH 8 . 0 , and EDTA-free protease inhibitor mix for His-Tag sequences ) . After incubation overnight at room temperature the streptavidin magnetic beads were washed in 3×500 µl of buffer-3 , 3×500 µl of buffer-3 with 2% SDS , 3×500 µl of buffer-3 without SDS , and then 3×500 µl of T1 ribonuclease buffer ( 150 mM KCl , 2 mM EDTA , 0 . 5 mM DTT , 50 mM Tris-HCl , pH 7 . 4 , and EDTA-free protease inhibitor mix for His-Tag sequences ( RPI Crop . ) ) . The streptavidin magnetic beads were resuspended in 0 . 5 ml of T1 buffer before RNase T1 ( Fermentas ) was added to obtain a final concentration of 40 U/ml and the bead suspension was incubated at room temperature for 15 minutes . Beads were washed three times with 500 µl of T1 wash buffer ( 500 mM KCl , 0 . 05% NP40 , 0 . 5 mM DTT , 50 mM Tris-HCl , pH 7 . 8 , and EDTA-free protease inhibitor mix for His-Tag sequences ( RPI Corp . ) ) and three times with 500 µl of polynucleotide kinase ( PNK ) buffer ( 50 mM NaCl , 10 mM MgCl2 , 5 mM DTT , 50 mM Tris-HCl , pH 7 . 4 , and EDTA-free protease inhibitor mix for His-Tag sequences ) . Beads were resuspended in 160 µl of PNK buffer before Thermosensitive Alkaline Phosphate ( TSAP ) ( Promega ) was added to obtain a final concentration of 0 . 15 U/µl , and SuperRNase inhibitor ( Ambion ) was added to obtain a final concentration of 1 U/µl . The bead suspension was incubated for 30 minutes at 37°C then was washed once with 500 µl of buffer-3 without SDS , and 3 times with 500 µl of PNK buffer . Streptavidin magnetic beads from the previous step were resuspended in 200 µl of PNK buffer then γ-32P-ATP ( MP Biomedicals ) was added to obtain a final concentration of 0 . 5 µCi/µl and T4 PNK ( New England Biolabs ) was added to obtain a final concentration of 1 U/µl . The bead suspension was incubated at 37°C for 30 minutes before non-radioactive ATP was added to obtain a final concentration of 100 µM , the incubation was continued for 10 minutes at 37°C . The bead suspension was washed 4 times with 650 µl of T4 RNA Ligase2 truncated ( Rnl2 ( 1–249 ) ) buffer ( 2 mM MgCl2 , 1 mM DTT , 50 mM Tris-HCl , pH 7 . 5 ) ( New England Biolabs ) . Streptavidin magnetic beads from the previous step were resuspended in 19 µl of Rnl2 ( 1–249 ) buffer combined with 19 µl of 50% Polyethylene Glycol 8000 ( PEG 8000 ) ( Promega ) . To the bead suspension was added 2 µl of 100 µM adenylated 3′ adapter oligodeoxynucleotide ( AppATCTCGTATGCCGTCTTCTGCTTGTC; IDT ) , 0 . 4 µl of 1 M MgCl2 , 2 . 5 µl of Rnl2 ( 1–249 ) ( 200 U/µl ) ( New England Biolabs ) , 1 . 25 µl of RNase inhibitor ( 40 U/µl ) ( Invitrogen ) . The bead suspension was incubated at room temperature for 4 hours then washed once with 500 µl of buffer-3 without SDS , and 3 times with 500 µl of PNK buffer . The washed streptavidin magnetic beads were re-suspended in 38 µl of PNK buffer , then to the bead suspension was added 2 µl 100 µM 5′ adapter oligonucleotide ( GUUCAGAGUUCUACAGUCCGACGAUC ) , 1 . 25 µl of RNase inhibitor ( 40 U/µl ) , 2 . 5 µl of T4 RNA ligase ( 10 U/µl ) ( Fermentas ) , 3 µl of 10 mM ATP ( New England Biolabs ) . The bead suspension was incubated overnight at 16°C then washed once with 500 µl of buffer-3 without SDS , and 3 times with 500 µl of Protease K buffer ( 75 mM NaCl , 6 . 25 mM EDTA , 1% SDS , 50 mM Tris-HCl , pH 7 . 5 ) . The streptavidin magnetic beads were resuspended in 200 µl of protease K buffer follow by the addition of Protease K ( Ambion ) to the final concentration of 1 . 2 mg/ml . After incubation at 55°C for 30 minutes , the supernatant was transferred to a new tube and another 200 µl of protease K buffer was added to the streptavidin magnetic beads . The incubation was continued for 5 minutes at 55°C before the supernatant was combined and the RNA was recovered by acidic phenol/chloroform extraction followed by a chloroform extraction and an ethanol precipitation . The pellet was dried and then dissolved in 12 µl of DPEC water . The recovered RNA was divided into 3×4 µl aliquots and kept in −80°C until further preparation . The recovered RNA was used to synthesize a cDNA library . Time course PCR amplification was performed in order to determine the optimum number of cycles for amplifying the cDNA library . One aliquot of recovered RNA was taken out of −80°C to thaw on ice , then reverse transcription oligonucleotide ( CAAGCAGAAGACGGCATACGA ) was added to the final concentration of 5 µM . The mixture was briefly centrifuged , heated at 70°C on a preset thermal cycler for 2 minutes and the tube was then placed on ice . To the iced tube containing the primer-annealed template RNA was added 4 µl of the premixed reverse transcription reaction mixture containing 2 µl of 5 X first strand buffer ( 375 mM KCl , 15 mM MgCl2 , 250 mM Tris-HCL , pH 8 . 3 ) ( Invitrogen ) , 0 . 5 µl of 12 . 5 mM dNTP mix ( Fermentas ) , 1 µl of 0 . 1 M DTT ( Invitrogen ) , and 0 . 5 µl of RNase inhibitor ( 40 U/µl ) ( Invitrogen ) . The tube was heated on the preset thermal cycler at 48°C for 3 minutes then Superscript II reverse transcriptase ( Invitrogen ) was added to the final concentration of 20 U/µl , the incubation was continued for 1 hour on the preset thermal cycler at 44°C . Time course PCR was carried out on a 50 µl scale using cDNA from the previous step and Phusion DNA polymerase ( Finnzymes ) ( 10 µl of the cDNA , 0 . 5 µl of 25 mM dNTP Mix , 10 µl of 5 x Phusion HF buffer ( Finnzymes ) , 1 µl of 100 µM primers , 0 . 5 µl of 2 U/µl Phusion DNA polymerase ) . PCR cycle conditions of 10 s at 98°C , 30 s at 60°C , 15 s at 72°C were used . Aliquots of 6 µl were removed every other cycle starting with cycle number 14 by temporarily pausing the PCR cycle at the end of the 72°C step . The maximum cycle of the PCR was set at 30 cycles . PCR aliquots were analyzed on a 6% Novex TBE PAGE gel ( Invitrogen ) and the optimal cycle number for cDNA amplification was chosen as five cycles prior to reaching the saturation level of PCR amplification . The optimal cycle number PCR was performed on a 50 µl scale using cDNA prepared from another aliquot of recovered RNA ( 4 µl ) . The amplified cDNA product was separated on a 6% Novex TBE PAGE gel ( Invitrogen ) and the band of interest was excised and eluted using 1 x gel elution buffer ( Illumina ) . RNA fragments with both adapters produce PCR fragments that span a range of 100–150 nt . The 5′-adapter-3′-adapter products without inserts may be detected after amplification of the cDNA as an additional PCR band at around 75 nt in length . The eluted DNA from gel extraction was ethanol precipitated followed by DNA analysis using Agilent 2100 Bioanalyzer . DNA was sequenced using an Illumina GAII sequencer ( University of California , Riverside ) . Sequences were trimmed using the function trimLRPatterns from the ShortRead package in Bioconductor [59] . Reads were aligned to the Sacchromyces cerevisiae genome using the short read aligner Bowtie [60] . The bowtie alignment allowed the read to have up to 1 mismatch and align once to the genome . Reads that aligned more than once to the genome will be discussed elsewhere [22] . All figures were made using the Mochiview genome browser [29] . The processed wig files for each dataset , along with fasta file of the yeast genome for alignment , can be downloaded from Gene Expression Omnibus [61] under the series number GSE31764 ( http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE31764 ) . Cells containing a TAP-tagged NRD1 gene were grown in yeast extract peptone dextrose ( YPD ) to an absorbance of about 0 . 8 and cross-linked with formaldehyde for 20 min at room temperature and processed for chromatin immunoprecipitation according to standard protocols [62] . ChIP and Input DNA were first amplified using linker adapted random primer and sequenase 2 and further amplified and labeled using PCR and Cy3- and Cy5- CTP , respectively . Cy dye labeled ChIP and Input DNA was combined with Agilent aCGH/ChIP hybridization buffer containing 2x Hi-RPM Hyb Buffer , Agilent blocking agent and Human cot-1 DNA and hybridized to Agilent 244K yeast tiling array ( G4491A ) for 40 hours on MAUI hybridization system with constant mixing . The hybridized array was washed using Agilent aCGH/ChIP-on-chip array washing buffer kit ( 5188–5266 ) and scanned on Axon GenePix Scanner ( GenePix A4300 ) and the raw data extracted using GenePix Pro 6 . 0 . Preliminary analysis of the ChIP data was carried out using CisGenome to identify ChIP enriched regions . Total RNA was extracted from yeast with hot acid phenol and run on a 1% denaturing formaldehyde MOPS agarose gel , visualized by ethidium bromide and Northern blotted as previously described [11] . Samples with clear rRNA bands and no visible degradation were analyzed by quantitative real-time RT-PCR . RNA was treated with turbo-DNA-free ( Ambion ) according to the manufacturer's instructions for the most stringent treatment . Reverse transcription was performed using the iScript cDNA Synthesis Kit ( BioRad ) . Real-time PCR was performed in triplicate 20 µl reactions on a CFX96 Real-time PCR detection system ( BioRad ) using iQ SYBR green supermix ( BioRad ) according to the manufacturer's instructions . Data from at least two replicate experiments were pooled using the Gene Study feature of the CFX96 real-time software , which normalizes for fluorescence intensity differences between plates . Expression was normalized to both ACT1 and 18S ribosomal RNA and the ratios were graphed relative to the wild-type control sample , which was set to 1 for each gene . Error bars represent the positive and negative range of the standard error of the mean . Tet-promoter yeast strains ( Open Biosystems ) and a control strain lacking a tet-responsive promoter were seeded to an initial A600 of 0 . 06 in YEPD . Cultures were grown at 30° for 2 . 5 hrs prior to the addition of doxycycline at a final concentration of 10 µg/ml and were grown an additional five hrs before collection .
|
Transcription in eukaryotes is widespread including both protein-coding transcripts and an increasing number of non-coding RNAs . Here we present the results of transcriptome-wide mapping of a set of yeast RNA–binding proteins that control expression of some protein-coding genes and a number of novel non-coding RNAs . The yeast Nrd1-Nab3-Sen1 pathway is required for termination and exosome-mediated processing of non-coding RNA polymerase II transcripts . Our data show that these components bind unexpected targets including a large number of antisense transcripts originating from the 3′ end of genes that are poorly expressed in the sense direction . We also show that Sen1 helicase , involved in termination of non-coding RNAs , is also present at the 3′ end of mRNAs , suggesting a more fundamental role in transcription termination . Mis-regulation of transcription is the underlying cause of many disease states . For example , mutation of the human Sen1 gene , senataxin , causes a range of neurodegenerative disorders . Understanding the roles of yeast RNA–binding proteins in controlling termination of coding and non-coding RNAs will be useful in deciphering the mechanism of these proteins in human cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome",
"expression",
"analysis",
"molecular",
"cell",
"biology",
"rna",
"processing",
"gene",
"regulation",
"genetics",
"gene",
"expression",
"molecular",
"genetics",
"biology",
"genomics",
"molecular",
"biology",
"genetics",
"and",
"genomics",
"dna",
"transcription"
] |
2011
|
Transcriptome-Wide Binding Sites for Components of the Saccharomyces cerevisiae Non-Poly(A) Termination Pathway: Nrd1, Nab3, and Sen1
|
Expansion of DNA trinucleotide repeats causes at least 15 hereditary neurological diseases , and these repeats also undergo contraction and fragility . Current models to explain this genetic instability invoke erroneous DNA repair or aberrant replication . Here we show that CAG/CTG tracts are stabilized in Saccharomyces cerevisiae by the alternative clamp loader/unloader Ctf18-Dcc1-Ctf8-RFC complex ( Ctf18-RFC ) . Mutants in Ctf18-RFC increased all three forms of triplet repeat instability—expansions , contractions , and fragility—with effect over a wide range of allele lengths from 20–155 repeats . Ctf18-RFC predominated among the three alternative clamp loaders , with mutants in Elg1-RFC or Rad24-RFC having less effect on trinucleotide repeats . Surprisingly , chl1 , scc1-73 , or scc2-4 mutants defective in sister chromatid cohesion ( SCC ) did not increase instability , suggesting that Ctf18-RFC protects triplet repeats independently of SCC . Instead , three results suggest novel roles for Ctf18-RFC in facilitating genomic stability . First , genetic instability in mutants of Ctf18-RFC was exacerbated by simultaneous deletion of the fork stabilizer Mrc1 , but suppressed by deletion of the repair protein Rad52 . Second , single-cell analysis showed that mutants in Ctf18-RFC had a slowed S phase and a striking G2/M accumulation , often with an abnormal multi-budded morphology . Third , ctf18 cells exhibit increased Rad52 foci in S phase , often persisting into G2 , indicative of high levels of DNA damage . The presence of a repeat tract greatly magnified the ctf18 phenotypes . Together these results indicate that Ctf18-RFC has additional important functions in preserving genome stability , besides its role in SCC , which we propose include lesion bypass by replication forks and post-replication repair .
DNA trinucleotide repeats are subject to frequent expansions and contractions in families affected by Huntington's disease ( HD ) and other inherited neurological disorders [1] , [2] . Some expanded triplet repeats also cause chromosome fragility , as in fragile X syndrome [1]–[3] . The complexity of triplet repeat instability in humans makes it likely that multiple mechanisms contribute to the problem . Two major sources of instability have been identified [1] , [2] , [4] . The first is erroneous DNA repair , which can account for instability in both proliferating and non-proliferating cells . Evidence for erroneous repair of triplet repeats includes the finding of fewer expansions of long CAG/CTG alleles in knockout mice deficient in DNA repair factors Msh2 , Msh3 , Pms2 , or Ogg1 ( summarized in [4] . Less is known in mammals about a causative role of repair on contractions or fragility , although CAG repeat contractions in a human cell line depend on elements of mismatch and nucleotide-excision repair [5] . The second major source of instability is aberrant DNA replication in proliferating cells . Many DNA replication mutants show altered levels of triplet repeat instability [1] , [2] , and treatment of human cell lines with DNA replication inhibitors affects expansions , contractions , and fragility [3] , [6] . Proliferating cells such as those in the male germ line are prone to expansions , although it is not known whether replication is causative in these cells . For example , expansions are present in pre-meiotic testicular germ cells from HD patients , with additional instability in meiotic and post-meiotic cells [7] . Sister chromatid cohesion is one important facet of DNA metabolism that has not been investigated for an effect on triplet repeat instability . A potential role of SCC in modulating triplet repeats is supported by the interplay of SCC with DNA repair [8] , [9] and with replication ( summarized in [10] . One protein complex that participates in SCC is the alternative clamp loader/unloader Ctf18-Dcc1-Ctf8-RFC ( Ctf18-RFC ) . In the absence of the Ctf18-RFC , SCC is compromised [11] , [12] . Biochemically , Ctf18-RFC can load and unload PCNA onto DNA [13]–[15] . It has been proposed that the PCNA unloading activity of Ctf18-RFC may be important to facilitate passage of the replication fork through the cohesin ring [15] , [16] . Ctf18-RFC has also been proposed to play a more general role in fork stabilization [17] . The ability of Ctf18-RFC to recruit PCNA to hydroxyurea-stalled replication forks [16] and to act together with the checkpoint mediator protein Mrc1 in the DNA replication checkpoint [18] is consistent with a more general role at stalled forks . This paper describes the discovery , through blind mutant screens , of yeast Ctf18-RFC mutants that destabilize triplet repeats . Genetic analysis indicates Ctf18-RFC likely acts through replication fork stabilization and/or post-replication repair ( PRR ) , not SCC , to prevent triplet repeat instability , chromosome fragility and cell cycle delays in S and G2/M phases . Our data also support a general role for the Ctf18-RFC complex in preventing DNA damage , a role which becomes more crucial in the presence of an at-risk sequence such as an expanded trinucleotide repeat tract .
Two independent genetic screens identified mutants in Ctf18-RFC as defective in stabilization of trinucleotide repeats ( Text S1 ) . In screen one , the yeast strain contained a ( CAG ) 20-URA3 reporter to monitor contractions ( Figure S1 ) . This strain was transformed with a gene disruption library , and transformants were screened for increased rate of contractions [19] . Following three rounds of testing with increasing stringency , vectorette PCR was used to identify a ctf18::LEU2 allele . The contraction phenotype was confirmed in a commercially obtained ctf18 strain . In screen two , mutants were sought that increased the rate of fragility for a ( CAG ) 85 tract on a yeast artificial chromosome ( YAC; Figure S2 ) [20] . Transfer of the YAC to the commercial haploid deletion strain set was followed by assays for increased fragility . Several rounds of screening showed that the dcc1 mutant reproducibly displayed the fragility phenotype . Thus , two different genetic screens for defects in regulating triplet repeat instability converged on the Ctf18 and Dcc1 components of the Ctf18-Dcc1-Ctf8-RFC complex . Subsequent analysis proved that ctf18 and dcc1 mutants exhibited increased levels of trinucleotide repeat contractions , expansions , and fragility . Contractions in these mutants were increased in every case , by up to 8-fold , for a wide range of repeat lengths ( short ( CAG ) 20 , medium ( CAG ) 70 , and long ( CAG ) 155 tracts; Figure 1A and 1B; Table S1 ) . In fact , the contraction phenotype of dcc1 and ctf18 mutants for long tracts was so pronounced ( Figure 1B ) that the fraction of unaffected cells , only 20–30% , was too low for meaningful analysis of expansions in these strains . Expansions of medium length tracts were increased 10- to 15-fold in ctf18 and dcc1 mutants ( Figure 1C ) ; however , these mutants did not increase expansion rates for very short ( CAG ) 13 repeats . Fragility was increased 2- to 3 fold for dcc1 and ctf18 mutants even in chromosomes without a repeat tract , but was further increased 3- to 5 fold in the presence of an expanded repeat in a length-dependent manner ( Figure 1D ) . In summary , inactivation of Ctf18-RFC substantially increased all three types of instability for CAG runs of 20–155 repeats and also increased general chromosome fragility . If these triplet repeat phenotypes are due to the sister chromatid cohesion ( SCC ) activity of Ctf18-RFC , then mutants in other SCC genes should show similar results . CTF4 and CHL1 were examined first , as these genes gave the closest match to Ctf18-RFC in a genetic interaction map of protein complexes involved in chromosome biology [21] . Mutants in CTF4 and CHL1 both show SCC defects [22]–[24] . Biochemically , Ctf4 couples polymerase a to Mcm and the replisome progression complex [25] , [26] , while Chl1 is a putative DNA helicase that associates with cohesion establishment factor Eco1 [24] . Inactivation of CTF4 did not affect short contractions , but increases were observed in contractions , expansions and fragility of the medium tract , with magnitudes similar to dcc1 and ctf18 mutants ( Figure 1A , 1C , 1D ) . In contrast , the chl1 mutant gave virtually no phenotype in triplet repeat assays ( Figure 1A , 1C , 1D ) . Since the ctf4 phenotype could be due to uncoupling of DNA polymerase α from the replication fork [25] , [26] rather than SCC , the ctf4 and chl1 results suggested the possibility of an SCC-independent phenotype of mutants in Ctf18-RFC and Ctf4 . Accordingly , assays were performed with scc1-73 and scc2-4 temperature-sensitive mutants defective in cohesion maintenance and establishment , respectively . At both permissive ( 23° ) and at semi-permissive ( 31° ) temperatures , scc1-73 and scc2-4 strains were indistinguishable from wild type in nearly every triplet repeat assay ( Figure 1A , 1C , 1D ) . The lack of a triplet repeat instability phenotype in scc1-73 , scc2-4 , and chl1 strains is in contrast to the clear SCC defect seen in these mutants [22] , [27] . While SCC cannot be rigorously excluded due to the essential nature of SCC1 and SCC2 , the most likely explanation for the lack of phenotypes in chl1 , scc1-73 , and scc2-4 mutants is that Ctf18-RFC mitigates triplet repeat instability in an SCC-independent manner . In addition to the canonical clamp loader composed of Rfc1 and the core of Rfc2-5 , there are three alternative clamp loaders; Ctf18-RFC , Elg1-RFC , and Rad24-RFC . The four clamp loaders have distinct biochemical properties , with Ctf18-RFC uniquely exhibiting efficient PCNA unloading [15] . Is Ctf18-RFC also distinct in regards to triplet repeat stabilization ? The instability profiles of dcc1 , elg1 , and rad24 mutants show distinct patterns ( Figure 2 ) . Deficiency in DCC1 increased six of seven types of instability: all forms of contraction , expansion ( save for short tracts ) , and fragility . In contrast , the elg1 and rad24 mutants showed elevated instability in only three or two assays , respectively ( Figure 2 ) . Aside from short tract expansions , the magnitude of elg1 or rad24 phenotypes was always weaker than for dcc1 . Short tract expansions were previously shown [19] to be especially sensitive to defects in the DNA damage response , including rad24 , consistent with the specificity for Rad24 seen in Figure 2 . We conclude that while all three alternative RFC complexes help stabilize CAG/CTG repeats , Ctf18-RFC has the most potent and wide-ranging impact in our assays . Crabbe et al came to a similar conclusion regarding the predominance of Ctf18-RFC in the DNA replication checkpoint [18] . Therefore it remained the focus of this study . Since SCC defects did not account for instability of triplet repeats , we tested the idea that strains deficient for Ctf18-RFC suffer enhanced DNA damage at the trinucleotide repeat , as suggested by the increased repeat fragility in mutants of the complex ( Figure 1 ) . If so , this damage might be susceptible to RAD52-dependent recombinational repair and therefore a rad52 background should alter the mutational spectrum in the absence of Ctf18-RFC . The results show that mutation of RAD52 suppressed , partially or completely , every dcc1 mutability phenotype–contractions of both short and medium CAG/CTG tracts and expansion of medium tracts ( Figure 3A ) . We conclude that Rad52-dependent repair in the absence of Ctf18-RFC does not proceed with fidelity in the context of a CAG repeat , since it results in expansions and contractions . A similar result was also observed in srs2 and mre11 mutants , where increased levels of medium- and long-tract repeat expansions and contractions were dependent on Rad52 [28] , [29] . Some medium tract contractions were Rad52-independent ( Figure 3A ) ; previous data indicated that an additional source of contractions could be processing of DSBs within the repeat tract followed by microhomology-mediated end joining [29] . Fragility was not suppressed or significantly increased in a dcc1 rad52 double mutant compared to the dcc1 single mutant ( Figure 3B ) , indicating that Rad52 does not contribute to fragility resulting from DNA damage in Ctf18-RFC deficient cells . In summary , in the absence of Ctf18-RFC , a Rad52-dependent pathway is operative that is responsible for the majority of the observed contractions and expansions . We considered the possibility that stabilization of replication forks by Ctf18-RFC explains its effects on triplet repeat mutations and fragility . This model is supported by studies showing Ctf18 localization to hydroxyurea-stalled forks in S . cerevisiae [16] , its association with replication origins in unperturbed S . pombe cells [17] , and its physical association with DNA polymerase ε [30] . If true , the fork stabilization model predicts that uncoupling DNA pol ε from the replicative helicase with an mrc1 mutation [31] should exacerbate the triplet repeat phenotype of Ctf18-RFC mutants . Mrc1 is important both for coupling the helicase and polymerase functions at the replication fork , and in signalling during the replication checkpoint and the DNA damage response [31]–[33] . We showed that triplet repeat expansions , contractions , and fragility are elevated in mrc1 mutants [19] , [34] , [35] . Also , mrc1 interacts genetically with ctf18 , ctf8 , and dcc1 [36] . To test mrc1 effects on triplet repeat instability in the absence of Ctf18-RFC , contraction rates were compared for short CAG tracts in single and double mutants of mrc1 , dcc1 , and ctf18 . The results in Table 1 show 4 . 1- to 7 . 1-fold increased contraction rates for single mutants of mrc1 , dcc1 , or ctf18 . The two double mutants gave effects that were significantly greater than additive: 20-fold for mrc1 dcc1 ( p = 0 . 05 ) and 16-fold for mrc1 ctf18 ( p = 0 . 04 ) . This result suggests that Ctf18-RFC and Mrc1 work in parallel to stabilize short repeats . Forward mutation rates at CAN1 were also significantly greater than additive in the mrc1 ctf18 double mutants: 23 . 4-fold over wt compared to 6 . 5-fold and 2 . 6-fold for the respective single mutants ( Table 1; p = 0 . 002 ) , indicating that these proteins act in different pathways in the context of a non-trinucleotide repeat sequence as well . In contrast , Mrc1 , Ctf18 , Ctf8 , and Dcc1 have been proposed to function in the same SCC pathway [22] . This difference is consistent with our earlier conclusion that Ctf18-RFC stabilizes triplet repeats independently of SCC . The mrc1 dcc1 and mrc1 ctf18 double mutants could not be assayed with medium and long tracts due to cell lethality . The results above suggested that Ctf18-RFC helps cope with triplet repeat-associated damage and in stabilizing replication forks , so we tested directly whether the Ctf18-RFC complex has a role in progression through the cell cycle . Cells from a log phase liquid culture were plated on solid media , and microscopic analysis was used to monitor the proportion of cells in each phase of the cell cycle: unbudded ( G1 ) , small budded ( bud size one-third or less the size of the mother cell ( S ) , and large budded ( G2/M ) . The results are quantified in Figure 4A , and representative micrographs are shown in Figure 4B . In wild-type cells with no CAG/CTG tract , there was a distribution of 30% unbudded , 12% small budded , and 55% large budded ( Figure 4A ) . The presence of a CAG/CTG tract changed this distribution in two ways . First , there were more small budded ( S phase ) cells , consistent with replication stress . Second , a new category of cells was observed that were either swollen with large buds or contained multiple buds ( Figure 4B ) , a phenotype that is indicative of unresolved damage in G2/M [37] . The proportion of the multi-budded/swollen cells rose with increasing repeat tract length to as much as 20% of the wild type population ( Figure 4A ) . In general the swelling was modest and most multi-budded clusters contained only one extra bud in wild-type cells ( Figure 4B ) . In dcc1 and ctf18 cells , even without a repeat , multi-budded/swollen cells comprised 18–30% of the population , a level significantly greater than wild-type cells with no tract ( Figure 4A ) . This indicates that the absence of the Ctf18-RFC complex leads to some level of repeat-independent damage that causes accumulation of cells in G2/M . Even more strikingly , the combination of an expanded repeat plus the lack of a functional Ctf18-RFC led to an increase in the multi-budded category to 42–54% of cells ( Figure 4B ) . In addition , the morphological defects in dcc1 and ctf18 mutants with a repeat tract often showed a more severe phenotype , with extreme swelling and many connected buds ( Figure 4B ) . Staining of nuclei revealed that some of the cells within multibudded clusters , often the more swollen ones , had fragmented or missing DNA ( example in Figure 6A ) . We conclude that Ctf18-RFC has an important function in helping resolve repeat-independent DNA damage , and that damage is persisting into the G2 or M phase . Since this phenotype is enhanced by an expanded triplet repeat , and since the expanded repeat causes replication stress , we also infer that Ctf18-RFC helps cope with repeat-induced replication stress during S phase . To measure cell cycle dynamics with more precision , we isolated unbudded G1 cells by micromanipulation and followed their progression through 2–3 cell cycles by microscopy . This single-cell approach measures the time spent in each phase of the cell cycle , and therefore it allows assignment of the cell cycle stage in which defects can first be detected . A schematic example of the approach and some representative data are shown in Figure 5A . The majority of wild type cells with no repeat spent ∼30 min in S phase , with a slight shift to longer S phases when a medium-length ( CAG ) 70 repeat was present ( Figure 5B ) . Cells containing a ( CAG ) 70 tract and lacking DCC1 or CTF18 exhibited several cell cycle phenotypes . First , they divided much more slowly . Average division time was 5 . 8 h for dcc1 ( range 2 . 5–8 . 5 h ) and 3 . 5 h for ctf18 ( range 2 . 0–6 . 0 h ) , compared to 2 . 0 h for wild type ( CAG ) 70 strain . The presence of the repeats enhanced the delay as the dcc1 and ctf18 mutants with no repeat averaged 2 . 5 h and 2 . 0 h per division , respectively . Second , some ctf18 and dcc1 cells stayed small budded 1–2 h , consistent with an S-phase delay , a phenotype that was exacerbated by the presence of the repeat ( Figure 5B ) . In contrast , all wild type cells completed S phase in 1 h or less , regardless of whether the repeat tract was present . Thus , single-cell analysis provides additional evidence for a role of Ctf18-RFC during S phase , as its absence leads to an extended S phase in some cells . Effects in G2/M were also evident from the single-cell analysis ( Figure 5B ) . All wild type cells had a G2/M phase of 2 h or less , regardless of the presence of the trinucleotide repeat . In contrast , some dcc1 and ctf18 cells were detected with G2/M phases of 2 hours or more , even when no repeat was present . The length of the dcc1 and ctf18 strain G2/M phase was greatly increased in the presence of the ( CAG ) 70 repeat , with some cells remaining in G2/M up to 6–8 h ( after which time yeast cells are able to adapt to DNA damage and continue through M even without repair [38] , [39] ) . This single-cell analysis also proved that arrested G2/M cells gave rise to the multi-budded cells described earlier . Finally , when the fate of the colony growth beyond 8 h was monitored , we observed that a majority of the dcc1 and ctf18 cells containing the ( CAG ) 70 repeat tract only accomplished a few additional cell divisions and did not form colonies visible by eye . This observation indicates that we likely underestimated the fragility and instability phenotypes obtained for ( CAG ) 70 repeats , and presumably for ( CAG ) 155 repeats . The cell cycle delays , increased fragility , and Rad52-dependent instability all suggested that DNA damage may be occurring at CAG repeats in the absence of Ctf18-RFC . To directly test for damage , we measured the proportion of wild type and ctf18 cells with a Rad52 focus in the presence or absence of CAG repeats ( Figure 6 ) . Rad52 focus formation occurs at DSBs or at broken replication forks , but not at forks stalled by HU [40] . In the absence of any repeat , only 1 . 1% of wild type cells had a Rad52-YFP focus ( Figure 6A and 6B ) . The presence of the repeats significantly increased the frequency of Rad52 foci to 3 . 3% and 3 . 8% for ( CAG ) 70 and ( CAG ) 155 , respectively ( Figure 6B ) . In ctf18 cells without a repeat tract , the incidence of Rad52 foci was elevated to 8% ( Figure 6B ) , indicating that significant levels of DNA damage are occurring in this background , consistent with the increased fragility observed above . In addition , there was a further increase in cells with foci in ctf18 cells with an expanded repeat , to 17% for both ( CAG ) 70 and ( CAG ) 155 ( Figure 6B ) . Considering that the repeat is in single copy , these data suggest a significant level of damage occurring at the repeat when the Ctf18-RFC complex is not functional . To determine when in the cell cycle the damage occurs , we visualized foci in cells at different stages ( Figure 6C and 6D ) . In wild type cells without a repeat , foci levels were very low , less than 2% , at all stages ( Figure 6D ) . Interestingly , wild type cells containing an expanded CAG repeat had a detectable increase in Rad52 foci in S phase , 40 min after release from α-factor ( Figure 6D ) . This timing coincides with replication through the repeat ( assessed by 2D gel electrophoresis; R . Anand and C . Freudenreich , data not shown ) indicating that Rad52-dependent events at the repeat may be replication-associated . The lower level of foci in cells arrested by nocodazole ( Figure 6D ) suggests that the DNA damage induced in S is usually repaired by G2/M in wild type cells . In ctf18 cells , the percentage of Rad52 foci was very low in G1 cells arrested by α-factor , suggesting that the complex does not have a genome protective function in G1 ( Figure 6D ) . In contrast , the proportion of cells with a Rad52 focus rose to 24% in S phase , and this number was dramatically increased in the presence of a repeat , to 65% for ( CAG ) 70 or 42% for ( CAG ) 155 ( which was a mixture of 155 and contracted tracts ) . These data show that Ctf18-RFC has an important S phase role . We also observed ctf18 S phase cells with more than one Rad52 focus ( Figure 6C ) . Contrary to the wild type situation , Rad52 foci frequently persisted into G2/M . 15–30% of the ctf18 cells still showed Rad52 foci when arrested in G2/early M by nocodazole ( Figure 6D ) . DAPI staining revealed that these foci were detectable in both G2 cells with nuclei at the bud neck as well as in cells entering M phase . Thus the repeat-induced damage sometimes persisted into M phase . Altogether , cell cycle analysis ( Figure 4 , Figure 5 , Figure 6 ) indicates that damage at the repeat tract likely initiates in S phase , and that presence of Ctf18-RFC is important for completing S phase without delay and without accumulation of DNA damage . Judging by the accumulation of cells and presence of Rad52 foci in G2 and even into M phase , a significant portion of damage persists beyond S phase in Ctf18-RFC deficient cells .
This work provides evidence for new functions of Ctf18-RFC in preserving genomic integrity , outside its role in sister chromatid cohesion ( SCC ) . These discoveries stemmed from the application of sensitive and specific genetic assays that revealed Ctf18-RFC's role in protecting a broad range of CAG/CTG repeat lengths from expansion , contraction , and fragility ( Figure 1 ) . Ctf18-RFC mutant phenotypes at triplet repeats were distinct from those shown by mutants in other SCC factors , such as chl1 , scc1-73 , scc2-4 , and mrc1 ( Figure 1 , Table 1 ) . Ctf18-RFC also was more important at trinucleotide repeats compared to the alternative RFC complexes Elg1-RFC or Rad24-RFC ( Figure 2 ) . A novel role for Ctf18-RFC in replication fork bypass of lesions that arise from triplet repeats was suggested by analysis of double mutants between Ctf18-RFC and either Mrc1 ( Table 1 ) or Rad52 ( Figure 3 ) . In agreement with the idea of a role at the replication fork , cells defective in Ctf18-RFC show an extended S phase and increased S-phase levels of Rad52 foci even in the absence of a triplet repeat ( Figure 5 , Figure 6 ) . These mutants also accumulate in G2/M , often with altered morphology and persistence of Rad52 foci ( Figure 4 , Figure 5 , Figure 6 ) , a phenotype consistent with unresolved DNA damage in G2 [41] . These altered cell cycle phenotypes occur even in the absence of a triplet repeat , but the presence of an expanded repeat tract severely exacerbated the mutant defects . Taken together , the results of this study suggest that Ctf18-RFC helps avoid DNA damage arising during replication , that Ctf18-RFC may also be important in coping with damage when it persists into G2 , and that triplet repeats make budding yeast especially dependent on Ctf18-RFC . Our observations are consistent with action of Ctf18-RFC in S phase , during or soon after passage of the replication fork . Previous work localized Ctf18 at or near hydroxyurea-stalled forks by Chromatin IP in S . cerevisiae [16] , and to replication origins in unperturbed S . pombe cells [17] . It was recently shown by DNA combing that fork speed is slowed 3-fold in ctf18 mutant cells [18] . Ctf18-RFC shows a physical association with DNA polymerase ε [30] , [42] , suggesting that it could act directly at the fork in a fork stabilization role . Although our results do not support a role for Ctf18-RFC in cohesion establishment , they are compatible with a role in facilitating replication through the cohesion ring , as proposed in [15] , [16] . One way to explain the repeat-specific effects we see is if Ctf18-RFC promotes dis-assembling or re-assembling the replisome to facilitate bypass through replication barriers , such as a cohesin ring or a hairpin structure . Another possibility is Ctf18-RFC could have a more general fork stabilizing function to prevent formation of hairpins or other aberrant secondary structures associated with trinucleotide repeats . In either case , fork integrity would be affected in ctf18 , dcc1 , or ctf8 mutants , leading to fragility and increased recombination , and ultimately to triplet repeat mutations . This model would also explain the ctf4 phenotype on instability ( Figure 1 ) , based on the role of Ctf4 in coupling DNA polymerase α to the replication fork [25] , [26] . Alternatively , the Ctf18-RFC S-phase role could be in a repair process that occurs behind the fork , but still in S phase , as described below . The striking G2/M accumulation phenotype and persistence of Rad52 foci was unexpected in mutants of Ctf18-RFC , in part because this phenotype suggests the presence of unresolved DNA damage that persists beyond S phase . Thus , either the damage incurred during S phase in ctf18 cells is often not easily repaired , or Ctf18-RFC also plays a role in helping resolve or repair damage during G2 . Recently , it was shown that post-replication gap repair can operate effectively when limited to the G2 phase , and the authors proposed that PRR occurs primarily on gaps left behind replication forks that have re-primed and continued [43] , [44] . Interestingly , their data suggest that error-free PRR , which is dependent on Rad5-catalyzed polyubiquitylation of PCNA , usually commences in S-phase and continues into G2/M . Thus with regard to timing , our data would be consistent with a role for Ctf18-RFC in error-free PRR . The dependence on HR for instability is also consistent with a role in error-free PRR , since strand invasion is needed for the template-switching step . In contrast , it is unlikely that our results with Ctf-18 RFC relate to error-prone PRR , since HR does not occur during translesion synthesis . Notably , the triplet repeat defects associated with absence of Ctf18-RFC are different than results when the PRR pathway is abolished by deletion of Rad5 or abolishing PCNA modification , as those mutants specifically increased expansions , but not contractions , of short repeats [45] . Thus , our results would be most consistent with aberrant PRR , rather than ablation of the pathway . How might Ctf18-RFC function biochemically to protect triplet repeats ? Ctf18-RFC was shown to load and unload PCNA in vitro in a manner that is more efficient on single stranded DNA and inhibited by RPA [15] . Significant unloading was not seen for the other complexes , RFC , Rad24-RFC , and Elg1-RFC . This unique biochemical activity of Ctf18-RFC mirrors the distinct pattern in our genetic observations ( Figure 2 ) . Possibly the loading/unloading function is especially important at triplet repeats either to bypass a previously formed hairpin structure or to minimize exposed single strands during replication and thereby reduce secondary structure formation and instability . A second possibility is that a defect in PCNA unloading during a gap repair event could lead to a persistent HR structure that would be prone to breakage or cleavage . In this model , contractions could occur during DSB repair end processing as proposed in [29] . In contrast , timely unloading of PCNA could facilitate proper resolution of a recombination intermediate without breakage , for example by Sgs1 dissolution . Sgs1 has a role in resolution of X-shaped SCJs that form upon replication of damaged templates [46] , and a recent study links Sgs1 to resolution of an intermediate that occurs during ubiquitylated PCNA-dependent gap repair [43] . Intriguingly , deletion of SGS1 also led to increased contractions and fragility , similar to ( though not as dramatic as ) ctf18 or dcc1 mutants , and in some contexts , the contractions were also Rad52-dependent [28] . Thus , for example , a defect in Ctf18-dependent unloading of ubiquitylated PCNA could lead to toxic recombination products , similar to a defect in Sgs1 activity . It will be interesting to learn how Ctf18-RFC functions on ubiquitylated PCNA , as we showed previously that expansion rates are elevated when PCNA ubiquitylation is blocked [45] . Does repeat stabilization by Ctf18-RFC extend to human cells ? Human Ctf18-RFC was shown to control the velocity , spacing , and restart activity of replication forks via acetylation of the cohesin ring [47] . Also , human Ctf18-RFC has been recently shown to be necessary for accumulation of polymerase ε during repair of UV lesions induced outside of S phase [48] . A key future study is to see whether , and how , human Ctf18-RFC protects triplet repeats .
Most strains were derived from BY4741 ( MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 ) or BY4705 ( MATα his3Δ200 leu2Δ0 lys2Δ0 met15Δ0 trp1Δ63 ura3Δ0 ) , isogenic derivatives of Saccharomyces cerevisiae strain S288C ( Open Biosystems; [49] . Isogenic derivatives were obtained commercially ( Open Biosystems ) or were created by targeted deletion of BY4741 or BY4705 . The scc1-73 , scc2-4 and wild type parent strains ( MATa ade2-1 can1-100 leu2-3 , 112 his3-11 , 15 ura3-1 trp1-1 , except scc1-73 strain that was TRP1 ) were provided by Philippe Pasero , CNRS , Montpellier , France . Strains used for foci experiments were derived from W303 ( MATa ADE2 his3-11 , 15 leu2-3 , 112 bar1::LEU2 trp1-1 RAD52-YFP ) , obtained from R . Rothstein , Columbia University , NYC , NY . The triplet repeat sequences reported here all have the CAG repeat on the lagging strand template , and CTG repeats on the Okazaki fragment . This CAG nomenclature is used throughout . Contraction and expansion rates for short CAG tracts were measured by fluctuation analysis , and authenticated by PCR , as previously described [19] . Statistical analyses were performed using the Wilcoxon Mann Whitney test . P values of less than 0 . 05 were considered statistically significant . Forward mutation rates for the CAN1 gene were determined by fluctuation analysis using canavanine at 60 µg/ml , and statistical analysis was performed using Student's t-test for comparison with wt and a two-way ANOVA with interaction tests [50] for comparison between single and double mutants . Expansions , contractions , and fragility of medium and long CAG tracts were measured using a YAC system , as described previously [20] . Diagrams of these assays are shown in Figure S2 . Contraction and expansion frequencies for medium ( CAG ) 70 and long ( CAG ) 155 tracts were determined as previously described previously [29] . For each strain , approximately 150 colonies were analyzed for CAG repeat length by colony PCR in at least three separate experiments , using primers flanking the CAG repeat ( P1 and P2 in Figure S2 ) . PCR products were separated on a 2% Metaphor gel ( Cambrex Bio Science Rockland , Inc . ) and sized . The frequency of repeat expansions and contractions in each strain background was calculated and statistical significance determined by the Fisher's exact test . Repeat lengths from 0 to ∼200 CAG repeats with an accuracy of +/−3 repeats can be obtained by this method . Fragility assays were performed as in [29] . Mutation rate was determined using the method of maximum likelihood [51] and data presented are an average of 3–5 experiments . Error bars indicate the standard error of the mean . Significance compared to the wild-type value for the same tract length was determined using a pooled variance t-test . Growth temperature was 30°C unless otherwise indicated . A summary of the whole instability and fragility data is represented in the Table S1 . Cell cycle distribution and cell morphologies ( Figure 4 ) were obtained as follows: YC-Leu-Ura liquid cultures were grown to mid-logarithmic phase and analyzed microscopically for the presence of unbudded cells , small-budded cells ( bud smaller than one-third of the mother cell ) , large-budded cells ( bud equal to or larger than one-third of the mother cell ) , and other cells ( malformed cells with protruded or multiple buds ) . The assays were repeated at least 3 times for each strain . Pictures of the cells were taken using a Zeiss AX10 microscope , under 63× magnification . Single cell assays ( Figure 5 ) were initiated by plating mid-logarithmic phase liquid cultures onto solid YC-Leu-Ura media . Single unbudded cells were isolated by micromanipulation , and their progression was monitored by microscopy every 30 min for 6 . 0–8 . 5 h . ( 1–4 cell divisions ) . The numbers of single cell lineages monitored were: wild type , no repeat ( 12 ) ; wild type , medium tract ( 12 ) ; ctf18Δ , no repeat ( 24 ) ; ctf18Δ , medium tract ( 24 ) ; dcc1Δ , no repeat ( 53 ) ; and dcc1Δ , medium tract ( 33 ) . Because we did not micromanipulate daughter cells away from each other , and because many of them in fact could not be separated ( indicating incomplete cell division and leading to a multi-budded cell ) , it was not always possible to distinguish the end of G2/M and beginning of G1 . In these cases , we grouped the two cell cycle phases ( G2+G1 ) . At the end of the experiment , cells still in G2 phase for a time less than the wild-type G2/M average time ( 1 h for no tract and 1 . 5 h for medium tract , respectively ) were not considered to be informative for their G2 phase , and were not counted . Cell imaging and fluorescent microscopy: examination of Rad52-YFP focus levels by microscopy was performed as previously described [52] . Briefly , cells were grown overnight in SC-Leu-Ura media at 23°C and exponentially growing cultures were prepared for microscopy . To visualize nuclear DNA by DAPI staining ( 50 ng/ml ) , cells were fixed in ethanol before mounting on the slide . Cell images were captured using a Zeiss AX10 microscope ( Carl Zeiss , Thornwood , NY ) equipped with a Retiga EXi camera ( Qimaging ) , and acquired using SlideBook software ( Intelligent Imaging Innovations , Denver , CO ) . All images were taken at 63-fold magnification . A single DIC image and 17 YFP images obtained at 0 . 3-µm intervals along the z-axis were captured for each frame , and Rad52-YFP foci were counted by inspecting all focal planes intersecting each cell . For each strain , ∼200–800 cells ( range 172–1483; Table S2 ) were scored for Rad52-YFP foci . Fisher's exact t-test was used to calculate significance . For cell synchronization , cells were arrested in G1 phase by treatment with α-mating factor ( 3 . 4 µg/ml; Sigma , St . Louis , MO ) for 2–5 hrs ( 2 for wt , 5 for ctf18 ) . S phase was evaluated by releasing cells from G1 arrest by washing three times in water , resuspension in SC-Leu-Ura medium , and 40 min incubation ( consistent with replication timing of the CAG repeat on the yeast artificial chromosome , R . Anand and C . H . Freudenreich , unpublished results ) prior to processing for fluorescence microscopy . Prolonged time in the presence of α-mating factor didn't eliminate the multi-budded category; ctf18 multi-budded clusters were excluded from the G1 and S phase Rad52 foci quantification . For G2/M arrest , cells were treated with 0 . 2 M nocodazole for 3–5 hrs . Because of the high frequency of multi-budded cells in the ctf18 strain , DNA DAPI staining was used as the reference to set the total number of cells .
|
DNA trinucleotide repeats are naturally occurring runs of three base-pairs . Genetic mutations that expand ( lengthen ) triplet repeats cause multiple neurological diseases , including Huntington's disease . Triplet repeats also contract ( shorten ) and break . This complex behavior suggests triplet repeats are problematic for DNA replication and repair enzymes . Here , we identified a cellular factor called Ctf18-RFC that helps yeast cells accurately replicate triplet repeats . We found that mutants lacking Ctf18-RFC show enhanced levels of expansions , contractions , and fragility over a wide range of triplet repeat lengths . Other labs showed that Ctf18-RFC helps replicated chromosomes stay together until mitosis , a process called sister chromatid cohesion . However , we found that Ctf18-RFC stabilizes triplet repeats in a different way , by helping the DNA replication machinery move through triplet repeats and by helping repair any resulting DNA damage . Another insight is that Ctf18-RFC provides these functions at other sites besides triplet repeats , but the presence of a triplet repeat makes the yeast cell especially dependent on Ctf18-RFC to prevent DNA damage and allow normal cell cycle progression . Our results implicate Ctf18-RFC as a new player in the triplet repeat story and indicate that it functions through novel roles to preserve genome integrity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"molecular",
"biology/dna",
"replication",
"molecular",
"biology/recombination",
"genetics",
"and",
"genomics/chromosome",
"biology",
"genetics",
"and",
"genomics/disease",
"models",
"genetics",
"and",
"genomics/gene",
"function",
"molecular",
"biology/dna",
"repair"
] |
2011
|
New Functions of Ctf18-RFC in Preserving Genome Stability outside Its Role in Sister Chromatid Cohesion
|
Metabolic network reconstructions represent valuable scaffolds for ‘-omics’ data integration and are used to computationally interrogate network properties . However , they do not explicitly account for the synthesis of macromolecules ( i . e . , proteins and RNA ) . Here , we present the first genome-scale , fine-grained reconstruction of Escherichia coli's transcriptional and translational machinery , which produces 423 functional gene products in a sequence-specific manner and accounts for all necessary chemical transformations . Legacy data from over 500 publications and three databases were reviewed , and many pathways were considered , including stable RNA maturation and modification , protein complex formation , and iron–sulfur cluster biogenesis . This reconstruction represents the most comprehensive knowledge base for these important cellular functions in E . coli and is unique in its scope . Furthermore , it was converted into a mathematical model and used to: ( 1 ) quantitatively integrate gene expression data as reaction constraints and ( 2 ) compute functional network states , which were compared to reported experimental data . For example , the model predicted accurately the ribosome production , without any parameterization . Also , in silico rRNA operon deletion suggested that a high RNA polymerase density on the remaining rRNA operons is needed to reproduce the reported experimental ribosome numbers . Moreover , functional protein modules were determined , and many were found to contain gene products from multiple subsystems , highlighting the functional interaction of these proteins . This genome-scale reconstruction of E . coli's transcriptional and translational machinery presents a milestone in systems biology because it will enable quantitative integration of ‘-omics’ datasets and thus the study of the mechanistic principles underlying the genotype–phenotype relationship .
High-throughput experimental technologies enable the production of heterogeneous data , such as expression profiles and proteomic data , for almost any organism of interest . A detailed mathematical representation of the in vivo cellular network is required to obtain a holistic understanding of cellular processes from these data sets and to quantitatively integrate them into a biological context . One such approach is the bottom-up network reconstruction , which builds manually networks in a brick-by-brick manner using genome annotation and component-specific information ( e . g . , biochemical characterization of enzymes ) [1] , [2] . This reconstruction procedure is well established for metabolic reaction networks and has been applied to many organisms , including Human [3] , Saccharomyces cerevisiae [4] , [5] , Leishmani major [6] , Escherichia coli [7] , Helicobacter pylori [8] , Pseudomonas aeruginosa [9] , and Pseudomonas putida [10] , [11] ( see http://systemsbiology . ucsd . edu/ for an continually updated table of metabolic reconstructions ) . These bottom-up metabolic networks differ from other network reconstructions as they are tailored to the genomic content of the target organism and built manually using biochemical , physiological , and other experimental information in addition to the genome annotation . Hence , these reconstructions can be thought of as biochemically , genetically , and genomically structured ( BiGG ) knowledge bases [12] . The reconstruction and modeling procedure is a 4-step process: 1 ) obtaining a draft reaction list based on genome annotation and biochemical databases , 2 ) refinement of reaction list using experimental information ( e . g . , from literature ) , 3 ) conversion of the reaction list ( reconstruction ) into a computable format and application of systems boundaries to define condition-specific models , and 4 ) the evaluation and validation of the model content using various mathematical methods ( see also [1] , [2] , [12] , [13] ) . By iterating step 2 to 4 , reconstructions that are self-consistent within their defined scope can be generated . Metabolic network reconstruction have demonstrated to be useful in at least 5 areas of applications [2]: ( i ) biological discovery [14] , ( ii ) phenotypic behavior [15] , ( iii ) bacterial evolution [16] , ( iv ) network analysis [17] , and ( v ) metabolic engineering [18] . This wide range of applications of the metabolic reconstructions is possible because they can be readily converted into predictive , condition-specific models . Unlike more traditional approaches to modeling metabolism , the constraint-based modeling approach ( COBRA ) requires few , if any , parameters [12] , [19] . The stoichiometric information encoded in the reconstruction ( i . e . , reaction list ) can be represented mathematically as a stoichiometric matrix , S , where the rows correspond to the components and the columns correspond to the reactions ( Figure 1 ) . While the COBRA approach has been successfully applied to metabolic networks , the same principles and assumptions can be also employed to reconstruct and model other cellular functions , such as signaling [20]–[22] , regulation [23] , and protein synthesis [24] . In this study , we extended and refined earlier work by Allen et al . , which proposed a stoichiometric formalism to model protein synthesis and illustrated it on some E . coli genes and operons [24] . We created a more detailed , gene-specific representation of the transcriptional and translational processes , which explicitly accounts for the sequence-specific synthesis of DNA , mRNA , and proteins . This reconstruction enables quantitative integration of high-throughput data such as gene expression , proteomic , and mRNA degradation data . Moreover , proteins are produced in high copy numbers in growing cells; thus , any quantitative mechanistic modeling and analysis of high-throughput data needs to account for the synthesis cost associated with these molecules . Numerous studies have been published that investigate protein synthesis using kinetic models [25]–[29] . These models are generally tailored to the questions they address making it difficult to readily apply them for modified problems . Since stoichiometric relationships are a common requisite for any type of mechanistic modeling , organism-specific BiGG knowledge bases can be used as templates to derive problem-specific , mechanistic models ( Figure 1 ) . In fact , network stoichiometry is a dominant feature of kinetic models as well [30] . Thus , network reconstruction serves as a platform for steady-state and kinetic modeling ( Figure 1 ) . In this study , we present a new generation of network reconstructions , which directly account for the synthesis of individual mRNA and proteins ( Figure 2A ) . We named the mathematical representation of this reconstruction the Expression matrix , or ‘E-matrix’ , since it encodes the expression of mRNA and proteins . All network reactions were formulated to account for gene-specific and E . coli-specific details , such as nucleotide composition , operon association , and sigma factor usage . Furthermore , we used information from three databases and more than 500 scientific publications to formulate mechanistically detailed and accurate reactions . This reconstruction is the first comprehensive database detailing the available information for these cellular functions and can thus be deemed a knowledge base . After conversion of the ‘E-matrix’ reconstruction into condition-specific models corresponding to different doubling times , we were able to accurately predict the ribosome production reported in literature , without any parameterization . Furthermore , we show that the ‘E-matrix’ can be used to study the effect of rRNA operon deletion . Our results predict that a high density of RNA polymerases is required on the remaining rRNA operons , to achieve the reported ribosome numbers . Finally , we show that proteins used in the ‘E-matrix’ could be grouped into functional modules which lead to a more simplified view of the network .
The conversion of a network reconstruction into a mathematical model can be achieved , analogously to metabolic networks [1] , by defining system boundaries and applying condition-dependent constraints on exchange and intracellular reactions ( Figure 1 ) [1] , [40] . Therefore , experimental data can be used to constrain the set of feasible network fluxes in a physiologically relevant manner . In the following section , we will illustrate the use of condition-specific models that were derived from the ‘E-matrix’ reconstruction . In this study , we present the first , mechanistically and chemically detailed , genome-scale network reconstruction of the transcriptional and translational machinery of E . coli . Biochemical components , reaction formulation , and quality control measures analogous to metabolic network reconstructions were used to incorporate bibliomic data from the last 50 years into one reconstruction ( Figure 2 ) . The corresponding knowledge base can be queried online ( http://bigg . ucsd . edu/E-matrix ) . This stoichiometric reconstruction represents a first step towards modeling this complex cellular function , and will require iterative refinement as new data becomes available . By describing the stoichiometric relationships between the components involved in transcription and translation , this reconstruction enables the quantitative integration of disparate ‘-omics’ data into a computational model ( Figure 5 ) . We demonstrated that low- and high-throughput data can be readily integrated and used as constraints on model reactions and the subsequent reduction of the feasible set of reaction fluxes results in physiological relevant predictions ( Figure 5B–D ) . Furthermore , we showed that the computational model can be used to accurately predict ribosome production under different growth conditions ( Figure 3 ) . The deletion of single or multiple rRNA operons from the ‘E-matrix’ predicted that a high density of RNA polymerases is required on the remaining rRNA operons to achieve the reported ribosome numbers ( Figure 4B ) . Computational analysis of the ‘E-matrix’ can provide further insight into the topologically local and global relationship between proteins in terms of functional modules ( Figure 6 ) . This ‘E-matrix’ reconstruction ushers in a new generation of cellular network models that account quantitatively for mRNA and proteins . The ‘E-matrix’ offers the potential to ( i ) serve as a platform for integrated , numerical analysis of heterogeneous , quantitative high-throughput datasets; ( ii ) increase our understanding of the relationship between mRNA and protein abundance; ( iii ) be integrated with metabolism by extending the transcriptional and translational reactions to metabolic genes; ( iv ) be integrated with regulatory events by formulating regulatory rules for the genes of the ‘E-matrix’ and extending the transcriptional and translational reactions to transcription factors; and ( v ) enable computation of the material and energetic cost of macromolecular synthesis . These capabilities are important milestones in moving towards a more comprehensive genome-scale in silico model of all cellular processes in E . coli . Furthermore , the underlying reconstruction methodology can be readily extended and applied to other prokaryotes . Such extension could lead to further insight into conserved and unique features of the transcriptional and translational machinery of prokaryotes . The history of E . coli metabolic reconstructions now spans more than 17 years , with numerous iterative reconstruction refinements and applications superseding initial expectations [63] . The reconstruction of transcriptional and translational machinery E . coli , and other prokaryotes , will have the same impact on systems biology , especially when integrated with metabolism , regulation , and condition-specific high-throughput data sets ( Figure 5 A ) . This work represents hence a crucial step towards the important and ambitious goal of whole cell modeling [64] .
The reconstruction of the transcriptional and translational machinery of E . coli was approached by first identifying the main components from genome annotation [32] , E . coli specific primary and review literature , as well as multiple databases ( Figure 2B ) . For each of these components the gene ID ( b-number ) , gene position , necessary metallo-ions and cofactors , and protein stoichiometry were extracted . The synthesis reactions for every network component were created using template reactions , which was possible since reaction mechanisms are similar for all network components ( see Text S1 for examples ) . These template reactions were carefully formulated and derived from primary and review literature ( Tables S15 , S16 , S17 ) . The template-based network reconstruction was performed using the scripting language , Perl ( http://www . perl . com/ ) . Each template reaction as well as protein complex formation reactions were generated manually based on legacy data ( Tables S15 , S16 , S17 , and S18 ) . Every network reaction was mass- and charged balances assuming a physiological pH of 7 . 2[1] . The basis for the reconstruction was the genome sequence , m56 [65] , the most current gene coordinates [32] , and the transcription unit definitions provided by EcoCyc ( version 10 . 6 , [36] ) . This information was also used to ( i ) calculate the formula and charge for each mRNA and protein species; ( ii ) individually adjust the template reactions for , e . g . , NTP requirement; and ( iii ) transcribe operons rather than genes . A complete list of all transcription units can be found in Table S9 . The genetic code used for this reconstruction is listed in Table S11 . Network gap analysis was performed after the initial reaction list was obtained . Multiple iterations of content refinement and evaluation ensured completeness of the network within its scope by including missing components and reactions ( Text S1 , Figure A–c ) . One network gap remained , which is the RNase PH that is annotated as pseudogene in Riley et al . [32] . The systems boundaries of the ‘E-matrix’ were defined by adding 76 exchange reactions for amino acids , NTP , and other metabolic components . Furthermore , demand reactions were added for each protein gene product ( Tables S9 and S12 ) . The ‘E-matrix’ model is available in Matlab format ( Dataset S1 ) . The mathematical model of the ‘E-matrix’ was represented by a stoichiometric matrix , S ( m rows×n columns ) , where m is the number of components and n is the number of reactions [1] . Reactions within the network were mass-balanced and assumed to be at steady state such that , where is flux vector . Additional constraints on upper , , and lower , , bounds were applied in form of on each reaction i . The lower limits were set to zero for irreversible reactions . The unit for each reaction flux was defined to be , where the doubling time ( ) is given in minutes , if not stated differently . The upper bounds on exchange reactions for NTPs and amino acids were constrained for all simulation conditions , while the lower bounds remained unconstrained . The fractional contribution of NTPs and amino acids were calculated based on experimental data [53] and scaled by RNA and protein content found at each doubling time ( Text S1 ) . The upper bounds of stable RNA transcription initiation reactions were constraint based on experimental data [42] using the following formula: where is the rRNA transcription initiation rate , is the copy number of the stable RNA gene i per cell due to gene dosage ( Table 3 ) , and TD the doubling time ( see Text S1 ) . The mRNA degradation rates were calculated using expression data in LB medium and mRNA half-life times [56] with where is the concentration of mRNA i in the cell , is the half-life time of mRNA i in LB medium , is the half-life time of mRNA i in M9 medium+glucose ( refer to Text S1 for detailed calculation ) . A total number of 4 , 600 mRNA per cell at 30 min doubling time was assumed [42] . The lower bound ( ) was set to be 0 . Since the expression data as well as the total mRNA number have experimental errors , the upper bound on each reaction flux had to be relaxed by multiplying each mRNA concentration with a factor of 10 . The upper bound on mRNA recycling , or CONV2 reactions , were constrained using the following formula: where is the doubling time ( s ) , is the length of mRNA i , and is the translation elongation rate at TD . This later set of reactions accounts for multiple translation rounds of an mRNA transcript between synthesis and degradation . The exchange flux rates and the transcription initiation rates of ribosomal RNA operons were constrained as described above . At each doubling time , the ribosome production rate ( DM_rib_50 ) was chosen as objective function , and the maximal possible production rate under the given set of constraints was calculated using linear programming . This analysis was carried out as illustrated in Figure 4 . First , the transcription initiation rates were applied as constraints to all rRNA operons for the different doubling times ( as described above ) . Using flux balance analysis ( FBA ) [66] , [67] , we optimized for ribosome production ( DM_rib_50 ) . For the strains deficient in one rRNA operon , we deleted each operon separately by setting the maximal possible transcription initiation rate to 0 ( ) , which corresponds the deletion of the reaction from the network . We optimized again for the ribosome production . For multiple rRNA operon deficient strains , all possible combinations of rRNA operon deletion were considered ( Table 3 ) , leading to the error bars in Figure 4 . The compensation factors were chosen arbitrarily ( 1 . 5 , 2 , 2 . 5 , and 4 ) and multiplied to all active rRNA operons in the mutant strains . Note that the unit for these simulations was . Flux variability analysis was performed as described by Mahadevan [68] using linear programming . Briefly , for every network reaction the minimal and maximal solution was determined by successively defining each network reaction as objective function . The lower bound of the ribosome production rate ( DM_rib_50 ) was constrained to . The pair-wise correlations between protein component recycling reactions ( PROT_RECYCL ) were determined in LB-medium using linear programming . The maximal reaction flux for reaction A was determined and its upper and lower bound was set to be the maximal flux value . The minimal and maximal reaction flux for reaction B was determined under this new set of constraints . The same procedure was repeated for the minimal flux rate through reaction A . The same approach was repeated for reaction B with respect to reaction A . This method resulted in pair wise dependency plots for all recycling reactions . The area of feasible flux rates was determined using a convex hull algorithm [69] and scaled by the maximal flux rates for each reaction . The reaction correlation was defined to be 1 minus the area between two network reactions . All calculation were performed using MatLab ( The MathWorks , Inc , Natick , MA ) and TomLab ( TomLab Optimization , Inc , Pullman , WA ) as linear programming solver . This knowledge base is freely available at http://bigg . ucsd . edu/E-matrix
|
Systems biology aims to understand the interactions of cellular components in a systemic manner . Mathematical modeling is critical to the integration and analysis of these components on a conceptual as well as mechanistic level . To date , detailed genome-scale reconstructions of metabolism have become available for a growing number of organisms . Although metabolism has an important role in cells , other cellular functions need to be considered as well , such as signaling , regulation , and macromolecular synthesis . For instance , the cellular machinery required for RNA and protein synthesis consists of a complex set of proteins . Here , we show that one can collect all of the necessary information for a prokaryotic organism to create a gene-specific , fine-grained representation of the macromolecular synthesis machinery . E . coli was chosen as a model organism because of the wealth of available information . The explicit representation of transcription and translation in terms of a mass-balanced network enables a detailed , quantitative accounting of the protein synthesis capabilities of E . coli in silico . Hence , this study demonstrates the feasibility of constructing very large networks and also represents a critical step toward building cellular models of growth that can account for gene-specific protein production in a stoichiometric fashion on the genome scale .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/microbial",
"physiology",
"and",
"metabolism",
"computational",
"biology/literature",
"analysis",
"cell",
"biology/microbial",
"growth",
"and",
"development",
"molecular",
"biology/bioinformatics",
"computational",
"biology/metabolic",
"networks",
"computational",
"biology/signaling",
"networks",
"computational",
"biology/systems",
"biology"
] |
2009
|
Genome-Scale Reconstruction of Escherichia coli's Transcriptional and Translational Machinery: A Knowledge Base, Its Mathematical Formulation, and Its Functional Characterization
|
Dengue is one of the most common infectious diseases . The aim of this study was to systematically review acute disseminated encephalomyelitis ( ADEM ) and to represent a new case . We searched for articles in nine databases for case reports , series or previous reviews reporting ADEM cases in human . We used Fisher’s exact and Mann-Whitney U tests . Classification trees were used to find the predictors of the disease outcomes . We combined findings using fixed- and random-effects models . A 13-year-old girl was admitted to the hospital due to fever . She has a urinary retention . The neurological examinations revealed that she became lethargic and quadriplegic . She had upper limbs weakness and lower limbs complete paraplegia . Her status gradually improved after the treatment . She was nearly intact with the proximal part of her legs had a mild weakness in discharge . The prevalence of ADEM among dengue patients was 0 . 4% [95% confidence intervals ( 95% CI ) 0 . 1–2 . 5%] , all neurological disorders among dengue was 2 . 6% [95% CI 1 . 8–3 . 8%] , and ADEM among neurological disorders was 6 . 8% [95% CI 3 . 4–13%] . The most frequent manifestation of ADEM was altered sensorium/consciousness ( 58% ) , seizures and urination problems ( 35% ) , vision problems ( 31% ) , slurred speech ( 23% ) , walk problems ( 15% ) then ataxia ( 12% ) . There was a significant difference between cases having complete recovery or bad outcomes in the onset day of neurological manifestations being earlier and in temperature being higher in cases having bad outcomes ( p-value < 0 . 05 ) . This was confirmed by classification trees which included these two variables . The prevalence of ADEM among dengue and other dengue-related neurological disorders is not too rare . The high fever of ADEM cases at admission and earlier onset day of neurological manifestations are associated with the bad outcomes .
Dengue , a worldwide prevalent mosquito-borne infectious disease , is a flavivirus spread by several species of Aedes type mosquitos , mainly Aedes aegypti [1] . Dengue has become a dangerous burden and is widely spread in more than 110 countries [2 , 3] . The incidence of dengue has increased to reach 30-fold throughout the past 50 years [4] . Annually , between 50 and 528 million people have the infection and about 10 , 000 to 20 , 000 deaths [5–8] . Dengue has a wide variety of manifestations , from fever to dengue shock syndrome and/or multiple organs failures [1 , 9] . There are a series of biological predictors such as immune cytokines [10–12] , circulating DNA [13] , microalbuminuria [14] , nonstructural protein 1 [15–17] , IgM , IgG [18] , IgA [19] and endothelial cell damage , as well as dysfunction predictors , have been evaluated [20] . However , no efficient marker for the prediction of severe dengue infection has been discovered [20–22] . Although neurological problems of dengue virus ( DENV ) have also been reported , the incidence of this group is uncommon between 0 . 5 and 6 . 2% [23] . A previous systematic review has investigated the factors associated with DENV and revealed that these factors included the neurological signs [24] . DENV associated neurological problems can be divided into DENV direct invasion and para- or post-infectious disease [3] . These neurological DENV include encephalopathy , encephalitis , immune-mediated syndromes as acute disseminated encephalomyelitis ( ADEM ) and Guillain-Barré syndrome ( GBS ) , neuromuscular complications as hypokalemic paralysis and dengue-associated stroke [25–30] . ADEM in dengue is very rare and it may occur during the acute phase or post-infectious phase of dengue . It is known to involve an immune-mediated mechanism in which the cytokine overproduction is triggered by DENV [25] . There was another theory which is the immune-mediated attack by autoantibodies and/or T-cells to central nervous system myelin structure . This leads to acute demyelination of the white matter of the brain , spinal cord or both . Thus , it results in an altered mental status and focal neurologic findings in ADEM patient such as paralysis [3] . Although ADEM causes a significant impact on dengue patients , data about this complication is still lacking . Understanding of this complication provides a potential insight into the clinical picture of DENV infection . Thus , this study aimed to conduct an extensive systematic review and meta-analysis of the literature on the ADEM manifestations in dengue with a new case report .
All the methods were performed in accordance with the relevant approved guidelines , regulations and declaration of Helsinki . The experimental protocols were approved by Children’s Hospital No . 2 in Ho Chi Minh City in Vietnam . Written informed consent was obtained from the parents to have their girl’s details and accompanying images published and approved by the aforementioned hospital . Moreover , all patient’s data was analyzed anonymously . This systematic review was performed according to the Preferred Reporting Items for Systematic Review and Meta-analyses statement ( PRISMA ) ( S1 Table ) [31] . We had developed and registered a protocol of methods ( CRD42016047583 ) . From inception to the 12th of September 2016 , we searched for suitable studies in nine databases including; PubMed , Google Scholar , Institute of Science Index ( Web of Science ) , Scopus , Popline , World Health Organization Global Health Library , Virtual Health Library , New York Academy of Medicine Grey Literature Report , System for Information on Grey Literature in Europe and cross-references from the included articles and previous reviews . The search strategy used was ( ADEM or encephalomyelitis ) and dengue . Three independent reviewers initially scanned primary titles and abstracts ( when available ) to select potential full-text articles for further scrutiny according to the inclusion and exclusion criteria . The inclusion criteria were as following; case reports , case series , previous literature reviews or systematic reviews discussing post-infectious immune-mediated ADEM of dengue infection in human . Exclusion criteria were as following; other study designs rather than case reports , case series , previous literature reviews or systematic reviews , other complications rather than ADEM , overlapped data sets , data which could not be extracted , duplicated studies and unreliable or incomplete data , no full-text available , abstract-only articles ( conference , letters , commentaries ) , or thesis , books , review editorial or author response . When the title and abstract were not rejected by any reviewer , the full-text of the article was obtained and carefully reviewed for inclusion by the three reviewers . Inclusion or exclusion of each study was determined by discussion and consensus between the three reviewers . When the disagreement occurred , a consensus decision was made following discussion with a senior reviewer . Data were extracted by three authors and were checked by at least another author . The disagreement was resolved via discussion and a consensus reached between the three authors . The data extraction form in an Excel file was developed by two authors based on a pilot review and extraction . The data extracted included the first author , year of publication , year of patient recruitment , study design , country of origin and characteristics of the population ( infant , children , adult ) , gender , age at examination of included individuals , the manifestations , the blood and CSF analyses , medications used , visual and neurological examinations , renal and liver function tests and outcome of each patient . If there were more than one value from the examination , the nadir value , e . g . the lowest platelets , the highest packed cell volume ( PCV ) , was extracted . Papers published by the same research group and studying the same factor were checked for potentially duplicated data based on the year of patient recruitment and hospital where the patients were recruited . When duplications were noted , the largest data set was used for our study . Fisher’s exact and Mann-Whitney U tests were used for categorical and continuous variables , respectively . The values with different units were converted into one common unit to make the values comparable . The classification tree models were used to find the independent predictors that best predict bad outcomes ( partial recovery or death ) versus complete recovery as well as complete recovery versus partial recovery [32] . In particularly , 20 variables including age , sex , clinical examination’s variables and the steroid treatment and its administration route were included to build classification tree models . We selected the maximum depth of the tree to be five to construct a tree of reasonable complexity . If the tree is too complex , it is difficult to apply . Likewise , we chose the minimum number of observation at each leaf node is equal to five to prevent the tree from sub-dividing into overly specific nodes that contain little supporting data . The performance measures of the tree were accuracy ( 1 –misclassification error ) and its 95% confidence intervals ( 95% CI ) , sensitivity , specificity , positive predictive ( PPV ) and negative predictive values ( NPV ) . The statistical significance was considered when the p-value was < 0 . 05 . Data were analyzed using SPSS version 23 . 0 , and R software version 3 . 3 . 2 . Meta-analyses were performed using Comprehensive Meta-analysis ( CMA ) software version 3 ( Biostat , NJ , USA ) when there was more than one study . Dichotomous variables were analyzed to compute pooled event rate ( ER ) . A fixed-effect model [33] was used when there is no evidence of a heterogeneity between studies , otherwise , a random-effects model was chosen [34] . Heterogeneity between studies was evaluated using the Q statistic and I2 test which describes the percentage of variability in the effect estimates that is because of heterogeneity beyond sampling error [34 , 35] . To evaluate the presence of publication bias , we performed Begg’s funnel plot [36] and Egger’s regression test [37 , 38] when there were five or more studies in the analysis . The publication bias was considered significant when the p-value was < 0 . 1 . If the publication bias was found , the trim and fill method of Duvall and Tweedie was performed by adding studies that appeared to be missing [39 , 40] to enhance the symmetry [41] . The adjusted pooled effect size and its 95% CI were computed after the addition of potential missing studies .
A 13-year-old girl was admitted to Children’s Hospital No . 2 in Ho Chi Minh City in Vietnam on 31st August 2016 because of fever for seven days . On admission , she was totally alert and had a low grade of fever . The examination found no focal neurologic deficits . Her total blood count showed that white blood cell ( WBC ) and platelet counts were 11 , 000/μL and 182 , 000/μL , respectively . C-reactive protein level was 3 mg/L . On the 8th day ( the day after admission ) of her disease course , she recovered from fever but first began to complain of no passage of urine . She was found to lose the sensation of urinating and have urinary retention . She then needed insertion of the indwelling urinary catheter . On the 9th day ( the 2nd day of admission ) , the serologic test revealed that serum dengue IgM was positive . Serum PCR dengue was negative . Laboratory tests of liver and renal functions and electrolytes did not show any particular abnormality . On the same day , she could not move her legs , began to lose her consciousness and showed signs of confusion . The neurological examination then found that she became lethargic and quadriplegic with no abnormal sign of cranial nerves . She had normal ocular fundus , opened her eyes with painful stimulation , answered her name and then drifted back to sleep . Muscle strength and tendon reflexes of the upper extremities were 2/5 and 2+ , respectively , with the weakness of upper limbs ( UL ) . The sensory functions were nearly intact . The muscle strength and tendon reflexes of the lower extremities were 0/5 and 2+ , respectively , with complete paraplegia of lower limbs ( LL ) . Bilateral ( B/L ) Hoffmann and Babinski tests were positive . Cerebrospinal fluid ( CSF ) collected on 10th day showed pleocytosis ( 61 cells/mL ) ; elevation of protein , 1 . 68 g/L; glucose , 0 . 43 g/L; chloride , 138 mmol/L; lactate 4 . 23 , mmol/L . Her CSF was positive for ELISA dengue IgM but negative for ELISA Japanese encephalitis virus IgM . PCR Zika virus in blood and in urine was negative . Magnetic resonance imaging ( MRI ) of the brain and spinal cord did not show any particular abnormality ( Fig 1 and S1–S3 Figs ) . An electromyogram showed that motor and sensory functions were normal on both UL and LL ( S4 and S5 Figs ) . From her clinical course and laboratory tests , she was diagnosed as ADEM following dengue infection without warning signs . For treatment of ADEM , high dose of methylprednisolone ( 30mg/kg/day ) for five days was given , beginning on the 3rd of September . The oral low dose of prednisone ( 1 mg/kg/day ) was then used for four weeks . Her alertness improved gradually . On the fourth day of methylprednisolone course , she opened her eyes responding to voice , oriented and answered word by word correctly . Although urinary retention remained , her muscle strength of upper and lower extremities increased to 4/5 and 2/5 , respectively . After two weeks of oral prednisolone , limbs weakness was significantly improved and after four weeks , sphincter function was back to normal . She was nearly intact with the proximal part of her legs had a mild weakness when she was discharged from the hospital after four weeks of admission ( five weeks since fever onset ) . From nine databases , we identified 690 potentially relevant publications . After excluding duplicates and screening titles and abstracts , we retrieved 34 articles for full-text review . Of these , 25 articles met our inclusion criteria . Four additional articles , from manual search , were identified . Finally , 29 articles were in this systematic review including; 15 case reports , 10 case series , 1 literature review with 1 case , and 3 literature reviews ( Fig 2 ) . The total sample size was 1 , 163 dengue patients including 165 patients with neurological complications . Of those 165 patients , there were 29 ADEM cases including three cases of rare types from ADEM; one case with neuromyelitis optica [42] , two cases with meningoencephalitis [43 , 44] , and our new case . The characteristics of the 29 cases are shown in Table 1 . Among 29 ADEM patients , there were three cases were described in a group of neurological manifestations with limited individual data [43 , 45] . There were more males than females ( 18 and 8 , respectively ) , the median age was 20 ( range 9 days to 65 years ) . Most of the dengue cases were diagnosed based on IgM followed by IgG . The three included literature reviews , in general , discussed the pathogenesis of DENV and its accompanying neurological complications , their pathogenesis , and their incidence . The first review discussed the neuropathogenesis of DENV illness , its neurological complications , the diagnosis , and treatment of these diseases . It discussed also the epidemiology of DENV and the increasing prevalence and incidence of the disease and its extension to new countries . The neuropathogenesis of DENV includes three ways; the metabolic disturbance causing encephalopathy , direct central nervous system invasion ( especially , by DENV-2 and -3 ) causing mainly encephalitis , and autoimmune reaction mechanism . The neurological complications discussed included encephalitis and meningitis , being the most common complication and caused by direct invasion , ADEM , and its rare type neuromyelitis optica , by an immune-mediated process , myelitis , either by an immune-mediated mechanism or by direct invasion , GBS , and mononeuropathies , by autoimmune mechanisms , and myositis [46] . While the second one discussed the various neurological complications , their diagnosis , and the treatment . The neurological complications included dengue encephalopathy , describing it as the most commonly reported neurological disorder associated with DENV and stating that in a retrospective study in Indonesia , 6% ( 152 ) of patients with DHF were admitted with encephalopathy . Encephalitis was described in five studies . Post-dengue immune-mediated diseases were discussed and included acute transverse myelitis , GBS , ADEM , and its rare type neuromyelitis optica . Also , cerebrovascular complications ( with unknown incidence ) and dengue muscle dysfunction ( ranging from 66 to 100% in different studies ) and neuro-ophthalmic complications ( about 10 to 40% in different studies ) were described [47] . Finally , the last one discussed the epidemiology , transmission of DENV , its clinical manifestations , its neurological complications and their pathogenesis , the diagnosis , and management of the disease . The pathogenesis of the neurological complications included immune-mediated reactions , metabolic disturbance , and direct invasion . This study described encephalopathy as the most common neurological manifestation of DENV infection . Other neurological complications were encephalitis , myelitis , GBS , myositis , and hypokalemic paralysis . However , ADEM was described as a rare complication [48] . The prevalence of neurological disorders ( n = 27 ) among dengue patients ( n = 1 , 024 ) in two studies was 2 . 6% [1 . 8–3 . 8%] without an evidence of heterogeneity ( Fig 3A ) . Pooling two studies enrolling all available dengue patients ( n = 1 , 024 ) revealed that the prevalence of ADEM ( n = 3 ) among dengue patients was 0 . 4% [0 . 1–2 . 5%] with a moderate heterogeneity ( Fig 3B ) . In seven studies recruiting dengue patients ( n = 144 ) with neurological disorders ( n = 8 ) , the prevalence of ADEM was 6 . 8% [3 . 4–13%] without an evidence of heterogeneity nor publication bias , Egger’s test ( p-value = 0 . 8 ) ( Fig 3C ) . We could only analyze manifestations in 26 cases due to the lack of information in three cases [43 , 45] . The onset day of neurological manifestations after initial dengue symptoms ranged from day 3 to day 19 ( median = 7 ) . Most of the cases had a fever on the ADEM onset ( 25 cases ) and 4 cases were not described ( ND ) . The reasons for ADEM admission were fever , vomiting , urination problems , arthralgia , seizures , walk problems , chills , altered sensorium , asthenia , hemiparesis , vision problems , paraparesis , paresthesia and hyperreflexia , abdominal pain , thrombocytopenia , lethargy , poor feeding and seizures , weakness , headache , rigors , and/or myalgia . The most frequent manifestations and signs related to dengue were fever ( 22/26 , 85% ) , thrombocytopenia and vomiting ( 13/26 , 50% ) , headache ( 11/26 , 42% ) , erythema/rash ( 9/26 , 35% ) , myalgia ( 8/26 , 31% ) , arthralgia ( 6/26 , 23% ) , chills ( 5/26 , 19% ) , leukocytopenia and restless ( 4/26 , 15% ) then retro-orbital pain , rigors and lethargy ( 3/26 , 12% ) . While the most frequent manifestations and signs related to ADEM were altered sensorium/consciousness ( 15/26 , 58% ) , seizures and urination problems ( 9/26 , 35% ) , vision problems ( 8/26 , 31% ) , slurred speech ( 6/26 , 23% ) , walk problems ( 4/26 , 15% ) then ataxia ( 3/26 , 12% ) ( Fig 4 and S2 Table ) . Liver function tests were normal in 5 cases , abnormal then normal in 1 case , abnormal in 9 cases and ND in 14 cases . The renal function tests were normal in 3 cases , 1 case had urea: 46 mg/dl and creatinine 1 . 1 mg/dl , 1 case had acute kidney injury and metabolic acidosis and ND in 23 cases . Urinalysis was normal in 3 cases and ND in 26 cases . Chest-X rays were normal in 4 cases , suggestive of acute respiratory distress syndrome ( ARDS ) in 1 case and B/L fluffy shadows as well as ARDS in another case and ND in 23 cases . The cardiovascular system examination was normal in two cases and ND in 27 cases . The pulses per min were normal in most cases , median ( range ) was 110 per minute ( 92–140 ) , stable in 2 cases but ND in 9 cases . The abdominal findings were normal in 1 case , distended urinary bladder , splenomegaly 2 cm and hepatomegaly 3 cm in another case , markedly distended and palpable urinary bladder and spleen tip palpable , 1 case had pain , mild hepatomegaly in another case and ND in 24 cases . The arterial blood pressure ( ABP ) was normal in most cases , median ( range ) of systolic blood pressure was 100 mmHg ( 90–130 ) while of diastolic blood pressure was 60 mmHg ( 60–80 ) and ND in 19 cases . The tourniquet test was positive in 3 cases and ND in 26 cases . The respiratory rate was 28/min in 1 case , 30/min in 1 case and stable in another case and ND in 26 cases ( S3 Table ) . The results from CSF analysis showed that most of the cases have elevated protein levels ( 13 cases ) , normal glucose ( 11 cases ) , pleocytosis ( 6 cases ) , positive for ELISA dengue IgM in 2 cases and ND in 12 cases . The results from MRI of brain and spinal cord showed that most of the cases have abnormalities such as T2 lesions , demyelination 2 cases and B/L hemorrhagic demyelination in 1 case , cervical/proximal dorsal cord edema deep white matter , cortical and Pontine swellings in 1 case , cervical/proximal dorsal cord edema in 1 case , cervicodorsal cord swelling in another case and no description for MRI of spinal cord in 20 cases while in 7 cases only for MRI of brain ( S4 Table ) . The power grade ranged from 1/5 to 5/5 . Moreover , deep tendon reflexes ( DTRs ) were normal ( 3 cases ) , increased/hyperreflexia ( 3 cases ) or brisk ( 3 cases ) , absent/hyporeflexia ( 2 cases ) and ND ( 18 cases ) . Plantar responses were B/L extensor ( 9 cases ) , B/L Babinski sign ( 1 case ) , flexor ( 1 case ) , B/L Hoffmann and Babinski signs ( 1 case ) , left Babinski sign and absent response in right foot ( 1 case ) nonresponsive B/L ( 1 case ) and ND in 15 cases . Cranial nerves were normal ( 6 cases ) , unable to be examined ( 3 cases ) facial nerve palsy ( 1 case ) , B/L ptosis ( 1 case ) and ND ( 18 cases ) . Furthermore , the ULs and LLs were abnormal in most cases , normal in 1 case only and ND in 11 cases for LLs and cases for ULs 13 cases . The motor system was ND in 25 cases and has abnormalities ( hemiparesis or quadriparesis ) in the remained ones ( S5 Table ) . The blood analysis showed an increased WBC in most cases , median ( range ) of WBC = 45 × 108/L ( 0 . 011–31×109 ) , while of platelet was 60 × 109 ( 1 . 1 × 109–328 × 109 ) , of hemoglobin ( Hb ) was 107000 mg/L ( 112–1620000 ) , of Glasgow coma scale ( GCS ) was 7 ( 6–9 ) , of PCV was 38 . 1% ( 30 . 1–48 . 4 ) , of alanine transaminase ( ALT ) was 214 U/L ( 36 . 3–123000 ) , of aspartate transaminase ( AST ) was 185 U/L ( 44–199000 ) , of albumin was 3 . 05 g/dL ( 2 . 3–4 . 2 ) , of creatine was 1 . 2 mg/dL ( 0 . 6–4 . 3 ) , of glucose was 103 . 8 mg/dL ( 69–138 ) , of urea was 46 mg/dL ( 21 . 9–102 ) ( S6 Table ) . The optic nerves were unable to be examined ( 3 cases ) , normal ( 3 cases ) , B/L ptosis ( 1 case ) , optic neuritis ( 1 case ) , B/L involvement ( 1 case ) , moderate B/L optic atrophy ( 1 case ) and ND ( 19 cases ) . The fundus examination was normal in 9 cases , showed pallor of optic discs in 1 case , B/L papilledema which was more severe in the right eye in another case and ND in 18 cases . Pupils have normal size and reaction to light in 5 cases , sluggish reaction to light in 1 case , mid-dilated and equal in size with sluggish reaction to light in 1 case , mid-dilated , symmetrical with sluggish reaction to light in 1 case , dilated in 1 case , B/L mid-dilated , symmetrical and sluggishly reacting to light in another case and ND in 19 cases . The visual acuity was deteriorated in left eye then in right eye then in both eyes gradually deteriorated in 1 case , another case had a severe visual impairment in right eye ( only light perception ) and a slight visual disturbance in left eye ( VA = 20/25 ) , 1 case with a severe right visual impairment and ND in 26 cases ( S7 Table ) . The follow-up period ranged from 28 days to 5 years ( S2 Table ) . Most specific treatments used for ADEM were oral or intravenous ( IV ) corticosteroids including methylprednisolone ( 11 cases ) , prednisolone ( 7 cases ) and dexamethasone ( 5 cases ) or human immunoglobulin ( 1 case ) . Other treatments used were anticonvulsant medications such as phenytoin ( 2 cases ) and phenobarbitone ( 1 case ) , oral or intravenous antipyretics and anticonvulsant such as dipyrone , paracetamol , pulse therapy , dopamine , noradrenaline , and lorazepam ( 1 case for each treatment ) . The outcomes in these cases were either death ( 3 cases ) , partial recovery ( 7 cases ) , complete recovery ( 16 cases ) or ND ( 3 cases ) . The cases with partial recovery were either; had mild B/L visual disturbance , dysuria , and dyschezia remained [49] , was able to walk with a minimal support [50] , wanted to carry further treatment in the hospital [51] , a slight residual cerebellar ataxia [52] , the frontal symptoms persisted [53] , mild ataxia and dysarthria [54] . The three cases died due to; myalgia , jaundice , conjunctival hemorrhage , hematuria , oliguria , shortness of breath , became stuporous , acute respiratory distress syndrome ( ARDS ) , acute kidney injury and metabolic acidosis [55] , intracranial tension [56] or B/L hemorrhagic demyelination [57] ( Table 2 ) . Our results showed that the body temperature levels in the complete recovery group were significantly lower than those of the partial recovery ( p-value = 0 . 026 ) and bad outcomes groups ( p = 0 . 03 ) . While there was a significant difference between cases having complete recovery or bad outcomes on the onset day of neurological manifestations which was found started earlier in cases having partial recovery ( p = 0 . 03 ) and bad outcomes ( p = 0 . 006 ) as compared to patients with complete recovery . Other factors including gender , steroid treatment , and its administration route , age , hemoglobin , platelet , WBC , GCS , PCV , ALT , AST , pulse , systolic blood pressure , diastolic blood pressure , urea , glucose , creatinine or albumin were not associated with the ADEM outcomes ( Table 3 ) . We then selected all of the aforementioned 20 variables to build classification tree models for bad outcomes ( partial recovery or died ) versus complete recovery and for partial recovery versus complete recovery . Interestingly , that both classification trees including the onset day of neurological manifestations ( the cut-off point at 9 . 5 days ) , and temperature ( the cut-off point at 101 . 2°F ) are the best models ( Fig 5 ) . Both trees enhance the results from uni-variable analysis indicating that the earlier onset day of the neurological manifestations ( < 9 . 5 days ) and higher fever when presenting ADEM ( ≥ 101 . 2°F ) were associated with the bad outcomes and partial recovery . The performance of tree that classified bad outcomes versus complete recovery is at the accuracy of 84 . 6% [65 . 1–95 . 6%] with the sensitivity of 80% , specificity of 87 . 5% , PPV of 80% and NPV of 87 . 5% . Whereas , the performance of tree that classified complete recovery versus partial recovery is at the accuracy of 82 . 6% [61 . 2–95 . 1%] with the sensitivity of 87 . 5% , specificity of 71 . 4% , PPV of 87 . 5% and NPV of 71 . 4% ( Fig 5 ) .
Our meta-analysis revealed that the prevalence of patients with neurological disorders within dengue patients is not rare ( 2 . 6% ) . Though ADEM was reportedly stated as a rare condition [57–59] , the incidence could be higher because of the high global burden of dengue infection . In a previous study , it revealed that dengue was present in 4–47% of patients with encephalitis in the endemic regions [60] . It is well-known that encephalopathy is the most common neurological disorder accompanying DENV infection [47 , 48] . Thus , this should push health cares to estimate the total global cases of ADEM from the annual incidence of dengue because the number of post-dengue ADEM is underestimated probably due to the neglect of the clinicians and patients , hence our current proposal is to screen all patients with neurological manifestations against dengue and most other flaviviruses such as Zika to investigate the number of ADEM cases in a multi-center study in Vietnam and Philippines . Moreover , we suggest adding such neurological complications in the dengue WHO guidelines , so they get no neglect . The onset of ADEM ranged from day 3 to day 19 from dengue infection . The clinicians should be aware that patients can present with early or late onset of ADEM symptoms and patients may not always mention a recent history of fever . Moreover , there was a significant difference between cases having complete recovery or bad outcomes only in two factors which were the onset day of the neurological manifestations being earlier and the temperature being higher in cases having bad outcomes or partial recovery through our uni-variable analysis . This finding is supported by classification tree models including the onset day of the neurological manifestations and temperature . Both trees indicate that the earlier onset day of the neurological manifestations ( < 9 . 5 days ) and higher fever when presenting ADEM ( ≥ 101 . 2°F ) were associated with the bad outcomes . These findings require an attention from physicians regarding the temperature of the dengue cases to be managed well once elevated . The most frequent manifestations related to dengue infection arranged from the most frequent to the least frequent were; fever , thrombocytopenia , and vomiting , headache , erythema/rash , myalgia , arthralgia , chills , leukocytopenia , and restlessness then retro-orbital pain , rigors , and lethargy . Noteworthy , vomiting , rash , and leukocytopenia are classified as dengue without warning signs in the WHO 2009 guidelines while thrombocytopenia , restlessness , and lethargy are classified as dengue warning signs [1] . While the main manifestations related to ADEM arranged from the most frequent to the least frequent were; altered sensorium/consciousness , seizures and urination problems , vision problems , slurred speech , walk problems , then ataxia . Similarly , altered consciousness is classified within the severe dengue signs , in the severe CNS involvement , in the WHO 2009 guidelines . However , other ADEM manifestations are not mentioned in it [1] , maybe because ADEM sometimes appears late after dengue . Hence , we suggest adding them to the guidelines of severe organs involvement stage . The results from MRI of the brain and spinal cord showed that most of the cases have abnormalities such as T2 lesions . In contrast to our results which showed no abnormalities . The attention for the MRI of either the brain or spinal cord findings should be paid more due to its unlimited importance in diagnosis and treatment of ADEM cases . Most of the outcomes in these cases were relatively good because most of them showed either partial recovery or complete recovery . There was no significant difference between cases with bad outcomes or complete recovery in the treatment used . Unlike a previous literature review which suggested that steroids are promising in the treatment of ADEM during its active phase [61] . Till now , there is no study described the mechanism of post-dengue ADEM . The neurological complications of dengue infection have been considered to be due to systemic complications of dengue and not related to its neurotropic nature [48 , 61–65] . After the demonstration of neural tropism of dengue virus , the neurological manifestations of dengue infection are categorized as ( 1 ) related to direct neurotropic effects of the virus ( myelitis , meningitis , myositis , rhabdomyolysis , and encephalitis ) , ( 2 ) related to systemic or metabolic complications of dengue ( encephalopathy , stroke ) and ( 3 ) post-infectious immune-mediated complications ( GBS , transverse myelitis , ADEM ) . ADEM usually occurs following a viral infection but may appear spontaneously , after bacterial , parasitic infection or vaccination . Most cases follow a nonspecific upper respiratory tract infection . Although it occurs in all ages , most reported cases are in infants and adolescents [66] . The post-infectious ADEM usually begins late in the course of viral infections including measles , chickenpox , mumps , rubella , influenza , EBV and nonspecific respiratory infections . The pathophysiology involves a transient auto-immune response directed at myelin or other self-antigens , possibly by a non-specific activation of auto-reactive T-cell clones or by molecular mimicry [63 , 67 , 68] . As with other viruses , the pathogenesis underlying dengue-associated ADEM may result from an immune system mediated-process [25] . A limitation of this study was the small number of included studies ( a total of 29 ADEM cases ) and reported cases , with some missing values , included in the uni-variable analysis , meta-analysis , and the classification tree models . Moreover , our results should be interpreted with caution because most cases depended on IgM ELISA which has a probable diagnosis [1] but with a high specificity [1 , 69–74] . In conclusion , our analysis of the case report and other included cases revealed that the onset day of neurological manifestations and temperature in the ADEM patients were associated with the disease outcome and can predict it . Moreover , we found that the most frequent dengue manifestations were fever , thrombocytopenia , vomiting , and headache while the most frequent ADEM manifestations were altered sensorium/consciousness , seizures urination problems , and vision problems . The serious manifestations after dengue infection continue to be reported . These manifestations should be considered in the diagnosis and management of patients with dengue infection . The prevalence of ADEM among dengue and other dengue-related neurological disorders is not too rare . Since the incidence of ADEM is not known well , future larger studies are necessary to accurately investigate ADEM .
|
We presented a 13-year-old girl of ADEM following dengue infection . She was totally alert and had a low grade of fever with no focal neurologic deficits , on admission . We revealed that the prevalence of either ADEM or all neurological disorders among dengue patients was not too rare . Moreover , we found that the most common manifestation of ADEM was altered sensorium/consciousness followed by seizures and urination problems then vision problems . These manifestations should be considered in the diagnosis and management of dengue-infected patients . Also , this requires shedding the light on the total global cases of ADEM from the annual incidence of dengue . The onset of ADEM can be early or late after dengue infection . Hence , clinicians should pay attention that it can be early or late that patients can forget about their fevers . Moreover , the onset day of neurological manifestations and patients’ temperature were significantly associated with the disease outcome .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"diagnostic",
"radiology",
"pathology",
"and",
"laboratory",
"medicine",
"vomiting",
"tropical",
"diseases",
"headaches",
"magnetic",
"resonance",
"imaging",
"urine",
"physiological",
"processes",
"signs",
"and",
"symptoms",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"imaging",
"techniques",
"dengue",
"fever",
"pain",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"anatomy",
"fevers",
"physiology",
"neurology",
"biology",
"and",
"life",
"sciences",
"myalgia",
"viral",
"diseases"
] |
2017
|
Post-dengue acute disseminated encephalomyelitis: A case report and meta-analysis
|
The characterization of functional elements in genomes relies on the identification of the footprints of natural selection . In this quest , taking into account neutral evolutionary processes such as mutation and genetic drift is crucial because these forces can generate patterns that may obscure or mimic signatures of selection . In mammals , and probably in many eukaryotes , another such confounding factor called GC-Biased Gene Conversion ( gBGC ) has been documented . This mechanism generates patterns identical to what is expected under selection for higher GC-content , specifically in highly recombining genomic regions . Recent results have suggested that a mysterious selective force favouring higher GC-content exists in Bacteria but the possibility that it could be gBGC has been excluded . Here , we show that gBGC is probably at work in most if not all bacterial species . First we find a consistent positive relationship between the GC-content of a gene and evidence of intra-genic recombination throughout a broad spectrum of bacterial clades . Second , we show that the evolutionary force responsible for this pattern is acting independently from selection on codon usage , and could potentially interfere with selection in favor of optimal AU-ending codons . A comparison with data from human populations shows that the intensity of gBGC in Bacteria is comparable to what has been reported in mammals . We propose that gBGC is not restricted to sexual Eukaryotes but also widespread among Bacteria and could therefore be an ancestral feature of cellular organisms . We argue that if gBGC occurs in bacteria , it can account for previously unexplained observations , such as the apparent non-equilibrium of base substitution patterns and the heterogeneity of gene composition within bacterial genomes . Because gBGC produces patterns similar to positive selection , it is essential to take this process into account when studying the evolutionary forces at work in bacterial genomes .
Comparative genomics is a fundamental key to the inner workings of genomes . The identification of genes and other functional elements such as regulatory regions , as well as the understanding of their influence on the fitness of organisms rely essentially on the detection of signatures of natural selection within genomes [1] . In that respect , devising a model of sequence evolution in the absence of selective constraints ( a neutral model ) is critical for the detection of functional sequences . Indeed , to explain the features of a given genomic segment , comparing the fit of a neutral model to that of a model that also invokes selection ( either purifying or positive ) is the operational way to infer evolutionary constraint and hence function . The base composition of genomic sequences varies widely , both across species and along chromosomes [2 , 3] . For instance , the genomic GC-content of cellular organisms ranges from 13% to about 75% [4 , 5] , with vast intra-genomic heterogeneity . These large-scale variations in base composition affect all parts of genomes , intergenic regions and genes—including all three codon positions [6]—and hence cannot be simply explained by selective constraints on the encoded proteins . Determining the underlying causes ( selective or neutral ) of these variations in GC-content is a major issue in genetics: if they result from selection , it implies that the genomic base composition per se is an important trait that contributes to the fitness of organisms; conversely , if these “genomic landscapes” are largely shaped by non-adaptive molecular processes , then characterizing these processes is essential for the reliable detection of selection ( see e . g . [7] ) . In mammals , the analysis of polymorphism data and substitution patterns along genomes demonstrated that the evolution of GC-content is driven by recombination , which tends to increase the probability of fixation of AT→GC mutations [8 , 9] . The impact of recombination on base composition in these genomes is most probably due to a phenomenon known as GC-biased gene conversion ( gBGC ) , which favours G/C nucleotides at polymorphic sites in the conversion of intermediates of recombination ( see review in [10] ) . Although gBGC as a process is unrelated to natural selection , it affects the probability of fixation of alleles in patterns similar to selection [11] . It has been shown to be an important confounding factor , which can mimic some marks of positive selection [7 , 12] and interfere with selection by actively promoting the fixation of deleterious alleles [13 , 14] . The process of gBGC has been observed directly in meiosis products from yeast and human [15 , 16] , and there is ample evidence , based on the analysis of relationships between recombination rate and substitution patterns within genomes , that this process affects many other eukaryotes [17–19] . In Bacteria and Archaea , several environmental factors potentially affecting genomic GC-content have been proposed ( such as the availability of oxygen or nitrogen in the environment , growth temperature , or the variety of environments encountered by an organism , see for instance [20] and ref . therein ) . Because these effects are weak and the nature of the selective pressures remain elusive , the major force driving genomic GC-content has long been considered to be mutational bias [21] . Recently however , two independent analyses have shown that in virtually all Bacteria , independently of their genomic GC-content , there is an excess of G/C→A/T mutations [22 , 23] . This suggests that an unknown process , selective or neutral , is opposing this universal mutational bias by favouring the fixation of G/C alleles Previously , an analysis of a large number of E . coli genomes had suggested a possible role of gBGC , based on the link between GC-content , recombination and the organization of the chromosome in this species [24] . However Hildebrand et al . [23] observed that the excess of G/C→A/T mutations was still present after removing datasets with evidence of recombination . Moreover they found no correlation between GC-content and recombination rate across bacterial species . They therefore concluded that this force could not be gBGC and hence that selection was driving an increase of genomic GC in Bacteria . The nature of this selective advantage remains however mysterious , though various hypotheses have been proposed [25 , 26] . Here we argue that the analyses performed by Hildebrand et al . [23] are not conclusive regarding the gBGC hypothesis , and we present evidence that variations in GC-content observed in Bacteria are influenced by gBGC . One pervasive signature of gBGC is that genomic regions undergoing high recombination rates will also acquire a high GC-content [6] . We thus studied the relationship between recombination and GC-content in 20 groups of Bacteria and one group of Archaea . This dataset covers a wide range of clades representative of the bacterial diversity . To avoid problems inherent to comparisons of recombination rates among species ( such as differences in polymorphism , genome samples , population size , mutation rates , an other life history factors ) , we examined the intragenomic variability for both recombination and GC-content . We show that in a wide variety of bacterial species , genes with evidence of recombination have a higher GC-content . We further show that this bias towards G/C nucleotides in recombining genes cannot be explained by selection on codon usage , and could interfere with the selection for AT-ending optimal codons . These two observations strongly suggest that homologous recombination , via gBGC , is a crucial factor universally influencing the nucleotide content of genes and genomes . If confirmed , gBGC can account for several pervasive yet unexplained features of bacterial genomes . Finally , we emphasize that because gBGC has the ability to both mimic and interfere with natural selection , gBGC must be considered by future studies geared at understanding processes driving bacterial genome evolution .
In Bacteria , recombination occurs in the form of gene conversion ( i . e . unidirectional transfer of genetic material from a donor sequence towards a homologous recipient sequence ) . To detect past gene conversion events in bacterial species , it is necessary to compare closely related genomes . We therefore selected in the database of homologous gene families HOGENOM ( release 6 ) [27] all groups of closely related species or strains encompassing at least 6 sequenced genomes . This dataset contains 20 bacterial groups and one archaeal group . For each gene family represented in these groups , we computed i ) the average GC-content at different positions of codons and ii ) the index of recombination provided by PHI [28] based on alignments of standardized length ( see methods for details ) . PHI is a rapid method for detecting recombination in multiple alignments at the scale of the gene , which has been shown to be more robust than most methods to variations in recombination rates , sequence divergence and population dynamics [28] . We used this test to determine if homologous gene families had experienced gene conversion events among members of the taxa of interest . One important feature of this test is that it measures whether there is sufficient phylogenetic signal in an alignment to tell if recombination has occurred . Only alignments with sufficient signal , whether recombinant or non-recombinant were retained for tests in the remaining of this study . We also used three other approaches for detecting recombination , and these confirm the robustness of our conclusions ( see Methods and Supplementary Material ) . In Eukaryotes , a general relationship between various estimates of recombination rate and the GC% of genes has been documented and provides indirect evidence for gBGC . Our first goal was to test this prediction in Bacteria and Archaea . To exclude a potential effect of the number of genes in the alignment on our estimates of recombination ( because alignments with more sequences are expected to give more power to detect recombination ) , we focused on single-copy genes of the core genome ( i . e . genes that are present in only one copy and found in each genome of a group ) . In 7 of the 21 groups , the proportion of single-copy genes of the core genome with evidence for recombination was very low ( <2% of all gene alignments tested ) , suggesting that these species are clonal or nearly so ( Table 1; shaded datasets in Fig . 1 ) . Indeed , the Burkholderia pseudomalei group , Chlamydia trachomatis , Francisella tularensis , Mycobacterium tuberculosis and Yersinia pestis are species known to be pathogenic clonal complexes with low polymorphism and probably very low recombination [29–33] , while Brucella spp . and Sulfolobus spp . are likely composed of ecologically isolated clades , because of their respective lifestyle as obligate intracellular pathogen or ecotypes endemic of hot springs [34–36] . In 11 of the 14 remaining groups , we found a significant positive difference in average GC-content at all and/or at the third position of codons ( GC3 ) between recombinant and non-recombinant genes ( Fig . 1 ) . In these 11 species , the difference in GC3 is always larger than that at all positions , suggesting that the effect of recombination on gene composition is stronger at synonymous positions ( probably because of purifying selection on protein sequences ) . Two notable exception to this pattern are i ) the bacterial species Helicobacter pylori , where GC-content seems to be lower in recombining genes and ii ) the Bacillus anthracis/cereus group , where GC at all positions and GC3 display opposite patterns , with GC3 being higher in recombining genes . Consistent results are obtained using alternative recombination detection methods ( S1 Fig . ) . Recombination is known to enhance the efficacy of selection by breaking linkage between neighboring selected sites . It is therefore possible that selection is more efficient in recombining genes . This effect ( referred to as Hill-Robertson interference ) should theoretically be more pronounced in the case of selection on codon usage [37] , which is relatively weak compared to selection on amino acid sequences . Thus recombination—in the absence of any gBGC—can potentially explain the pronounced effect observed on GC3: if Hill-Robertson interference leads to a higher frequency of optimal codons in highly recombining genes , and if optimal codons tended to be GC-rich , this effect could explain a relationship between GC-content and recombination . The “selection model” sketched above predicts that the frequency of all optimal codons ( both GC-ending and AU-ending ) should increase with recombination . In contrast , a model incorporating the effect of gBGC predicts that GC-ending codons ( not specifically optimal codons ) should be enriched in recombining regions , and that AU-ending codons ( and possibly AU-ending optimal codons if gBGC is strong enough to override selection on codon usage ) should display the opposite pattern . We therefore looked specifically at the frequency of the different types of codons , i . e . optimal and non-optimal , in recombining and non-recombining genes . There is a debate over the best way to define optimal codons , based on their over-representation in either ribosomal protein genes ( RP ) , or genes with the highest codon bias ( HCB ) [38–40] . We therefore analyzed the frequency of GC-ending and AU-ending optimal codons ( FopGC and FopAU ) and non-optimal codons ( FnopGC and FnopAU ) according to both RP and HCB definitions ( Fig . 2 , S3 Fig . , and S4 Fig . , respectively ) . The higher GC3 of recombining genes means that GC-ending codons are over-represented in recombining genes , but this is true for optimal GC-ending codons in only 2 ( RP optimal codons ) or 4 ( HCB optimal codons ) species out of 11 . This effect is hence essentially due to non-optimal codons ( FnopGC is significantly higher in recombining genes than non-recombining genes in respectively 9 and 8 species for RP and HCB definitions ) . Moreover , optimal AU-ending codons are significantly depleted in recombining genes for 8 ( resp . 5 ) species for RP ( resp . HCB ) codons . In fact , only two species , S . pyogenes and Nesseiria meningitidis ( using the HCB method—only S . pyogenes using the RP method ) exhibit a pattern partially compatible with the selection hypothesis presented above . All species display either an increase of FnopGC and/or a decrease of FopAU in recombining genes , a fact that cannot be explained by a higher efficiency of selection . This pattern excludes the possibility of pervasive selection for codon usage promoting a better adaptation to the pool of tRNA for genes in regions of high recombination , but is compatible with the predictions of gBGC . In principle , gBGC should affect all genomic regions where recombination occurs , including intergenic regions . Intergenes are generally shorter than coding regions . Furthermore , they evolve more rapidly and hence are more difficult to align . Hence , the methods that we used to detect recombination in coding regions cannot be applied with intergenes . We therefore used the recombinant or non-recombinant status of the neighboring protein-coding genes as a proxy of the status of the intergenes . In 11 of the 14 taxonomic groups , we observed that intergenes flanked by recombining genes have a higher GC-content than intergenes flanked by non-recombining genes ( S2 Fig . ) . The difference in GC-content between the two classes is weaker than that observed in coding regions , and when considered individually , only one comparison is statistically significant ( Streptococcus pyogenes , p < 0 . 01 ) . This is possibly because the prediction of recombination status of intergenic regions is indirect ( based on the status of flanking genes ) and therefore less accurate than that of coding regions . However , the number of cases where the difference in GC-content between the two classes is positive ( 11 out of 14 comparisons ) is significantly higher than expected by chance ( Chi-squared test: p = 0 . 03 ) . To quantify the relationship between recombination and base composition , we first analyzed the genome of S . pyogenes , one of the species for which the signature of gBGC is strong ( Fig . 1 , see also Sup . Mat . ) . We used ClonalOrigin [41] to compute the population-scaled recombination rate ( rho ) for each gene of the core genome . The correlation between rho and the GC-content at third codon position of each gene ( GC3 ) is slight but significant ( R2 = 0 . 034 , p<10−4 , n = 478 ) . Interestingly , when we exclude genes for which the estimate of rho is less reliable ( about 10% of the data , see Sup . Mat . ) , the correlation strongly increases ( R2 = 0 . 087; p<10−9 , n = 437 ) . Due to the low amount of data available in a gene-scale alignment , the measure of rho is expected to be noisy . Thus , the observed correlation is probably an underestimate . To try to get more robust estimates of rho , we binned the dataset into 20 groups of genes according to their GC3 , and we computed the correlation between the average GC3 and the average rho of each bin . Using this approach , we observed a strong correlation between the GC3 and recombination ( R2 = 0 . 60; Fig . 3A ) . For a comparison , we performed a similar analysis in humans: we randomly selected 600 human genes ( which corresponds to the average number of genes analyzed in our bacterial data sets ) , binned the data set into 20 groups of genes according to their GC3 , and we computed the correlation between the average GC3 and the average recombination rate as obtained from population-wide surveys [42] . The average correlation ( computed after repeating the random sampling 1 , 000 times ) is R2 = 0 . 55 ( with 95% of the R2 values in the interval [0 . 26–0 . 78]; one representative example is presented in Fig 3A ) . Using this binning approach , we noted that the proportion of genes in a bin that are detected as recombinant by PHI is strongly correlated with the average value of rho ( R2 = 0 . 70; p<10−5 ) . This suggests that this index ( hereafter referred to as PREC , for ‘proportion of recombinant’ ) is a good proxy for the average recombination rate in a bin . In fact , we observed that GC3 correlates more strongly with PREC ( R2 = 0 . 68; Fig . 3B ) than with rho . Given that the computation of rho with ClonalOrigin is extremely time consuming , we decided to use PREC to evaluate the correlation between GC3 and recombination rate in other species . Correlations were positive and significant for 11 of the 14 species , with R2 values ranging from 0 . 24 to 0 . 68 , and we did not observe any significant negative correlation ( Fig . 3B ) . This shows that the correlations between GC-content and recombination rate in bacteria are of similar magnitude to what is observed in humans .
Our results suggest that recombination affects the GC-content of genes in most bacterial phyla . We analyzed genes of the core genome to compare the base composition of genes with or without evidence of recombination . In 11 of the 14 species in which a significant amount of recombination could be detected , we observed that the GC-content ( measured either at the third codon position or along the entire coding region ) is higher among recombining genes compared to other ( hereafter labeled as “non-recombining” ) core genes . Several hypotheses have been proposed to explain the variations in GC-content among bacterial genomes [25] . Several studies have revealed that the genomic GC-content of bacterial genomes is always higher than what would be predicted from mutational bias [22 , 23 , 43] . Hence , it seems inescapable that some other evolutionary force is driving the genomic GC-content towards higher values in virtually all bacterial species , except maybe for the most AT-rich genomes [22 , 43] . Recombination is known to enhance the efficiency of selection by breaking linkage among sites . It is therefore conceivable that our results merely reveal a universal selective pressure favoring GC-rich alleles . But the mechanism underlying such selection would have to be acting more efficiently on synonymous sites than non-synonymous sites because the difference of GC% between recombining and non-recombining genes is higher at the third position of codons . This excludes potential selection on amino-acid content . One selectable trait that may influence synonymous positions is codon usage . If optimal codons tended to be GC-rich , recombination could drive GC% higher by favoring the adaptation of genes to better translation efficiency . However , we observed a higher GC-content in recombining genes even in species favoring A/U-ending codons ( Fig . 2 ) . Moreover , A/U-ending and G/C-ending codons show opposite relationships with recombination in most species , irrespective of their optimality . These observations suggest that the evolutionary force explaining our results is also largely independent from selection on codon usage . This conclusion is further supported by the fact that the relationship between GC-content and recombination is also observed in intergenic regions ( S2 Fig . ) . In fact , as suggested by Hershberg and Petrov [22] , who observed that the intergenic regions of bacterial genomes also have higher GC% than expected from their mutational pattern , it seems likely that the process is unlinked to gene expression or function . Hence , either there is selection acting simultaneously on each nucleotide of a bacterial genome to become G or C , or GC-biased gene conversion , which has now been observed in a variety of Eukaryotes is also at work in Bacteria . The hypothesis that gBGC plays a role in bacterial genome evolution has been considered previously [23] . Hildebrand et al . analyzed the correlation between genome-wide measures of recombination rate ( scaled by effective population size ) with genomic GC-content among 34 species , covering different bacterial phyla . As they did not find any significant correlation , they concluded that there was no evidence of gBGC in Bacteria [23] . However , we argue here that this observation is not conclusive . In fact , the strength of gBGC depends on four variables: the effective population size ( Ne ) , the rate of recombination per bp per generation ( r ) , the length of conversion tracts ( L ) and the intensity of the repair bias ( b0 ) ( for review , see [6] ) . In a haploid organism , the population-scaled gBGC coefficient is: B = 2 N e ⋅ r ⋅ L ⋅ b 0 ( 1 ) Similar to selection , the impact of gBGC on genome evolution depends on its intensity relative to genetic drift , and becomes negligible when B ≪ 1 . There is evidence that besides Ne and r , both L and b0 can vary strongly across species . For example , in budding yeast , when a GC/AT heterozygote site is involved in a gene conversion event , the GC-allele is transmitted with a probability pGC = 0 . 507 ( which is significantly higher than the expected Mendelian transmission ratio; [15] ) , whereas in humans , a recent analysis of gene conversions tracts associated to non-crossover recombination showed that GC-alleles are transmitted with a probability pGC = 0 . 70 [16] . Thus , the parameter b0 ( b0 = 2 ∙ pGC – 1 ) is about 30 times higher in humans than in yeast . Conversely , gene conversion tracts are on average about 4 times longer in yeast than in mammals [15 , 44] . Thus , for a same population-scaled recombination rate ( Ne r ) , the intensity of gBGC would be about 7 times stronger in humans than in yeast . This example illustrates that because of variations in L and b0 , the gBGC model does not necessarily predict a good correlation between population-scaled recombination rate and GC-content across species . In fact , to test the predictions of the gBGC model , it is more appropriate to investigate correlations between base composition and recombination rate within genomes , so that the other parameters ( Ne , L and b0 ) can be controlled for . The observations presented previously are qualitatively consistent with the gBGC model . However , they do not provide a quantification of the impact of recombination on bacterial genomes: to what extent might this model account for the strong variations of GC-content observed across bacterial species ? The gBGC model predicts that , all else being equal , the present-day GC-content of a genome should directly reflect its average recombination rate over long evolutionary time . To test this prediction , it is important to take into account two difficulties . First , recombination rates measured in extant populations reflect recent events ( more recent than the coalescent time , i . e . of the order of Ne generations ) , and hence may not correspond to the average recombination rate over times necessary for genomic GC-content to evolve significantly ( i . e . inter-species divergence times ) . Second , the precision in the estimate depends on the physical scale at which recombination is measured . To illustrate these points let us consider the human genome , where the impact of gBGC is well documented [6] . At the gene scale , the correlation between present-day recombination rate ( measured in a 10-kb window , centered on the middle of the gene , using HapMap genetic map [42] ) , and the gene GC-content ( at third codon position ) is significant but quite weak ( R2 = 0 . 035 , p<10−10 ) . However , at 1Mb scale the correlation is much stronger ( R2 = 0 . 15; [9] ) . Furthermore , when GC-content variations and recombination rates are measured over the same evolutionary time period , the correlation becomes very strong ( 1Mb scale: R2 = 0 . 64; [45] ) . To test whether the impact of gBGC in bacteria was comparable to what is observed in mammals , we first focused on Streptococcus pyogenes , one of the species for which the signature of gBGC is strong ( Fig . 1 , see also Sup . Mat . ) . We computed the population-scaled recombination rate ( rho ) for each gene of the core genome , using ClonalOrigin [41] . The correlation between rho and the GC-content at third codon position of each gene ( GC3 ) is higher than what is observed in humans ( R2 = 0 . 087; p<10−9 ) . This result is remarkable , given that recombination rates are measured here at the gene scale ( typically about 1kb ) . To go further , we binned the data set into 20 groups of genes according to their GC3 , and we computed the correlation between the average GC3 and the average rho of each bin . Our reasoning is that by computing average values , we should get estimates of rho that are more robust to measurement noise and to possible temporal variations in recombination rates . Using this approach , we observed a strong correlation between the GC3 and recombination ( R2 = 0 . 60; Fig . 3A ) . To investigate the amplitude of this relation in the other bacterial species studied here , we used PREC ( an index based on the proportion of genes in a bin that are detected as recombinant by PHI ) , which provides a good estimate of the average recombination rate in a bin , and is much easier to compute than ClonalOrigin’s rho . We observed a significant correlation for 11 of the 14 species , and these significant correlations were positive in all cases , with R2 values comprised between 0 . 24 and 0 . 68 ( on average R2 = 0 . 43; Fig . 3B ) . Thus , in many bacteria , the average recombination rate in a bin is a good predictor of its average GC-content . We performed an analogous analysis in human genes using jackknife sampling of the dataset to scale it to the size of bacterial datasets . The average correlation observed in humans is R2 = 0 . 55 ( Fig 3A ) . Hence , in many bacteria , the intensity of the relationships between GC-content and recombination is comparable to that observed in humans , where the impact of gBGC on base composition is known to be strong [45] . This is consistent with the hypothesis that on the long term , the gBGC process can have a major influence on the evolution of base composition in bacteria . If the base composition of a genome is at evolutionary equilibrium then , by definition , the number of A/T→G/C substitutions must be equal to the number of G/C→A/T substitutions . Hildebrand and colleagues [23] noted that in a large majority of bacterial genomes ( 94/149 ) , the number of G/C→A/T changes ( inferred from the comparison of closely related organisms ) exceeds the number of A/T→G/C changes . Given that genomic base composition strongly fluctuates over long evolutionary times ( as demonstrated by the wide distribution of GC-content across bacterial species ) , it is not surprising that many genomes are not at equilibrium . However , what is unexpected is that this non-stationarity predominantly leads to loosing GC-content: a priori , at the scale of the entire bacterial biodiversity , one would expect to observe as many GC-increasing genomes as GC-decreasing genomes . One possible explanation is that the observed excess G/C→A/T changes among closely related genomes corresponds to polymorphic mutations , which eventually do not reach fixation because either selection or gBGC favors GC-alleles over AT-alleles [23] . Hildebrand and colleagues observed an excess of G/C→A/T changes even in bacterial genomes that show no evidence of recent recombination population-wise ( i . e . Ner = 0 ) [23] . They therefore rejected the hypothesis that the fixation bias could be due to gBGC . However , this conclusion relies on one important assumption: that the Ner parameter measured in extant populations reflects the long-term average recombination rate . In fact it is expected that Ne ( and hence Ner ) should fluctuate over time , as populations go through periods of bottlenecks and expansion . Immediately after a bottleneck , Ner would be close to 0 , and hence genomes should accumulate G/C→A/T substitutions . However , on the long term , this can be compensated by an increase in GC-content when the effective population size becomes larger ( and hence B > 1 ) . Thus , the base composition of genomes may remain above the mutational equilibrium on the long term , even if many lineages go through periods during which Ner is null ( and hence B = 0 ) . Interestingly , the rare species for which the long-term recombination rate is effectively null ( typically endosymbiotic bacteria ) , generally have very AT-rich genomes [46] , as predicted by the gBGC hypothesis . Several lines of evidence suggest that the pattern of spontaneous mutations is biased towards AT nucleotides , in Eukaryotes as well as Prokaryotes [22 , 23 , 47] . Under such a bias , it is expected that the selective pressure to reduce mutation rate should generally favor a GC-biased DNA repair machinery . Recombination and repair are tightly linked processes that use many common pathways . In yeast , the analysis of conversion tracts in meiotic product indicates that the conversion bias is most probably due to the mismatch repair machinery ( MMR ) [48] . The MMR components involved in homologous recombination ( MutS , MutL ) are generally conserved between Bacteria and Eukaryotes . Hence , in Bacteria as well as Eukaryotes , gBGC could be the secondary effect of a selection for biased repair mechanisms . However , population genetics model show that when gBGC is strong , it drives the fixation of deleterious mutations [49] . Thus , one interesting hypothesis is that in highly recombining species , selection might favor unbiased repair mechanisms ( i . e . , values of b0 close to 0 in equation ( 1 ) ) , so as to limit the deleterious consequences of strong gBGC . H . pylori is notorious for being highly recombining [50 , 51] , as confirmed by our results . It is interesting that we have found no evidence for gBGC in this species , and recombining genes even seem to have slightly lower GC-content than non recombining genes . This trend is however relatively weak because there is no significant trend of variation of recombination rates among classes of homogeneous GC-content , as shown in Fig . 3B . One possible explanation for the absence of gBGC in H . pylori could hence be that b0 is null in this species , in compensation of a very high rate of recombination . Although the MMR pathway is a good candidate as a molecular source of gBGC in Bacteria , the association of gBGC with MutSL genes is not straightforward . These genes are absent from three of our genome datasets , C . jejuni , H . pylori and Bifidobacterium longum , resulting from ancestral losses in Delta-Proteobacteria and Actinobacteridae , respectively [52] . In H . pylori , we indeed find no evidence of gBGC while the genomes are recombining at high frequency [50 , 51] . In C . jejuni and B . longum , however , we observe patterns similar to the other bacterial datasets that are in support of the existence of gBGC , indicating that it does not depend on the presence of a typical MutSL complex . The existence of gBGC in Bacteria and Eukaryotes however suggests that it may have been present in the last universal common ancestor of all cellular life forms ( LUCA ) . Unfortunately , the only archaeal dataset matching our criteria was a group of Sulfolobus sp . genomes for which our analysis showed few evidence of recombination ( Table 1 ) , in agreement with the previously described isolation of endemic clades in this group [36] . We do not claim that gBGC is the unique determinant of base composition in bacterial genome: in fact there is evidence that mutation patterns vary significantly among species [22] , and these variations are expected to contribute to differences in genome base composition . However , the model we propose provides a simple explanation for several important results of comparative bacterial genomics . First , gBGC can explain why bacterial genomes can maintain a high GC-content , even though mutation is universally AT-biased [22 , 23] . Second , gBGC can explain some of the intragenomic heterogeneity in GC-content observed in bacterial genomes . Indeed , we observe that genes with evidence of recombination display on average substantially higher GC-content than other genes . This observation also suggests that the probability of recombination is variable among genes in the genome , as proposed under some speciation models [53] . Furthermore , given that recently acquired genes tend to be AT-rich , gBGC would contribute to their progressive enrichment in GC-content [54 , 55] . The variations of GC-content in Bacteria have long remained unexplained . The results presented here highlight a strong relationship between the GC-content of genes and their history of recombination . This result , and the observation that bacterial genomes are generally above the GC-content predicted from their mutational bias towards AT , are fully consistent with the existence of gBGC . To explain our results under a selective model , one would have to hypothesize: i ) that all bacterial species are under the same selection for higher GC throughout their genome , ii ) that this selective pressure affects all positions of a genome , independently of gene function and expression , and iii ) that the efficiency of selection varies with recombination ( Hill-Robertson interference ) . Incidentally , if the correlations between GC-content and recombination ( Fig . 3B ) were due to Hill-Robertson interference , this would imply that all regions in a genome ( except possibly the most GC-rich ) are mal-adapted . We favor the gBGC model because it is much more parsimonious , and it relies on mechanisms that have already been uncovered in Eukaryotes . Ultimately , it will be possible to test experimentally the existence of gBGC by analyzing recombination products in bacteria . Our discovery is important because gBGC has been shown to interfere with the efficiency of selection in Eukaryotes , and to lead to false positives in the search for regions under positive selection in a genome . The prevalence—if not universality—of this phenomenon underlines the importance of incorporating gBGC in the set of evolutionary forces to be considered when searching for signature of adaption in genomes .
We used the HOGENOM database [27] to select sets of genome sequences comprising at least six closely related strains or species . The selection of closely related genomes was based on a genomic distance derived from the HOGENOM6 database . For each pair of genomes present in the database , we computed the best identity score , as obtained using BLAST , within each family of homologous proteins as defined in HOGENOM , and averaged these scores to obtain a global similarity score s . We then took 1-s as a distance and selected groups of genomes with at least 6 members and a distance lower than 0 . 15 . This criterion left 21 groups of species representing a variety of bacterial and archeal species ( Table 1 ) . For each gene family , CDSs were extracted using ACNUC Python API [56] and re-aligned with MUSCLE [57] using default parameters . GC% and codon frequencies were computed using custom Python scripts . Detection of recombination based on multiple alignments is expected to be sensitive to both the number of sequences aligned , and the total length of the alignment . For this reason , we used only the universal unicopy genes of each species , and selected the 900 central positions of each nucleotide alignment ( remaining positions as well as genes shorter than this threshold were ignored in our analysis ) . We used the software Phipack implementing the PHI test [28] ( parameters: window size of 100bp , p-value computed from 1 , 000 permutations ) to test if a gene family alignment contained evidence for recombination . Only families for which the site permutation test could be performed were considered , i . e . where the phylogenetic signal was sufficient to accept or reject the hypothesis of recombination . An alignment was determined to be “recombinant” if the p-value of the permutation test was lower than 0 . 05 . To confirm the results obtained with PHI , we also used three other tests of recombination: NSS [58] , MaxChi2 [59] and Geneconv [60] . NSS and MaxChi2 statistics were computed together with PHI ( as implemented in Phipack package ) using the same permutations to compute p-values . A consensus of PHI , NSS and MaxChi2 was done , classifying as recombinant or non-recombinant those gene families consistently detected by all those three methods as recombinant or non-recombinant , respectively; families with disagreeing results were discarded for the consensus analysis . Geneconv was run with the parameters “-GScale = 1 -Numsims = 10000 -Maxsimglobalpval = 0 . 05” , and a gene family was considered recombinant when at least one significant global recombinant fragment was reported; other families were considered non-recombinant . As Geneconv differs from the other methods on the nature of the reported evidence , it was not considered when defining a consensus classification . Individual tests and their consensus yielded quite different classifications of gene families , but all led to qualitatively very similar results ( see S1 Fig . , S3 Fig . and S4 Fig . ) . To test for the presence of gBGC , we performed Student’s t-tests comparing the mean GC-content ( at all or at each separate codon position ) of recombinant versus non-recombinant core gene family sets . All alignments as well as the results of the various tests for recombination are available at ftp://pbil . univ-lyon1 . fr/pub/datasets/DAUBIN/Lassalle_PloS_Genet_2015 We used ClonalOrigin [41] to estimate the recombination rate on full-length core gene family alignments of 900bp or more . As ClonalOrigin inferences are highly demanding in computation time and power , we only performed this analysis on the moderately sized dataset of S . pyogenes genomes ( Spyo ) . The program was used with default parameter weights , running the MCMC with 1 , 000 , 000 burn-in generations and 500 , 000 generations sampled every 1 , 000 ( command line was: ‘warg -a 1 , 1 , 0 . 1 , 1 , 1 , 1 , 1 , 1 , 0 , 0 , 0 -x 1000000 -y 500000 -z 1000’ ) . Families for which this task could not be finished in less than 1 month computation were discarded , yielding a total dataset of 478 gene families ( out of 496 core ≥900bp-long ones ) . A number of genes also exhibited very high variance in their estimates of rho and their exclusion yielded better correlations with GC3 . Intergenic regions are evolving fast , leading to inaccurate alignments that are not amenable to robust detection of recombination as performed on coding sequences . We thus chose to test the relationship of intergenic GC% with recombination in a gene-centered way: for each core protein-coding gene tested with PHI , we considered non-coding regions up to 400bp on both sides of a CDS and averaged their GC% , provided they were of a size larger than 50bp each ( to avoid stochastic errors due to too small number of observed nucleotides ) . A measure of intergenic GC% per core gene family was obtained as the mean over all genomes of the previous values , and was associated to the recombinant/non-recombinant status of the gene family for further testing . Optimal codons for each amino acid were computed by comparing synonymous codon frequencies within CDSs encoding ribosomal proteins ( as defined from HOGENOM family annotations ) versus all other CDSs ( RP method ) . Under the hypothesis of selection for highly expressed genes to be adapted to the tRNA pool , codons statistically enriched in ribosomal proteins ( Chi-squared test based on a 2×2 contigency table of counts of occurrence at a focal codon against those of its synonyms in ribosomal vs . other protein-coding genes , with p-value < 0 . 001 ) were considered as “optimal”; others were classified as “non-optimal” . ( S2 Table ) . We then computed the absolute frequency of optimal ( Fop ) and non-optimal codons ( Fnop ) over all coding sequences . Fop and Fnop were calculated separately for codons pooled by composition at the third position , i . e . ending in A/U or G/C . As there is a debate on whether this method is appropriate to define optimal codons [38–40] , we also used an alternative definition and took optimal codons datasets from a previous exhaustive survey of Hershberg and Petrov [38] ( HCB method ) . A set of optimal codons was selected when determined in Hershberg and Petrov [38] study for several strains with a strong consensus , i . e . when >60% documented strains agreed on the preferred codon and remaining strains preferred a codon with the same composition ( A/U or G/C ) at the third position ( found in all datasets but A . baumanii ) ( S3 Table ) .
|
Classical population genetics models indicate that the efficiency of selection , and hence adaptation , depends on a number of non-selective factors , such as the size of a population or the intensity of recombination . In the last 10 years , evidence has accumulated that another mechanism called GC-Biased Gene Conversion ( gBGC ) can interfere with selection and even mimic its effects . This phenomenon , which arises from a particularity of the recombination machinery , was first thought to be restricted to sexual eukaryotic organisms . Here , we show that this mechanism probably exists in Bacteria and has a strong impact on their genome evolution . This discovery not only explains many previously unconnected features of bacterial genome evolution , but also highlights the importance of non-adaptive evolutionary processes in Bacteria .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
GC-Content Evolution in Bacterial Genomes: The Biased Gene Conversion Hypothesis Expands
|
We evaluated the profile of patients referred to the Fiocruz Outpatient Clinic , a reference center for the diagnosis and treatment of leprosy in Rio de Janeiro , RJ , and analyzed the origins and outcomes of these referrals . This is an observational retrospective study based on information collected from the Leprosy Laboratory database at Fiocruz , Rio de Janeiro , RJ , Brazil . A total of 1 , 845 suspected leprosy cases examined at the reference center between 2010 and 2014 were included . The originating health service referrals and diagnostic outcomes were analyzed as well as the clinical and epidemiological data of patients diagnosed with leprosy . Our data show that the profile of the patients treated at the Clinic has changed in recent years . There was an increase in both the proportion of patients with other skin diseases and those who had visited only one health service prior to our Clinic . Among the total 1 , 845 cases analyzed , the outcomes of 1 , 380 were linked to other diseases and , in 74% of these cases , a biopsy was not necessary to reach a diagnostic conclusion . A decrease in new leprosy case detection among our patients was also observed . Yet , among the leprosy patients , 40% had some degree of disability at diagnosis . The results of the present study demonstrated the importance of referral centers in support of basic health services within the decentralization strategy . But , the success of the program depends on the advent of new developmental tools to augment diagnostic accuracy for leprosy . However , it should be emphasized that for new diagnostic methods to be developed , a greater commitment on the part of the health care system regarding research is urgently needed .
In 1991 , the World Health Organization ( WHO ) adopted the goal of eliminating leprosy as a public health issue worldwide by the year 2000 . Elimination was defined as achieving a prevalence rate of lower than 1/10 , 000 inhabitants [1] . The widespread implementation of the multidrug therapy ( MDT ) program has been a success , resulting in a substantial reduction in global prevalence . Nonetheless , new case detection rates have not decreased as rapidly as expected in certain countries , especially in Brazil , India , and Indonesia , which have remained endemic . In 2014 , 213 , 899 new leprosy cases were detected around the world while , in the Americas , new cases numbered 33 , 789 , 94% of which were in Brazil [2] . Concurrently in the same year , the detection rate in the State of Rio de Janeiro was 7 . 36/100 , 000 inhabitants [3] . Since the advent of the reform within the Brazilian Health System in 1989 and , with it , the initial implementation of the family health strategy , the prevention , diagnosis and treatment of diseases , including leprosy , were moved to the primary care level [4] . Within this system containing an extensive network of health facilities and recognized that the rate of leprosy new case detection has basically not wavered over the last 20 years , it was decided to decentralize leprosy control strategies throughout the country . There was a shift from a vertical model associated with general dermatological services to one in which the diagnosis and treatment of the disease would be integrated into the primary care level facilities . By 2009 , the State of Rio de Janeiro had implemented the new strategy in 40% of all its municipal primary health care services [5] . At present , the City of Rio de Janeiro has 147 basic health care that follow the family health model and 59 that are specialized in the traditional sense . In this context , leprosy reference centers continued to be primarily responsible for diagnosing complex cases , managing difficult reactional episodes , treating the side effects of MDT , evaluating relapse cases and developing research projects [6] . However , once patient demand for leprosy reference centers was affected , in any way , by integrating leprosy diagnosis into basic health care services , it was necessary to determine the pattern of all these referrals in the light of the new decentralization policy . Previous studies have focused on the consequences of decentralization on the indicators used to evaluate leprosy control , i . e . , the proportion of MB patients , new cases under the age of 15 , patients diagnosed with physical disabilities as well as on the clinical and epidemiological profiles of leprosy patients [7] [8] . In this study , we decided to investigate the origins and outcomes of suspected leprosy patient referrals to the Fiocruz Outpatient Clinic , a reference center for the diagnosis and treatment of leprosy in Brazil , after the implementation of the decentralization policy .
The present study is an observational retrospective study of the information gathered from the Leprosy Outpatient Clinic database at Oswaldo Cruz Foundation ( Fiocruz ) , a reference center for the diagnosis and treatment of leprosy located in the City of Rio de Janeiro , RJ , Brazil . Patients referred to the Fiocruz Clinic for diagnostic confirmation from January 2010 thru December 2014 were included from the moment we began registering the origin of these referrals in our database . All suspected cases of leprosy who arrived at the Clinic from both public and private health services were included . Likewise , all patients who arrived spontaneously along with the household contacts of new leprosy cases that had some skin or neurological suspicious lesions became participants . Conversely , those who attended at the Clinic for therapeutic management and not for diagnosis ( e . g . control of reactions , suspicion of recurrence ) and those who abandoned the study prior to receiving a diagnostic conclusion were excluded . The sample under study consisted of 1 , 845 cases . The Fiocruz Outpatient Clinic serves individuals from the metropolitan area of the City of Rio de Janeiro as well as other cities in the State . Patients are referred to the Clinic by both public and private health care services , arrive spontaneously , or are household contacts of a leprosy case with suspected skin or neurological lesions . For diagnostic purposes , patients are routinely undergo dermato-neurological evaluation , skin smears , skin biopsies for histopathological analysis [9] and , if necessary , polymerase chain reaction is performed [10] . Those diagnosed with leprosy are treated and followed up at the Fiocruz Clinic or referred to the original service for treatment . Those diagnosed with other dermatoses or neuropathies are sent to referral services for these specific diseases . Socio-economic , clinical , and epidemiological information , laboratory parameters , and case outcomes are recorded onto a database . For the present study , the use of these data was approved by the Ethics in Research Committee of the Oswaldo Cruz Foundation number 976 . 330–10/03/2015 . The variables analyzed in this study were: i ) age , sex , place of residence , originating health service ( public , private or spontaneous demand ) , and number of health services the patient consulted before coming to our Clinic ( 1 , 2 , 3 or more ) ; ii ) clinical diagnosis without a biopsy: leprosy or other disease ( OD ) ; iii ) histopathological diagnosis: leprosy or other disease; iv ) the presence or absence of disability at leprosy diagnosis; and v ) patient destination after outcome: continuation at the Fiocruz Clinic , return to the originating service , or referral to another health care service . These clinical and epidemiological aspects of patient referrals during 2010–2014 are presented in the tables; and bivariate analysis were conducted for categorical variables using the chi square test . Also , in the present study , data during 2005–2014 were analyzed and compared with the 2010–2014 . We retrieved data from the official leprosy case data reported by the City and State of Rio de Janeiro obtained at the SINAN ( the National Disease Information System ) database . We used the following variables: the annual number of new leprosy cases at the Fiocruz Outpatient Clinic and in the City and State of Rio de Janeiro; the annual proportion of newly-diagnosed leprosy cases compared to other diseases at Fiocruz Clinic; the annual percentage of the number of health services visited prior to Fiocruz referrals . Means were calculated to compare the 2005–2009 and 2010–2014 periods . The statistical analysis was performed using SPSS version 22 software .
The leprosy data regarding both the City and State of Rio de Janeiro reveal an accentuated decline in the number of new leprosy cases between 2005 and 2014 , the same being true for the Fiocruz Clinic , as seen in Fig 1 . The reduction in the mean of cases between 2010–2014 as compared to 2005–2009 was 51% and 34% in the City and State respectively , while at the Fiocruz Clinic , the decline was lower , 27% ( Fig 1 ) . Analysis of total patient demand at the Fiocruz Clinic ( new leprosy cases plus those with other skin diseases ) showed that the mean proportion of leprosy patients seen at the Fiocruz Clinic between 2005 and 2009 was 28% while the mean proportion of patients with other dermatoses in the same period was 47% . During the 2010–2014 , there was an increase of 16% in the proportion of patients with other skin diseases ( mean proportion = 55% ) compared to the previous period and a proportional decrease of approximately 33% among patients with leprosy ( mean proportion = 18% ) ( p <0 . 001 ) ( Fig 2 ) . Observing that the trend in our Clinic had shifted during 2010–2014 in comparison to 2005–2009 , it was decided to evaluate how many health care services had been consulted by these patients before being referred to the Fiocruz Clinic , whose results are showed in Fig 3 . The mean proportion of patients who visit only one clinic prior their Fiocruz referral was 20% during 2005–2009 while in the 2010–2014 time period this mean was double ( 40% ) ( Fig 3 ) . Analyzes of the clinical and epidemiological aspects of patients seen after 2010 showed that , of the 1 , 845 cases evaluated , 1 , 380 ( 75% ) had other dermatoses or neuropathies and , in 74% of these cases , biopsies were not taken because a leprosy diagnosis could be excluded after a neuro-dermatological evaluation . Only 25% ( 465 ) of the cases evaluated at our Clinic for suspected leprosy had a confirmed diagnosis of leprosy . Fifty-one per cent ( 375 ) of all the biopsied were diagnosed with leprosy and 49% ( 359 ) , with other dermatoses ( Table 1 ) . Analyzing the association between original referral service and outcome , there was no difference in the positive leprosy outcome between public and private health services referral source . ( Table 2 ) . Regarding suspected cases of leprosy who spontaneously arrived at Fiocruz Outpatient Clinic , a significant association with the diagnosis of other diseases was observed . Among these , only 30 ( 11 . 3% ) were found to have leprosy ( Table 2 ) . Table 3 presents the clinical and epidemiological characteristics of cases of confirmed leprosy diagnoses . The majority of patients ( 70 . 1% ) were between 15 and 59 years of age and 21 . 6% were over 60 . Male patients predominated representing 59% in our cohort . Considering the operational classification , there was a slight difference between the proportion of paucibacillary ( PB ) versus multibacillary ( MB ) patients ( 49% and 51% , respectively ) in this study . The disability grade ( DG ) of two-hundred-and-sixty-two confirmed leprosy patients were registered in the database: 14% had grade 2 of disability and 40% had some degree of disability at diagnosis . There was no statistical significance between the presence of physical disabilities and whether the referrals came from public or private health care facilities . Sixty-nine percent of the diagnosed leprosy patients were treated and followed up at Fiocruz Outpatient Clinic ( 320/465 ) ( Table 3 ) .
The results of the present study provide strong evidence that the profile of the patients diagnosed and treated at the Fiocruz Outpatient Clinic have changed in the past 5 years possibly as a result of the public policy shift towards decentralization . The proportion of patients with other skin diseases and those who visited only one health service before our Clinic increased . It is noteworthy that , in 74% of the cases with other diseases , a biopsy was not necessary for diagnosis indicating that general practitioners in the primary care facilities could not distinguish easily diagnosed skin diseases ( other than leprosy ) . After dermatological examination by a specialist at our clinic , the diagnoses of other dermatoses were clearly defined , and were more often eczematous diseases , psoriasis , superficial mycoses or dyschromia ( S1 Table ) . Nevertheless , a decrease in new leprosy case was not only detected in our Clinic but was also observed in the City and State of Rio de Janeiro . Smith et al . have suggested the possibility that the global decline in case detection in conjunction with the rise in disabilities may be linked to the move from vertical leprosy control activities to integrated approaches [11] . On the other hand , some authors have indicated that the impact of decentralizing policies have pointed to such major gains as reduced prevalence rates , an increase in early detection , and maintenance of the quality of care [12] [13] . Others studies have emphasized the importance of the ability of trained dermatologists to accurately diagnose the disease , suggesting that the greater the success in reducing the disease burden , more important is the role of the specialist with knowledge of the disease and its differential diagnoses [14] [15] . Regarding the epidemiological and clinical aspects of the leprosy cases diagnosed at Fiocruz Outpatient Clinic from 2010 to 2014 there were no important differences concerning age and gender between our results and the proportions recorded in Brazil in 2014 . Considering the operational classification , there was a slight difference between the proportion of PB and MB patients ( 49% and 51% , respectively ) , while last year , new cases detected in the country was 66% MB [2] . This result may be biased , however , since the differential diagnosis of paucibacillary ( PB ) leprosy with other dermatoses is often more difficult to achieve than the diagnosis of multibacillary ( MB ) . Among the leprosy patients , considered in the present study , 40% had some degree of disability at diagnosis . In Brazil , within the leprosy new cases diagnosed in 2014 , the percentage of grade 2 disabilities was 6 . 56% [2] . Both the City and State of Rio de Janeiro have registered rising percentages of this indicator . In fact , the latest five-year averages were 10% and 10 . 5% , respectively [3] . The Ministry of Health deems a grade 2 disability ≥10% high [16] . It is possible that the fact that this study was carried out in a reference center specialized in the treatment of leprosy had an impact on theirs result due to a selection and measurement biases . However a previous analysis of leprosy patients also treated at the Fiocruz reference center between 2003 and 2007 showed that 12 . 2% had grade 2 disability and that 32 . 5% had some degree of disability at diagnosis at that time [17] . These data strongly indicate that the referrals to the Outpatient Clinic of this reference center were , in actuality , delayed , which is particularly surprising in a State with a large number of units and health professionals . On the other hand , this delay may more accurately reflect the difficulties involved in diagnosing leprosy in the primary care health units in the absence of specialized health professionals or laboratory tests to aid in diagnosing the disease , especially since the initial presentation of leprosy may be a slight injury of nerve or a discreet and asymptomatic skin lesion , which would further complicate early diagnosis . One of the advantages attributed to the integration of leprosy diagnosis and treatment into basic health care units is the increased access of the general population to these services [18] [19] [20] . Theoretically , the integration strategy should contribute to more effective disease control as it would increase the chances of early diagnosis , avoid the occurrence of sequels , break the transmission chain , and increase patient adherence to treatment . But there is a consensus that for integration to be truly successful , health professionals must receive constant theoretical and practical training [19] [20] . A major obstacle is that in many Brazilian states , the permanence of doctors and nurses in primary care health units has proved to be difficult [21] [22] . In the State of Rio de Janeiro , for example , the Public Health Department analyzed the impact of the integration policy for 4 years after its initial implementation into the system . It found that the high professional turnover rate in the primary care health units along with the hardships encountered in obtaining sufficient financial resources to adequately train new professionals are issues that adversely impact the effectiveness of leprosy control measures in the State [18] . In this context , the results of the present study demonstrated the importance of our referral center in support of the basic health care services by accepting cases from all over the State and performing differential diagnosis of skin diseases and neuropathies . Studies performed in other countries have demonstrated that specialized services are necessary and continue to provide significant support within an integrated health care system approach towards the diagnosis and management of leprosy [23] [24] . In addition to medical assistance , reference centers remain committed to their role in developing research that contributes to leprosy control , specifically searching for new tools to more rapidly identify early signs of the disease . For example , many recent reports have shown that PCR-based assays are excellent adjuncts in clinical and histopathological analyses toward the definitive identification of M . leprae [25] [26] . Other studies aimed at the identification of biomarkers profiles associated with the early onset of type 1 leprosy reactions [27] in addition to antigens that could be used to monitor treatment efficacy in leprosy patients have shown great promise [28] . At specialized leprosy referral centers in Bangladesh and Brazil , Walker et al performed a severity scale for leprosy Type 1 reactions to help diagnose reactional episodes and improve the management of this disabling complication of leprosy [29] . Moreover , the Fiocruz Leprosy Reference Center has developed studies to identify the major risk factors associated with the incidence of leprosy among household contacts in order to support monitoring programs with the use of screening procedures able to spot high-risk individuals thereby widening the opportunities for early diagnosis and treatment . For that purpose , serological test using anti-PGL1 has been performed among leprosy household contacts [30] . In recent years , studies carried out at the Fiocruz Leprosy Laboratory have demonstrated the great value of qPCR in the clinical management of suspected cases of paucibacillary leprosy [10] as well as pure neural leprosy [31] . But , to develop new diagnostic methods , that could be used in a variety of field conditions , to augment diagnostic accuracy , a greater commitment on the part of the health care system regarding research is urgently needed .
|
Leprosy , a neglected disease , remains endemic in some developing countries despite the existence of a successful program to treat and cure patients . While has been a drastic decrease in the number of patients , but we still have a stable number of new cases that is still very high in countries like India and Brazil in which more than 30 . 000 new cases were observed in 2014 . Over the past ten years , Brazil has changed the strategies regarding of public health so that leprosy diagnostic , treatment and surveillance functions would predominantly be performed in primary care health units . The decentralization of leprosy diagnosis and treatment was expected to impact early cases detection and contribute to decrease in the number of cases with nerve damage . We analyzed and compared the demand of patient referrals to the Fiocruz Outpatient Clinic , a reference center for the diagnosis and treatment of leprosy in Rio de Janeiro , RJ , Brazil , prior and subsequent to the implementation of the decentralization strategy . Our results indicated that the profile of patients treated at the Fiocruz Clinic changed after the diagnosis and treatment of leprosy was integrated into the primary health services . There was an increase in the proportion of patients with other skin diseases . At the same time , 40% of the patients with leprosy had a higher disability grade at diagnosis , indicating late diagnosis . The initial presentation of leprosy may be a discrete skin or neural lesion , representing a challenge even for trained dermatologists . These results are probably the consequence of difficulties encountered in diagnosing leprosy in the primary health units without specialized health professionals or adequate laboratory tests . Although decentralization strategies have several advantages integrating the diagnosis of leprosy into basic health care units , the support of referral centers in diagnosing complex cases , managing difficult reactional episodes , and treating of side effects is central to the control of the epidemic .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"dermatology",
"medicine",
"and",
"health",
"sciences",
"outpatient",
"clinics",
"disabilities",
"biopsy",
"tropical",
"diseases",
"geographical",
"locations",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"health",
"care",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"skin",
"diseases",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"primary",
"care",
"south",
"america",
"brazil",
"people",
"and",
"places",
"diagnostic",
"medicine",
"health",
"care",
"facilities",
"leprosy"
] |
2016
|
Impact of a Reference Center on Leprosy Control under a Decentralized Public Health Care Policy in Brazil
|
Many interspecies hybrids have been discovered in yeasts , but most of these hybrids are asexual and can replicate only mitotically . Whole-genome duplication has been proposed as a mechanism by which interspecies hybrids can regain fertility , restoring their ability to perform meiosis and sporulate . Here , we show that this process occurred naturally during the evolution of Zygosaccharomyces parabailii , an interspecies hybrid that was formed by mating between 2 parents that differed by 7% in genome sequence and by many interchromosomal rearrangements . Surprisingly , Z . parabailii has a full sexual cycle and is genetically haploid . It goes through mating-type switching and autodiploidization , followed by immediate sporulation . We identified the key evolutionary event that enabled Z . parabailii to regain fertility , which was breakage of 1 of the 2 homeologous copies of the mating-type ( MAT ) locus in the hybrid , resulting in a chromosomal rearrangement and irreparable damage to 1 MAT locus . This rearrangement was caused by HO endonuclease , which normally functions in mating-type switching . With 1 copy of MAT inactivated , the interspecies hybrid now behaves as a haploid . Our results provide the first demonstration that MAT locus damage is a naturally occurring evolutionary mechanism for whole-genome duplication and restoration of fertility to interspecies hybrids . The events that occurred in Z . parabailii strongly resemble those postulated to have caused ancient whole-genome duplication in an ancestor of Saccharomyces cerevisiae .
A whole-genome duplication ( WGD ) occurred more than 100 million years ago in the common ancestor of 6 yeast genera in the ascomycete family Saccharomycetaceae , including Saccharomyces [1 , 2] . Recent phylogenomic analysis has shown that the WGD was an allopolyploidization—that is , a hybridization between 2 different parental lineages [3] . One of these parental lineages was most closely related to a clade containing Zygosaccharomyces and Torulaspora ( ZT ) , whereas the other was closer to a clade containing Kluyveromyces , Lachancea , and Eremothecium ( KLE ) . The ZT and KLE clades are the 2 major groups of non-WGD species in family Saccharomycetaceae . The WGD had a profound effect on the genome , proteome , physiology , and cell biology of the yeasts that are descended from it , but the genomes of these yeasts have changed substantially in the time since the WGD occurred , with extensive chromosomal rearrangement , deletion of duplicate gene copies , and sequence divergence between ohnologs ( pairs of paralogous genes produced by the WGD ) . These changes have made it difficult to ascertain the molecular details of how the WGD occurred . Ancient hybridizations are rare in fungi ( or at least difficult to detect [4] ) , but numerous relatively recent hybridizations have been identified using genomics , particularly in the ascomycete genera Saccharomyces [5 , 6] , Zygosaccharomyces [7–9] , Candida [10–12] , and Millerozyma [13] . Marcet-Houben and Gabaldón [3] proposed 2 alternative hypotheses for the mechanism of interspecies hybridization that led to the ancient WGD in the Saccharomyces lineage . Hypothesis A was hybridization between diploid cells from the 2 parental species , perhaps by cell fusion . Hypothesis B was mating between haploid cells from the 2 parental species to produce an interspecies hybrid zygote , followed by genome doubling . Under both hypotheses , the product is a cell with 2 identical copies of each parental chromosome . These identical copies should be able to pair during meiosis , leading to viable spores . While there are no known examples of natural yeast hybrid species formed by diploid–diploid fusion ( hypothesis A ) , 3 examples have been discovered in which hybrid species were apparently formed simply by mating between haploids of opposite mating types from different species ( hypothesis B ) . These are Candida metapsilosis [11] , C . orthopsilosis [10 , 12] , and Zygosaccharomyces strain ATCC42981 [8 , 14] . These interspecies hybridizations occurred by mating between parents with 4%–15% nucleotide sequence divergence between their genomes . However , none of these 3 hybrids can sporulate , which could be either because the homeologous chromosomes from the 2 parents are too divergent in sequence to pair up during meiosis or because pairing occurs but evolutionary rearrangements ( such as translocations ) between the parental karyotypes result in DNA duplications or deficiencies after meiosis [15–18] . None of these 3 hybrids has undergone the genome-doubling step envisaged in hypothesis B . Several groups [3 , 18–20] have proposed that genome doubling could occur quite simply by means of damage to 1 copy of the MAT locus in the interspecies hybrid , which could cause the hybrid cell to behave as a haploid , switch mating type , and hence autodiploidize . This proposal mimics laboratory experiments carried out by Greig et al . [21] in which hybrids between different species of Saccharomyces were constructed by mating . The hybrids were unable to segregate chromosomes properly and were sterile , but when 1 allele of the MAT locus was deleted , they spontaneously autodiploidized by mating-type switching and were then able to complete meiosis and produce spores with high viability . Each spore contained a full set of chromosomes from both parental species [21] . While genome doubling via MAT locus damage is an attractive hypothesis consistent with hypothesis B above [3] , no examples of it occurring in nature have been described . We show here that Z . parabailii has gone through this process . There are 12 formally described species in the genus Zygosaccharomyces [22] . The most studied of these is Z . rouxii , originally found in soy sauce and miso paste [23 , 24] . Others include Z . mellis , frequently found in honey [25] , and Z . sapae from balsamic vinegar [26 , 27] . Species in the Z . bailii sensu lato clade ( Z . bailii , Z . parabailii , and Z . pseudobailii; [28] ) are of economic importance because they are exceptionally resistant to osmotic stress and low pH . Their resistance to the weak organic acids commonly used as food preservatives makes them the most frequent spoilage agent of packaged foods with high sugar content , such as fruit juices and jams , or with low pH , such as mayonnaise [29–33] . These same characteristics make Zygosaccharomyces relevant to biotechnology since high stress tolerance and rapid growth are often desirable traits in microorganisms to be used as cell factories . The strain we analyze here , Z . parabailii ATCC60483 , has previously been used for production of vitamin C [34] , lactic acid [35] , and heterologous proteins [36] . Despite the diversity of the genus , genome sequences have been published for only 2 nonhybrid species of Zygosaccharomyces: the type strains of Z . rouxii ( CBS732T; [37] ) and Z . bailii ( CLIB213T; [38] ) . The genus also includes many interspecies hybrids with approximately twice the DNA content of pure species ( 20 Mb instead of 10 Mb; [7 , 8 , 14 , 39] ) . Mira et al . [39] sequenced the genome of Zygosaccharomyces strain ISA1307 and found that it is a hybrid between Z . bailii and an unidentified Zygosaccharomyces species . In 2013 , Suh et al . [28] proposed that some strains that were historically classified as Z . bailii should be reclassified as 2 new species , Z . parabailii and Z . pseudobailii , based on phylogenetic analysis of a small number of genes . The sequences of the RPB1 and RPB2 genes that they obtained from Z . parabailii and Z . pseudobailii contained multiple ambiguous bases , consistent with a hybrid nature [39] . In the current study , we sequenced the genome of a second hybrid strain , ATCC60483 . We show that ATCC60483 and ISA1307 are both Z . parabailii and are both descended from the same interspecies hybridization event . By sequencing ATCC60483 using Pacific Biosciences ( PacBio ) technology , we obtained near-complete sequences of every Z . parabailii chromosome , which enabled us to study aspects of chromosome evolution in this species that were not evident from the Illumina assembly of ISA1307 [39] .
We first tried to sequence the Z . parabailii genome using Illumina technology , but even with high coverage , we were unable to obtain long contigs . The data indicated that the genome was a hybrid , so instead we switched to PacBio technology , which generates long sequence reads ( 6 kb on average in our data ) . Our initial assembly had 22 nuclear scaffolds , which we refined into 16 complete chromosome sequences with a cumulative size of 20 . 8 Mb by manually identifying overlaps between the ends of scaffolds and by tracking centromere and telomere locations . We annotated genes using the Yeast Genome Annotation Pipeline ( YGAP ) , assisted by RNA sequencing ( RNA-Seq ) data to identify introns . The nuclear genome has 10 , 087 protein-coding genes , almost twice as many as Z . bailii CLIB213T ( Table 1 ) . Most of the chromosome sequences extend into telomeric repeats at the ends . The consensus sequence of the telomeres is tgtgggtgggg , which matches exactly the sequence of the template region of the 2 homeologous TLC1 genes for the RNA component of telomerase that are present in the genome . Chromosome sequences that do not extend into telomeres instead terminate at gene families that are amplified in subtelomeric regions or contain genes that are at chromosome ends in the inferred Ancestral ( pre-WGD ) gene order for yeasts [41] indicating that they are almost full length , except for 3 chromosome ends that appear to have undergone break-induced replication ( BIR ) and homogenization with other chromosome ends . We identified 1 scaffold as the mitochondrial genome , which maps as a 30-kb circle containing orthologs of all S . cerevisiae mitochondrial genes . We also found a plasmid in the 2-micron family ( 5 , 427 bp ) , with 99% sequence identity to pSB2 , which was first isolated [42] from the type strain of Z . parabailii ( NBRC1047/ATCC56075 ) . Visualization of the genome using a Circos plot [43] shows that most of the genome is duplicated , indicating a polyploid origin ( Fig 1 ) . However , although most genes have a homeolog , the chromosomes do not form simple collinear pairs . Instead , sections of each chromosome are collinear with sections of other chromosomes . Comparison to Z . bailii CLIB213T shows that for each region of the Z . bailii genome , there are 2 corresponding regions of the Z . parabailii genome: 1 almost identical in sequence and 1 with approximately 93% sequence identity , which demonstrates a hybrid ( allopolyploid ) origin of Z . parabailii and suggests that Z . bailii was one of its parents . To analyze this relationship in detail , we estimated the parental origin of every Z . parabailii ATCC60483 gene based on the number of synonymous substitutions per synonymous site ( KS ) when compared to its closest Z . bailii homolog ( Fig 2A ) . This analysis revealed a bimodal distribution of KS values in which 47 . 1% of the ATCC60483 genes are almost identical to CLIB213T genes ( KS ≤ 0 . 05 ) and a further 42 . 5% are more divergent ( 0 . 05 < KS ≤ 0 . 25 ) . From this relationship , we infer that Z . parabailii ATCC60483 is an interspecies hybrid formed by a fusion of 2 parental cells , which we refer to as Parent A ( purple ) and Parent B ( green ) . Parent A was a cell with a genome essentially identical to Z . bailii CLIB213T . Parent B was a cell of an unidentified Zygosaccharomyces species with approximately 93% overall genome sequence identity to Z . bailii , corresponding to a synonymous site divergence peak of KS = 0 . 16 ( Fig 2A ) . We refer to the 2 sets of DNA in Z . parabailii that were derived from Parents A and B as the A-subgenome and the B-subgenome , respectively . We refer to the A- and B-copies of a gene as homeologs , and we use a suffix ( “_A” or “_B” ) in gene names to indicate which subgenome they come from . The genome contains ribosomal DNA ( rDNA ) loci inherited from each of its parents . Our assembly includes 2 complete rDNA units with 26S , 5 . 8S , 18S , and 5S genes . Phylogenetic analysis of their internal transcribed spacer ( ITS ) sequences shows that the rDNA on chromosome 11 is derived from Z . bailii ( Parent A ) , whereas the rDNA on chromosome 4 is derived from Parent B and contains an ITS variant seen only in other Z . parabailii strains ( S1 Fig ) . A third rDNA locus in our assembly ( at 1 telomere of chromosome 15 ) is incomplete and does not extend into the ITS region . The rDNA unit on chromosome 4 is also telomeric , whereas the unit on chromosome 11 is located at an internal site 165 kb from the right end . None of the genes in the interval between this rDNA and the right telomere of chromosome 11 have orthologs in Z . bailii CLIB213T . Z . parabailii has 16 chromosomes . We identified its 16 centromeres bioinformatically , which correspond to 2 copies ( A and B ) of each of the 8 centromeres in the Ancestral pre-WGD yeast genome ( Table 2 ) [41 , 44] . In contrast , Z . rouxii has only 7 chromosomes because of a telomere-to-telomere fusion between 2 chromosomes followed by loss of a centromere [44] . The missing centromere in Z . rouxii is Ancestral centromere Anc_CEN2 , which maps to Z . parabailii centromeres CEN4 and CEN11 , located between the genes MET14 and VPS1 . The Z . rouxii centromere must have been lost after it diverged from the Z . bailii/Z . parabailii lineage . Alignment of the Z . rouxii MET14-VPS1 intergenic region with the Z . parabailii CEN4 and CEN11 regions shows that the CDE III motif of the point centromere has been deleted in Z . rouxii ( S2 Fig ) . Z . parabailii inherited the mitochondrial genome of its Z . bailii parent . A complete mitochondrial genome sequence for Z . bailii is not available , but we identified 55 small mitochondrial DNA ( mtDNA ) contigs in the CLIB213T assembly , which together account for most of the genome , and calculated an average of 96% sequence identity between these and ATCC60483 mtDNA . CLIB213T lacks 2 of the 5 mitochondrial introns that are present in ATCC60483: the omega intron of the large subunit mitochondrial rDNA and intron 2 of COX1 . Intraspecies polymorphism for intron presence/absence and comparable levels of intraspecies mtDNA sequence diversity have been reported in other yeast species [45 , 46] . When genes in the Circos plot are colored according to their parent of origin , it is striking that many Z . parabailii chromosomes are either almost completely “A” ( purple ) or almost completely “B” ( green ) ( outer ring in Fig 1 ) , even though the chromosomes do not form collinear pairs . This pattern can be seen in more detail in a dot-matrix plot between Z . bailii and Z . parabailii ( Fig 3 ) . From this plot , it is evident that most of the A-subgenome is collinear with Z . bailii scaffolds , whereas the B-subgenome contains many rearrangements relative to Z . bailii . For example , Z . parabailii chromosome 1 is derived almost entirely from the B-subgenome but maps to about 12 different regions on the Z . bailii scaffolds . In contrast , Z . parabailii chromosome 3 is derived from the A-subgenome and is collinear with a single Z . bailii scaffold . In total , from Fig 3 we estimate that there are approximately 34 breakpoints in synteny between the Z . parabailii B-subgenome and Z . bailii but no breakpoints between the A-subgenome and Z . bailii , when posthybridization rearrangement events ( described below ) are excluded . This difference in the levels of rearrangement in the A- and B-subgenomes relative to Z . bailii indicates that the 2 subgenomes were not collinear at the time the hybrid was formed . Therefore , most of the rearrangements between the 2 subgenomes are rearrangements that existed between the 2 parental species prior to hybridization . The 2 parents both had 8 chromosomes , but their karyotypes were quite different . Because each event of reciprocal translocation or inversion creates 2 synteny breakpoints [47] , we estimate that about 17 events of chromosomal translocation or inversion occurred between the 2 parents in the time interval between when they last shared a common ancestor and when they hybridized . The situation in Z . parabailii ( hybridization between parents differing by 17 rearrangements and 7% sequence divergence ) contrasts with that in the hybrid Millerozyma sorbitophila ( only 1 detectable rearrangement between the parents , despite 15% sequence divergence [13] ) . Although the Z . parabailii genome largely contains unrearranged parental chromosomes , there have been 2 major types of rearrangement after hybridization . First , posthybridization recombination between the subgenomes at homeologous sites has formed some chromosomes that are partly “A” and partly “B . ” Second , a process of homogenization has occurred at some places in which 1 subgenome overwrote the other , resulting in gene pairs that are A:A or B:B . This process is commonly called loss of heterozygosity ( LOH ) or gene conversion . Based on their KS distances from Z . bailii , the genome contains 4 , 153 simple A:B homeologous gene pairs , 300 A:A pairs , and 84 B:B pairs . To examine the genomic locations of LOH and rearrangement events in more detail , we further classified genes using a scheme that takes account of their pairing status as well as their divergence from Z . bailii . Genes were defined as “A” or “B” as before or “N” if a KS distance from Z . bailii could not be calculated ( Fig 2B and 2C ) . We then assigned each gene to 1 of 7 categories such as “B-gene in an A:B pair” or “A-gene , unpaired” and plotted the locations of genes in each category . The resulting map of the genome ( Fig 4 ) allows LOH and recombination events to be visualized . N-genes ( black in Fig 4 ) are seen to be mostly located near telomeres . Several points of recombination between the A- and B-subgenomes are apparent , such as in the middle of chromosome 4 . LOH tends to occur in stretches that span multiple genes . For example , on chromosome 13 , LOH has formed 8 runs of consecutive A-genes in a chromosome that is otherwise “B”; these A-genes are members of A:A pairs . They were probably formed by homogenization ( gene conversion without crossover ) , although they could also be the result of double crossovers followed by meiotic segregation of chromosomes . Patches of LOH are frequently seen adjacent to sites of recombination between the 2 subgenomes ( Fig 4 ) . Three large regions of apparently unpaired A-genes near the ends of chromosomes ( 1L , 5L , and 9R; light blue in Fig 4 ) are probably artefacts caused by BIR , which is a process that can make the ends of 2 chromosomes completely identical from an initiation point out to the telomere [48] . These regions have 2x sequence coverage in our Illumina data , and we can identify the probable locations of an identical second copy of each of them at other chromosome ends ( Fig 4 ) . The Z . parabailii genome contains 2 MAT loci ( one of which is broken ) and 4 HML/HMR silent loci ( Fig 5 ) . In S . cerevisiae , mating-type switching is a DNA rearrangement process that occurs in haploid cells to change the genotype of the MAT locus [49] . During switching , the active MAT locus is first cleaved by an endonuclease called HO , and its a- or α-specific DNA is removed by an exonuclease . The resulting double-strand DNA break at MAT is then repaired by copying the sequence of either the HMLα or HMRa locus . This process converts a MATa genotype to MATα , or vice versa . Repeated sequences , called Z and X , located beside MAT and the HM loci act as guides for the DNA strand exchanges that occur during this repair process . The HM loci are “silent” storage sites for the a and α sequence information because genes at these loci are not transcribed due to chromatin modification; only MAT is transcribed [49] . We infer that the parents of Z . parabailii each contained a MAT locus and 2 silent loci ( HMLα and HMRa ) , similar to S . cerevisiae and Z . rouxii haploids [50] . Fig 5A shows that Z . parabailii has a MAT locus on chromosome 7 , flanked by Z and X repeats and full-length copies of the genes SLA2 and DIC1 , similar to the MAT loci of many other species [50 , 51] . This MAT locus is derived from Parent A . Chromosome 7 also contains HMLα and HMRa loci ( derived from Parent B ) near its telomeres . However , the B-subgenome’s MAT locus is broken into 2 pieces . Most of it is on chromosome 2 , but its left part ( the 3′ end of MATα1 , the Z repeat , and the neighboring gene SLA2 ) is on chromosome 16 ( Fig 5A ) . Chromosomes 2 and 16 also each contain an HMLα or HMRa locus from the A-subgenome . Examination of the breakpoint in the B-subgenome’s MAT locus shows that the break was catalyzed by HO endonuclease , because it occurs precisely at the cleavage site for this enzyme ( Fig 5B ) . In S . cerevisiae , HO has a long ( approximately 18 bp ) recognition sequence that is unique in the genome , and it cleaves DNA at a site within this sequence , leaving a 4-nucleotide 3′ overhang [52] . Although the recognition and cleavage sites of HO endonucleases in other species have not been investigated biochemically , they can be deduced because the core of the HO cleavage site ( cgcagca ) invariably forms the first nucleotides of the Z region in each species [51] . Moreover , the HO cleavage site corresponds to an amino acid sequence motif ( faqq ) in the MATα1 protein that is strongly conserved among species . The 2 parts of the broken MAT locus are located beside the genes GDA1 and YEF1 ( Fig 5A ) , which are neighbors in Z . bailii CLIB213T and in the Ancestral yeast genome [38 , 41] . Therefore , after HO endonuclease cleaved the “B” MAT locus , the broken ends of the chromosome apparently interacted with the GDA1-YEF1 intergenic region of the A-subgenome , causing a reciprocal translocation . This site is the only synteny breakpoint between the A-subgenome of Z . parabailii and the genome of Z . bailii ( scaffold 9; Fig 3 ) . Comparison of the DNA sequences at the site ( Fig 5B ) shows no microhomology between the 2 interacting sequences and that DNA repair led to duplications of a 5-bp sequence ( acaac ) from the GDA1-YEF1 intergenic region and a 2-bp sequence ( ca ) from MATα1 , suggestive of nonhomologous end joining ( NHEJ ) as the repair mechanism . We hypothesize that this genomic rearrangement occurred during a failed attempt to switch mating types , which resulted in a reciprocal translocation instead of normal repair of MAT by HML or HMR . While the B-subgenome’s MATα1 gene is clearly broken , its MATα2 gene also appears to be nonfunctional . MATα2 has 2 introns , and our RNA-Seq data show how both homeologs of this gene ( ZPAR0G01480_A and ZPAR0B05090_B ) are spliced . A point mutation at the 3′ end of intron 2 of the B-gene changed its AG splice acceptor site to AC , with the result that splicing now uses another AG site 2 nucleotides downstream ( Fig 5C ) . This change results in a frameshift , truncating the B-copy of the α2 protein to 57 amino acid residues instead of 211 and presumably inactivating it . Surprisingly , the Z . parabailii genome does not contain any MATa1 ( or HMRa1 ) gene . This gene codes for the a1 protein , which is 1 subunit of the heterodimeric a1-α2 transcriptional repressor that is formed in diploid ( a/α ) cells and which acts as a sensor of diploidy by repressing transcription of haploid functions such as mating while permitting diploid functions such as meiosis [53] . The a1 gene is present in Z . rouxii and Z . sapae [27 , 37 , 50] , but it is also absent from Z . bailii CLIB213T and must have been absent from Parent B . The Z . bailii CLIB213T MAT organization is not fully resolved [38] , but it contains a MAT locus with α1 and α2 genes on scaffold 14 and an HMR locus with only an a2 gene on scaffold 19 . Evolutionary losses of MATa1 have previously been seen in some Candida species [54 , 55] , but not in any species of family Saccharomycetaceae . In contrast , the gene for the other subunit of the heterodimer , MATα2 , is present in all Zygosaccharomyces species and is probably maintained because it has a second role in repressing a-specific genes in this genus [56] . Solieri and colleagues have reported evidence that a1-α2 is nonfunctional in a Z . rouxii/pseudorouxii hybrid in which its 2 subunits are derived from different species [14] . The 2 subgenomes apparent in the Illumina scaffolds of the Zygosaccharomyces hybrid strain ISA1307 , previously sequenced by Mira et al . [39] , are both 99%–100% identical in sequence to the A- or B-subgenomes of ATCC60483 . Therefore , ISA1307 is also a strain of Z . parabailii . Importantly , the ISA1307 genome sequence contains the same HO-catalyzed reciprocal translocation between MATα1 of the B-subgenome and the GDA1-YEF1 intergenic region of the A-subgenome ( Fig 5A ) . Because this rearrangement is so unusual and because it did not involve recombination between repeated sequences , it is highly unlikely to have occurred twice in parallel . The rearrangement is much more likely to have occurred only once , in a common ancestor of the 2 Z . parabailii strains after the hybrid was formed . It cannot pre-date the hybridization because it formed junctions between the A- and B-subgenomes , which originated from different parents . ATCC60483 and ISA1307 are independent isolates of Z . parabailii , both from industrial sources . ATCC60483 was isolated from citrus concentrate used for soft drink manufacturing in the Netherlands [57 , 58] , and ISA1307 was a contaminant in a sparkling wine factory in Portugal [39 , 59–61] . We found several examples in which the 2 strains differ in their patterns of LOH , which confirms that they have had some extent of independent evolution . All 3 large regions of BIR ( on chromosomes 1 , 5 , and 9; Fig 4 ) are unique to ATCC60483 . ISA1307 contains A:B homeolog pairs throughout these regions , whereas ATCC60483 has only A-genes , which we infer to be in A:A pairs . Other examples of differential LOH include a 4-kb region around homologs of the S . cerevisiae gene YLR049C , which exists as B:B pairs in ATCC60483 but A:B pairs in ISA1307 , and the gene KAR4 , which is an A:B pair in ATCC60483 but only a B-gene ( single contig ) in ISA1307 . Notably , the section of the RPB1 gene ( also called RPO21 ) that Suh et al . [28] used for taxonomic identification of Z . parabailii and Z . pseudobailii exists as an A:B pair in ATCC60483 , but only as an A-gene in the ISA1307 genome . The absence of the B-copy of RPB1 made Mira et al . [39] hesitant to conclude that ISA1307 is Z . parabailii . Both ATCC60483 and the type strain of Z . parabailii ATCC56075T have previously been reported to be capable of forming ascospores [28 , 57 , 58] . We confirmed that our stock of ATCC60483 is able to sporulate ( Fig 6A and 6B ) . On malt extract agar plates , we observed that sporulation occurs directly in zygotes formed by conjugation between 2 cells , resulting in asci in which the 2 former parental cell bodies typically contain 2 ascospores each . Such dumbbell-shaped ( conjugated ) asci , indicative of sporulation immediately after mating , are characteristic of the genus Zygosaccharomyces [25] and have previously been described in other Z . bailii ( sensu lato ) strains [25 , 62–66] . The presence of conjugating cells in a culture grown from a single strain indicates that ATCC60483 is functionally haploid ( capable of mating ) and that it is homothallic ( capable of mating-type switching ) . Since the zygote proceeds immediately into sporulation without further vegetative cell divisions , the diploid state of Z . parabailii appears to be unstable . Although Suh et al . [28] reported that asci of the type strain of Z . parabailii contain 2 spores , we consistently observed that asci occur in pairs of mated cells connected by a conjugation tube ( Fig 6A and 6B ) , indicating that 4 spores are formed per meiosis . We dissected tetrad asci from ATCC60483 , grew colonies from the spores , and then used colony PCR to determine their genotype at the intact MAT locus on chromosome 7 . Among 13 tetrads analyzed , 9 showed a ratio of 2 MATa colonies to 2 MATα colonies ( Fig 6C and 6D ) . Two tetrads showed 1:3 or 3:1 ratios , and the other two yielded both MATa and MATα PCR products from some single-spore colonies . The genotype of the ATCC60483 starting strain is MATα from the A-subgenome ( designated MATα_A ) , so the presence of MATa genotypes in colonies derived from spores made by this strain confirms that mating-type switching occurred at some point . We sequenced the PCR products and found that the A- and B-subgenome HMRa loci were both used as donors for mating-type switching: among the pure MATa colonies , 18 were MATa_A , and 7 were MATa_B ( Fig 6D ) . Quite surprisingly , 4 tetrads with 2a:2α segregation had 1 MATa_A and 1 MATa_B spore colony , which is inconsistent with simple meiotic segregation from an a/α diploid . Because all the spores contain a functional HO gene , the genotypes of these 4 tetrads ( #1 , #7 , #19 , and #20 ) probably result from additional switches during the early growth of some colonies . Similarly , switching during early colony growth may explain the presence of MATα_B genotypes in tetrad #11 and the colonies with mixed a+α genotypes ( in tetrads #11 and #13 ) , as well as the presence of faint PCR products corresponding to the alternative MAT genotype in some other colonies ( Fig 6C ) . In S . cerevisiae , homothallic diploid ( HO/HO MATa/MATα ) strains show 2:2 segregation of MAT alleles in tetrads , but after spore germination the haploid cells can then switch mating types as often as once per cell division [67] , leading to mating and colonies that contain mostly diploid cells [68]; by contrast , most ( but not all ) of the Z . parabailii spore-derived colonies contained a single mating type ( Fig 6C and 6D ) . We found that almost all the genes involved in mating and meiosis that Mira et al . [39] reported to be missing from the Z . parabailii ISA1307 genome are in fact present in both ATCC60483 and ISA1307 ( S1 Table ) . For example , we annotated A- and B-homeologs of IME1 , UME6 , DON1 , SPO21 , SPO74 , REC104 , and DIG1/DIG2 as well as MATa2 , MATα1 , and MATα2 . We also identified genes for the α-factor and a-factor pheromones ( MFα and MFa ) . The MFα genes code for an unusually high number of copies ( 10–14 ) of a 13-residue peptide whose consensus sequence , ahlvrlspgaamf , is quite different from that of other yeasts , including Z . rouxii ( 7/13 matches ) and S . cerevisiae ( 4/13 matches ) [2] . Z . parabailii and Z . bailii do lack most of the ZMM group of genes , involved in crossover interference during recombination [69] , even though these are present in Z . rouxii ( S1 Table ) . Interestingly , identical sets of ZMM genes have been lost in Z . bailii/Z . parabailii relative to Z . rouxii , as were lost in most Lachancea species relative to Lachancea kluyveri [70]: ZIP2 , CST9 ( ZIP3 ) , SPO22 ( ZIP4 ) , MSH4 , MSH5 , and SPO16 are absent , as well as MLH2 , which is not known to be a ZMM gene , whereas ZIP1 is retained . A similar loss of ZMM genes has occurred in Eremothecium gossypii relative to E . cymbalariae [71] . A small number of Z . parabailii ATCC60483 genes have “disabling” mutations—frameshifts or premature stop codons that prevent translation of a normal protein product . The majority of these mutations are present in only 1 subgenome of ATCC60483 and are unique to this strain . For example , there is a 1-bp insert in the A-homeolog of the DNA repair gene MLH1 that is not present in the B-homeolog or in ISA1307 or CLIB213T . In a systematic search , we found a total of 10 A-genes and 9 B-genes that were inactivated only in strain ATCC60483 ( S2 Table ) . In each case , the other homeolog was intact , and the mutations , discovered in the PacBio assembly , were confirmed by our Illumina contigs of the ATCC60483 genome . We found a further 8 disabling mutations that are shared between ATCC60483 and ISA1307 . One of these is the AC-to-AG splice site mutation in the B-homeolog of MATα2 described above ( Fig 5C ) . Another is the HO endonuclease gene , whose A-homeolog contains an identical 1-bp deletion in both ATCC60483 and ISA1307 , whereas the B-homeolog of HO is intact in both strains ( S2 Table ) . It is perhaps surprising that the HO gene that degenerated is the A-homeolog , whereas the broken MAT locus is the B-homeolog , but the 2 endonucleases are likely to have had identical site specificities because the HO cleavage site is well conserved among species . The existence of these 8 shared disabling mutations provides further support for the idea that the 2 strains of Z . parabailii are descended from the same hybrid ancestor , because these mutations may not be viable in the absence of the intact homeologous copies of these genes . Only one of them is present also in CLIB213T ( S2 Table ) . We annotated 447 introns in the Z . parabailii ATCC60483 genome , most of which are confirmed by our RNA-Seq data . There are 428 intron-containing genes , including 19 that have 2 introns . We did not find any examples of intron presence/absence differences between homeologs . Interestingly , we found several genes with an in-frame intron—that is , an intron that is a multiple of 3 bp long and contains no stop codons , so that both the spliced and unspliced forms of the mRNA can be translated into proteins . Genes with in-frame introns are likely to undergo alternative splicing , making 2 forms of the protein with different functions . One of these loci is PTC7 ( ZPAR0J04940_A and ZPAR0A06900_B ) . Both of the Z . parabailii homeologs contain a 69-bp in-frame intron within the open reading frame ( ORF ) of the gene . It has previously been shown that alternative splicing of a similar in-frame intron in S . cerevisiae PTC7 leads to the translation of a mitochondrial protein isoform from the spliced mRNA and a nuclear envelope protein isoform from the unspliced mRNA and that the intronic region codes for a transmembrane domain of the protein [72] . Thus , the alternative splicing mechanism in PTC7 is conserved between Saccharomyces and Zygosaccharomyces . We also found in-frame introns in the Z . parabailii orthologs of S . cerevisiae NUP100 , NCB2 , and HEH2 , identically in their A- and B-homeologs . None of these genes is known to be alternatively spliced in S . cerevisiae . In each of these examples , there are typical splice donor , branch , and acceptor sequences within the long form of the ORF . Programmed “+1” ribosomal frameshifting , a process whereby the ribosome skips forward by 1 nucleotide when translating an mRNA , is known to occur in 3 genes in S . cerevisiae: OAZ1 , ABP140 , and EST3 [73] , and we found that +1 frameshifting is also required to translate the Z . parabailii orthologs of these 3 genes , in both the A- and B-homeologs . We also found 2 new loci that apparently undergo +1 frameshifting . Translation of both homeologs of BIR1 ( ZPAR0O02690_A and ZPAR0I02720_B ) requires a +1 frameshift at a sequence identical to the EST3 frameshifting site: CTT-A-GTT , where the A is the skipped nucleotide . Translation of both homeologs of YJR112W-A ( ZPAR0O02960_A , ZPAR0I02990_B ) requires a +1 frameshift at a sequence identical to the ABP140 frameshifting site: CTT-A-GGC . In S . cerevisiae , the CUP1 locus confers resistance to copper toxicity by a gene amplification mechanism . CUP1 codes for a metallothionein , a tiny cysteine-rich copper-binding protein . The reference S . cerevisiae genome sequence contains 2 identical copies of CUP1 duplicated in tandem , but under copper stress this locus can become amplified to contain up to 18 tandem copies of the gene [74 , 75] . There are at least 5 different types of CUP1 repeats in different S . cerevisiae strains , which must have originated independently from progenitors with a single CUP1 gene [75 , 76] . In Z . parabailii , we found a slightly different organization . At homeologous loci on chromosomes 2 and 7 , ATCC60483 has multiple identical copies of a 1 , 454-bp repeating unit . Each unit contains 2 metallothionein genes , MT-58 and MT-47 , coding for proteins of 58 and 47 residues , respectively . There is only 56% amino acid sequence identity between MT-58 and MT-47 proteins . The chromosome 7 locus contains 5 copies of the repeating unit , and the chromosome 2 locus contains 2 copies , so ATCC60483 has 14 metallothionein genes in total . These loci are not syntenic with S . cerevisiae CUP1 , but they are syntenic with metallothionein genes in C . glabrata and Z . rouxii [77 , 78] .
Our results show that Z . parabailii is a hybrid species that was formed by fusion between two 8-chromosome parental species , one of which was Z . bailii . The low sequence divergence of the ATCC60483 A-subgenome from the type strain of Z . bailii ( the modal synonymous site divergence is less than 1%; Fig 2A ) and the almost complete collinearity of these genomes ( Fig 3 ) indicate that the A-parent of Z . parabailii should be regarded as Z . bailii itself , and not merely as a species closely related to Z . bailii . The unusual MAT locus structure of this hybrid raised questions about how it was formed and whether Z . parabailii currently has a full sexual cycle . At first glance , the MATα/MATα hybrid genotype of ATCC60483 might suggest that Z . parabailii could not have been formed by mating . However , this genotype could also be the result of mating-type switching . We propose that the following steps occurred ( Fig 7 ) . Z . parabailii was formed by mating between strains of parent A ( Z . bailii ) and parent B , of opposite mating types . These parental genomes already differed by about 34 chromosomal rearrangement breakpoints , so the hybrid was unable to produce viable spores by meiosis . The hybrid also had no MATa1 gene , so it could not form the a1-α2 heterodimer that stabilizes the diploid state in S . cerevisiae [68] . One of the roles of the a1-α2 dimer in S . cerevisiae is to repress transcription of HO endonuclease , which is only required in haploid cells . We suggest that in the newly formed Z . parabailii hybrid , transcription of HO was not repressed . Continued expression of this gene resulted in genotype switching at the MAT loci ( perhaps several consecutive switches between a and α ) and , eventually , breakage of the B-subgenome MAT locus due to an illegitimate recombination with the GDA1-YEF1 intergenic region instead of HML or HMR . At some point after hybridization , the HO gene from the A-subgenome also degenerated by acquiring a frameshift mutation . The breakage of the “B” MAT locus can be inferred to have been one of the first rearrangement events that occurred after the hybridization but also to have been recent . It must have been one of the first posthybridization events , because the GDA1-YEF1 breakage that occurred simultaneously with it is the only point of noncollinearity between the A-subgenome and the Z . bailii genome ( apart from sites of interhomeolog recombination or homogenization; Fig 4 ) . It must have been recent because the pseudogene fragments of the broken MAT locus have not yet accumulated any other mutations . There are no nucleotide differences in 2 , 298 bp between the broken MATα_B locus on chromosome 2 and HMLα_B on chromosome 7 . Together , these 2 observations suggest that the interspecies mating that formed Z . parabailii occurred less than 105 generations or 1 , 000 years ago [79] . Such a recent origin is consistent with the very low numbers of gene inactivations that have occurred since hybridization , with the fact that most of these are not shared between the 2 sequenced Z . parabailii strains ( only 8 of 27 inactivating mutations are shared; S2 Table ) , and with the retention of rDNAs from both parents . We expect that , if the Z . parabailii lineage survives , it will accumulate extensive inactivations and deletions of redundant duplicated genes over the next few million years , as seen in older WGDs . The net result of the evolutionary changes to the genome is that Z . parabailii now has 16 chromosomes ( all different in structure but containing homeologous regions ) , 1 active MAT locus , 1 active HO gene , and 4 silent HML/HMR loci . A genome with this structure resembles haploid S . cerevisiae [1] and is potentially capable of both mating-type switching and mating . We confirmed that both of these processes occur in ATCC60483 . Z . parabailii has a life cycle in which 16-chromosome haploids mate to produce 32-chromosome diploids ( Fig 7 ) that sporulate immediately because the diploid state is unstable; there is no MATa1 gene , and hence , there is no a1-α2 heterodimer . Thus , Z . parabailii is an allopolyploid that regained fertility by genome doubling after interspecies mating , as a consequence of damage to 1 copy of its MAT locus . Two previous reports that Z . parabailii strains produce only mitotic spores [31 , 80] can be reinterpreted in view of the hybrid nature of the genome . Their experimental data are fully compatible with the meiotic sexual cycle we propose for Z . parabailii . Rodrigues et al . [80] made a derivative of ISA1307 in which 1 copy of ACS2 was disrupted by the G418-resistance marker APT1 and the other copy was not . After sporulation of this strain , all 80 spores they tested were G418 resistant , and all 16 spores from 4 tetrads contained both an intact copy of ACS2 and an acs2::APT1 disruption , which led Rodrigues et al . [80] to conclude that the spores were made by mitosis . However , this inheritance pattern is exactly the pattern expected if the 2 copies of ACS2 are homeologs ( different Mendelian loci ) rather than alleles of a single Mendelian locus and if ISA1307 is a haploid that autodiploidized before it sporulated . Thus , their strain could be described as haploid acs2_a::APT1 ACS2_B , where the ACS2_A and ACS2_B loci have independent inheritance ( they are on chromosomes 10 and 13 in our genome sequence ) . Similarly , Mollapour and Piper [31] disrupted 1 of the 2 copies of YME2 in strain NCYC1427 with a kanMX4 cassette and found that all the spores produced by this strain retained both an intact YME2 and yme2::kanMX4 . They concluded that the spores were vegetative , but again , the result is consistent with meiotic spore production if the 2 YME2 loci have independent inheritance ( they are on chromosomes 4 and 6 ) and if the disruption was made in a haploid strain that autodiploidized before sporulating . The sequence data in [31] allow NCYC1427 to be identified as Z . parabailii and not Z . bailii as originally described . Furthermore , in both ISA1307 [80] and NCYC1427 [25 , 31] , spores are formed in pairs of conjugated cells , similar to Fig 6A . We conclude that ISA1307 and NCYC1427 have sexual cycles identical to the one we describe for ATCC60483 . The evolutionary steps that formed Z . parabailii by interspecies mating , and restored its fertility by damage to one of its MAT loci , are essentially identical to one of the mechanisms ( hypothesis B ) proposed for the origin of the ancient WGD in the S . cerevisiae lineage [3 , 18–20] . Our study therefore validates genome doubling after MAT locus damage as a real evolutionary process that occurs in natural interspecies hybrids , enabling them to resume mating and meiosis . The Z . parabailii hybridization was very recent , so any period of clonal reproduction that elapsed before fertility was restored must have been short , which is as expected because there is no selection to maintain meiosis genes during clonal growth [18 , 20] . The possible role of MATa1 in the ancient WGD remains unclear . In Zygosaccharomyces , the absence of this gene makes zygotes proceed into sporulation . In the ancient WGD , it is likely that a MATa1 gene was present in the initial zygote , in which case the zygote would have been stable until it sustained MAT locus damage , but this is not certain because the ZT parent might have lacked MATa1 . The specific cause of damage to the MAT locus in Z . parabailii was incorrect DNA repair after cleavage by the mating-type switching endonuclease HO . The HO gene is present in the ZT clade , but not in the KLE clade [51 , 81] , and these 2 clades were the 2 parental lineages of the interspecies hybridization that led to the ancient WGD [3] . Species that contain HO show evolutionary evidence of repeated deletions of DNA from beside their MAT loci , caused by accidents during mating-type switching [51] . Indeed , the disappearance of the MATa2 gene from Saccharomycetaceae genomes , which occurred at approximately the same time as the WGD , must have been due to some sort of mutational damage to the MAT locus . Although HO-mediated damage can only occur in the small clade of yeasts that contain HO , other types of mutational damage to 1 copy of MAT are a plausible mechanism for fertility restoration in other fungal interspecies hybrids .
The strain analyzed here originally came from the collection of Thomassen & Drijver-Verblifa NV in the Netherlands [57 , 58] and was called “Saccharomyces bailii strain 242” in those studies . It was isolated from citrus concentrate being used as raw material for soft drinks . It was later deposited at the American Type Cultures Collection as ATCC60483 . Suh et al . [28] identified it as Z . parabailii by molecular methods . ATCC60483 genomic DNA was prepared using the Blood & Cell Culture DNA Mini Kit ( Qiagen ) , according to the manufacturer’s manual . To prevent fragmentation of the DNA , the sample was not vortexed . The final genomic DNA amount was 15 μg as determined by Qubit Fluorometer ( Thermo Scientific ) . PacBio sequencing was carried out by the Earlham Institute ( Norwich , United Kingdom ) using 8 SMRT cells , which generated 218x mean coverage for the nuclear scaffolds . We assembled the raw data using the computational facilities at the Irish Centre for High-End Computing ( ICHEC ) , with the HGAP3 protocol of the SMRT Analysis suite version 2 . 3 . 0 [82] . We initially obtained 22 nuclear scaffolds , which we reduced to 16 chromosomes by manually identifying overlaps between scaffolds . In parallel , we also obtained 198x Illumina read coverage of the genome ( Genome Analyzer IIx; University of Milano-Bicocca , Department of Clinical Medicine ) , which we assembled separately into contigs that were used to verify the status of rearrangement points and pseudogenes discussed in the text . The Z . parabailii chromosomes were annotated using an improved version of our automated YGAP [83] , which uses information in the Yeast Gene Order Browser [78] and the Ancestral ( pre-WGD ) gene order [41] to generate a synteny-based annotation . The automated annotation was curated using transcriptome data from ATCC60483 cultures grown in a bioreactor; Illumina RNA-Seq was generated at Parco Tecnologico Padano ( Italy ) . We made a de novo transcriptome assembly using Trinity [84] and compared the transcripts against YGAP’s gene models using PASA [85] and by manual inspection of spliced mRNA reads . Chromosomes were numbered 1 to 16 from largest to smallest . Genes were given systematic names by YGAP such as ZPAR0D01210_B , where ZPAR indicates the species; 0 indicates the genome sequence version; D indicates chromosome 4; 01210 is a sequential gene number counter that increments by 10 for each protein-coding gene ( genes that were added manually have numbers that end in 5 or other digits ) ; and the suffix _B indicates that this gene is assigned to the B-subgenome as described below . NCBI nucleotide sequence database accession numbers are CP019490–CP019505 ( nuclear chromosomes ) , CP019506 ( mitochondrial genome ) , and CP019507 ( 2-micron plasmid ) . The mitochondrial genome of Z . bailii CLIB213T was not reported with the rest of this strain’s genome [38] and is highly fragmented in the assembly . We identified mitochondrial contigs in the original CLIB213T assembly by BLASTN using the ATCC60483 mtDNA as a query , assembled these contigs into 55 larger contigs using the CAP3 assembler and SSPACE3 [86 , 87] , and calculated a weighted average nucleotide identity of 96% from nonoverlapping alignments totaling 23 , 197 bp . We assigned most genes in Z . parabailii ATCC60483 to either the A-subgenome ( highly similar to the Z . bailii CLIB213T genome ) or the B-subgenome ( derived from the other parent in the hybridization ) , using their levels of synonymous nucleotide sequence divergence from CLIB213T genes . For this purpose , we used BLASTP [88] to compare every annotated protein from ATCC60483 to the CLIB213T proteome and designated the best hit as a homolog . The corresponding ATCC60483 and CLIB213T DNA sequence pairs were then aligned using CLUSTALW [89] , and their levels of sequence divergence were calculated using the yn00 program from the PAML suite [90] . ATCC60483 genes were assigned to the A-subgenome if the level of synonymous divergence was KS ≤ 0 . 05 and to the B-subgenome if 0 . 05 < KS ≤ 0 . 25 and given an _A or _B suffix on the gene name accordingly . Genes for which KS > 0 . 25 or for which no Z . bailii homolog was identified were given the suffix _N . To identify inactivated genes systematically , we searched the annotated A:B gene pairs for cases in which one of the homeologs was less than 90% of the length of the other , and we then examined these cases manually ( S1 Table ) . Note that our use of the labels “A” and “B” differs from the scheme used by Mira et al . [39] for strain ISA1307 . We designated each gene ( homeolog ) as either “A” or “B” based on its divergence from Z . bailii CLIB213T , with “A” always indicating the Z . bailii-like homeolog . Some chromosomes therefore contain mixtures of “A” and “B” genes due to posthybridization recombination or homogenization between the 2 subgenomes . In contrast , Mira et al . [39] identified homeologous pairs of scaffolds in their assembly and arbitrarily designated 1 scaffold as “A” and the other as “B” so that each scaffold is homogeneous , but there is no consistent relationship between the “A” and “B” labels and the parent-of-origin of a homeolog in their scheme . Cells were left for sporulation on malt extract ( 5% ) agar for 5 days . A small loop of cells was washed in sterile distilled water , resuspended in a 1:20 dilution of Zymolyase 100T , and incubated for 10 min at 30°C . The Zymolyase solution was removed by centrifugation , and the pellet resuspended in distilled water ( 500 μl ) . A 10-μl drop was placed in the middle of a YPD plate , and dumbbell-shaped asci were dissected using a Singer Sporeplay dissection microscope . The YPD plate was incubated for 2 days at 30°C . Individual spore-derived colonies were used for MAT locus genotyping by colony PCR using Q5 polymerase high-fidelity 2x master mix ( NEB ) and annealing temperature 55°C . Sequences of PCR primers A–F are given in S3 Table . Primers E and F were designed to bind equally to the HMR regions of the A- and B-subgenomes . Primers A–D are specific for the A-subgenome .
|
It has recently been proposed that the whole-genome duplication ( WGD ) event that occurred during evolution of an ancestor of the yeast S . cerevisiae was the result of a hybridization between 2 parental yeast species that were significantly divergent in DNA sequence , followed by a doubling of the genome content to restore the hybrid’s ability to make viable spores . However , the molecular details of how genome doubling could occur in a hybrid were unclear because most known interspecies hybrid yeasts have no sexual cycle . We show here that Z . parabailii provides an almost exact precedent for the steps proposed to have occurred during the S . cerevisiae WGD . Two divergent haploid parental species , each with 8 chromosomes , mated to form a hybrid that was initially sterile but regained fertility when 1 copy of its mating-type locus became damaged by the mating-type switching apparatus . As a result of this damage , the Z . parabailii life cycle now consists of a 16-chromosome haploid phase and a transient 32-chromosome diploid phase . Each pair of homeologous genes behaves as 2 independent Mendelian loci during meiosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"fungal",
"spores",
"chromosome",
"structure",
"and",
"function",
"centromeres",
"sequence",
"assembly",
"tools",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"genome",
"analysis",
"fungal",
"reproduction",
"saccharomyces",
"research",
"and",
"analysis",
"methods",
"genome",
"complexity",
"mycology",
"genomics",
"chromosome",
"biology",
"genetic",
"loci",
"yeast",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"computational",
"biology",
"introns",
"organisms",
"chromosomes"
] |
2017
|
Evolutionary restoration of fertility in an interspecies hybrid yeast, by whole-genome duplication after a failed mating-type switch
|
Asymmetric division of zygote is critical for pattern formation during early embryogenesis in plants and animals . It requires integration of the intrinsic and extrinsic cues prior to and/or after fertilization . How these cues are translated into developmental signals is poorly understood . Here through genetic screen for mutations affecting early embryogenesis , we identified an Arabidopsis mutant , zygotic arrest 1 ( zar1 ) , in which zygote asymmetric division and the cell fate of its daughter cells were impaired . ZAR1 encodes a member of the RLK/Pelle kinase family . We demonstrated that ZAR1 physically interacts with Calmodulin and the heterotrimeric G protein Gβ , and ZAR1 kinase is activated by their binding as well . ZAR1 is specifically expressed micropylarly in the embryo sac at eight-nucleate stage and then in central cell , egg cell and synergids in the mature embryo sac . After fertilization , ZAR1 is accumulated in zygote and endosperm . The disruption of ZAR1 and AGB1 results in short basal cell and an apical cell with basal cell fate . These data suggest that ZAR1 functions as a membrane integrator for extrinsic cues , Ca2+ signal and G protein signaling to regulate the division of zygote and the cell fate of its daughter cells in Arabidopsis .
Asymmetric cell division is a fundamental process which produces daughter cells with different size and cell fate during embryonic and postembryonic development . Irrespective of cell types or organisms , it requires a common set of coordinated events including the establishment and transduction of polarity , and cytokinesis [1–3] . Cell polarity can be set up by intrinsic or/and extrinsic factors [4] . During early embryogenesis , the zygotic polarity is established either by maternal determinants prior to fertilization in Drosophila , or by sperm entry at fertilization in Caenorhabditis elegans . In higher plants , studies have been largely focused on easily accessible epidermal and root cells [5 , 6] , little is known about molecular mechanisms underlying the asymmetric division of zygote . In Arabidopsis , a YODA-mediated MAPK pathway was shown to be critical for asymmetric cell division in stomata and zygote development [5 , 7–9] . So far , the BASL-MAPK signaling feedback was shown to control stomata asymmetric division . The phosphorylation of BASL ( BREAKING OF ASYMMETRY IN THE STOMATAL LINEAGE ) mediated by MPK3/6 , is required for the cortical localization of BASL . YODA and MPK3/6 are subsequently recruited by the phosphorylated BASL . The feedback loop is further promoted by MPK3/6 . It is demonstrated that cell polarity and fate determination are reinforced and connected by a positive feedback loop of BASL-MAPK during asymmetric division [8 , 9] . YODA encodes a MAPKK kinase that promotes zygote elongation and the basal extra-embryonic cell fate [10] . The MAPKK kinase cascade , on the other hand , is likely activated by the paternal SHORT SUSPENSOR ( SSP ) [11 , 12] . However , the kinase activity of SSP is not required for YODA activation . A small nuclear protein , GROUNDED ( GRD ) , is also required for zygote elongation and the first asymmetric division to establish the basal cell fate [7 , 13] . Recently , it was reported that EMBRYO SURROUNDING FACTOR 1 ( ESF1 ) peptides from central cell before fertilization act with SSP to promote suspensor elongation through the YODA pathway [14] . These suggest that the conserved MAPK cascade plays a key role in zygote asymmetric division and basal cell fate determination . In addition , WOX ( Wuschel-related homeobox ) genes also play critical roles during early embryogenesis and serve as cell fate determinants of the apical and basal cell lineages [15 , 16] . WOX genes , on the other hand , are directly activated by other transcription factors like WRKY2 [17] . In general , extracellular stimuli are received by membrane receptor kinases , and subsequently integrated and transduced inward via numerous signaling molecules [18] . Question remains to be elucidated that how the extracellular stimuli are perceived during early embryogenesis , and how the receptor kinases activate downstream MAPK signaling cascade need to be identified , too . To gain insights into molecular mechanisms controlling zygote development , a detailed screen of our Ds insertion collections for mutations affecting early embryogenesis was performed [19] . A Ds insertion mutant , zygotic arrest 1 ( zar1 ) , whose zygote elongates normally but fails to perform asymmetric division , was identified . Furthermore , cell fate specification of both apical and basal cells is affected by the mutation manifested by the mis-expression of cell-specific markers . The enforcement of the basal and apical cell fate is likely dependent on ZAR1 and AGB1 functions . ZAR1 encodes a leucine-rich repeat receptor-like kinase ( LRR-RLK ) that contains a putative CaM-binding domain and a Gβ-binding motif within its intracellular kinase region . Our data indicate that ZAR1 kinase activity is activated through its direct interaction with CaM1 and the heterotrimeric G protein Gβ ( AGB1 ) . We hypothesize that ZAR1 integrates extracellular stimuli with intracellular Ca2+ and G-protein signaling , to modulate zygotic division in Arabidopsis .
Double fertilization is a unique reproductive process of flowering plants , in which two female gametes ( the egg and the central cell ) in the embryo sac ( Fig 1A ) fuse with two male gametes ( the sperms ) , to produce zygote ( ( Fig 1B ) and endosperm , respectively . Following a quiescent stage after fertilization , the zygote undergoes a series of morphological changes that lead to the establishment of zygote polarity and the zygote elongates to about three folds along the future apical-basal embryo axis ( Fig 1B and 1C ) . Subsequently , an asymmetric division occurs and a small apical cell which gives rise to the embryo proper , and a large basal cell that forms the suspensor connecting the embryo and the mother tissue , are produced ( Fig 1D–1F ) . The uppermost cell of the basal lineage forms the hypophysis that is ultimately incorporated into the embryonic root . This stereotyped development in Arabidopsis serves as a model for genetic dissection of early embryo development in flowering plants [20 , 21] . In zar1-1+/- ( for the simplicity , the heterozygous mutant is marked as zar1-1+/- and the homozygous as zar1-1 or zar1-1-/- ) siliques , three days after pollination ( DAP ) , 73 . 2±0 . 34% ( n = 1286 ) ovules contain early globular embryo as in the wild-type ( Fig 1G ) , while the other ovules contain either an elongated zygote with an obvious nucleus almost localized at the central ( 8 . 6±1 . 5% , n = 1286 ) ( Fig 1H ) or two similar cells from a nearly symmetric division ( 15 . 7±1 . 9% , n = 1286 ) ( Fig 1I ) . In total , they account for about 26 . 8% ( n = 1286 ) of the ovules . This indicates that the mutation does not affect zygote elongation but hinders it from division or disrupts the first asymmetric division . These ovules are also defective in endosperm development ( S1 Fig ) . Genetic analysis showed that the KanR/KanS ratio in selfed zar1-1+/- progenies is 1 . 8:1 ( n = 1038 ) , close to the ratio 2:1 ( χ2 = 1 . 35 ) for embryo lethal mutation; the KanR/KanS is 1:1 . 03 ( n = 788 ) when zar1-1+/- was used as male or 1:1 . 19 ( n = 879 ) as female in crosses with the wild-type . These indicate that homozygous zar1-1-/- is embryo lethal , but male or female gametogenesis is not impaired by the mutation . The ratio of the normal to abnormal ovules is 2 . 7:1 , in agreement with a typical recessive mutation of 3:1 segregation ( P < 0 . 05 , χ20 . 05 = 3 . 84 ) . Meanwhile , we obtained two weak alleles , zar1-2 ( a Ds insertion line ) and zar1-3 ( a T-DNA mutant ) . Both zar1-2 and zar1-3 are homozygous with very similar phenotype . In the zar1-2 siliques at 3 DAP , the seed set rate is 98±2 . 02% ( n = 526 ) . However , about 19 . 41±0 . 52% ( n = 328 ) zygotes undergo symmetric , or approximately symmetric division in zar1-2 siliques ( Fig 1J and 1K ) , although other zygotes undergo asymmetric division ( Fig 1L ) as the wild-type . Statistical analysis indicates that the length of zar1-2 basal cell ( 39 . 33±5 μm , n > 30 ) is significantly reduced compared to the wild-type ( 58 . 9±5 . 1 μm , n > 30 ) ( see Fig 2A , 2C and 2G ) , because of an approximately symmetric division of zygote . Nevertheless , an early globular embryo is observed in every ovule two days later ( Fig 1M ) , suggesting that there is an endurable disruption during early embryogenesis . Together , these indicate that ZAR1 plays a critical role in zygote asymmetric division , and genetic redundancy exists for ZAR1 function . To test whether the mutation affects the cell fate of the apical and basal cell lineages , we checked the expression of a basal cell-specific marker pWOX8gΔ:NLS-vYFP3 [15] and an apical cell-specific marker pWOX2:DsRed2 [17] . As previously reported in wild-type plants , the YFP is only detected in zygote and the basal cell lineage , but absent in the apical cell lineage of pWOX8gΔ:NLS-vYFP3 transgenic plants ( Fig 3A–3F ) . The apical cell lineage marker DsRed signal is detected in zygote , apical cell , the apical cell lineage and cells adjacent to embryo proper from basal cell lineage ( Fig 3A–3F ) . In zar1-1 arrested ovules , very weak YFP signal was detected in elongated zygote , but no YFP was observed in apical cell , or basal cell ( Fig 3G and 3H , S2I–S2L Fig ) , indicating that the identity of these cells is impaired in zar1-1 plants . In zar1-2 , however , the YFP signal was strong in zygote ( Fig 3I ) and both the apical and basal cells ( Fig 3J ) as in the wild-type ( Fig 3A and 3B ) , and unexpectedly present in the embryo proper besides the basal lineage ( Fig 3K–3N , S2 Table ) as compared to the wild-type ( Fig 3C–3F ) . In comparison , the DsRed was only detected in elongated zygote ( Fig 3G ) but not in the apical or basal cells in zar1-1 arrested ovules ( Fig 3H ) . We further analyzed a trans-heterozygote made with zar1-1+/-/pWOX8gΔ:NLS-vYFP3/pWOX2:DsRed2 pollinated with zar1-2-/-/pWOX8gΔ:NLS-vYFP3/pWOX2:DsRed2pollen grains and checked the expression of pWOX8gΔ:NLS-vYFP3/pWOX2:DsRed2 in zygote and proembryo , we found that in about half seeds ( n = 142 ) , the expression pattern of WOX2 and WOX8 is consistent with the signal in zar1-1 ( S2A–S2D Fig ) , however , the other half seeds are similar to the wild type . This analysis showed that the pWOX8gΔ:NLS-vYFP3 and pWOX2:DsRed2 markers are mis-expressed in arrested zygotes and daughter cells from the symmetrical division in zar1-1 , indicating the zar1-1 mutation impairs zygote and the apical/basal cell fate . The ectopic expression of the pWOX8gΔ:NLS-vYFP3 marker in the apical lineage during early embryogenesis in zar1-2 suggests that the zar1-2 mutation affects the apical and basal cell lineages . Similarly , we checked the expression of pWRKY2:NLS-vYFP3 [16] . The YFP signal is found in zygote , the apical cell and the basal cell in wild-type ovules ( Fig 3O and 3P ) ; but no signal is detected in these cells in zar1-1 arrested ovules ( Fig 3Q and 3R ) ; conversely , expression of WRKY2 is detected ectopically in endosperm besides zygote and its daughter cells in zar1-2 ovules ( Fig 3S and 3T ) . This shows that WRKY2 expression is also impaired in zar1 mutants . Taking together , we concluded that the mutations of ZAR1 have an effect on the cell fate specification of both the apical and basal cell lineages during early embryogenesis in Arabidopsis . Molecular analysis indicates that zar1-1+/- contains a single Ds insertion which causes a 377 bp deletion from +1516 bp in the intron to +1893 bp in the second exon of At2G01210 . And in the two weak alleles , zar1-2 and zar1-3 , which contain a Ds insertion and a T-DNA , respectively , the extracellular region of At2G01210 is interrupted in both zar1-2 and zar1-3 ( Fig 4A and S3 Fig ) . To investigate whether At2G01210 is the ZAR1 gene , we introduced a 3 . 5 kb genomic DNA fragment spanning the locus into zar1-1+/- mutant plants . The transgene decreased the seed abortion rate from 26 . 3% ( n = 962 ) in zar1-1+/- to 19 . 8% ( n = 951 ) to 8 . 4% ( n = 965 ) in different T1 transgenic lines ( S4A Fig ) . Statistical analysis showed the abortion rate of ovules is decreased significantly in transgenic plants compared to that of zar1-1+/- . These demonstrate that the transgene can rescue zar1-1+/- phenotype . We also introduced this 3 . 5 kb ZAR1 genomic fragment into zar1-2 mutant and checked the phenotype of transgenic plant . It shows that the selected ZAR1 genomic fragment can rescue the zar1-2 phenotype and the zygote division in zar1-2 Is very similar to the wild-type plants . Interestingly , the wild-type plants transformed with pzar1:ZAR1ΔK-GFP , in which the kinase domain of ZAR1 was replaced by GFP to mimic the Ds insertion in zar1-1 , reduced the seed set from 95 . 81% ( n = 682 ) to 76 . 03–93 . 04% ( n > 630 ) in different transgenic lines . The lowest seed set is close to zar1-1+/- ( 73 . 22% ) ( n = 785 ) . Quantitative RT-PCR analysis indicates that the mRNA level of the ZAR1ΔK-GFP correlates tightly with the seed set reduction ( S4B Fig ) . Moreover , the extra-long zygote or daughter cells and approximately symmetric division of zygote seen in zar1-1+/- were also observed in pzar1: ZAR1ΔK-GFP plants ( S5 Fig ) . Together , these data confirmed that the At2G01210 is indeed the ZAR1 gene . The above data also suggest that the phenotypic difference may be resulted from different nature of the alleles . To investigate this possibility , we first checked if the truncated mRNA is produced . Indeed , there is a truncated transcript of about 800 nt in zar1-1+/- plants besides the 1500 nt wild-type mRNA ( S6A and S6B Fig ) . Furthermore , the presence of a 42 KD truncated protein in zar1-1+/- , but not in zar1-2 , was confirmed with Western analysis using ZAR1 antibody ( S6C Fig ) . These data indicate that zar1-2 is a null mutation and the sterile phenotype in zar1-1+/- is most likely caused by the truncated protein . ZAR1 is a protein of 716 amino acids with an N-terminal signal peptide , sequentially followed by seven LRR repeats , a transmembrane domain ( TM ) , and a C-terminal serine/threonine kinase domain which consists of eleven subdomains ( VI-XI ) ( Fig 4A and S3 Fig ) . There is a putative CaM-binding motif ( CaMBD ) downstream the TM , and a putative Q ( D/E ) RQQ-type Gβ-binding motif [22] between subdomain VII and VIII . ZAR1 is grouped to the Type III LRR-RLK subfamily [23] . The Ds is inserted into the subdomain V in zar1-1 , and a Ds or T-DNA is inserted in the region between LRR7 and TM domain in zar1-2 , or zar1-3 , respectively ( Fig 4A and S3 Fig ) . To investigate ZAR1 subcellular localization , CaMV 35S promoter driving ZAR1-GFP fusion gene was constructed and introduced into Arabidopsis . Confocal microscopy on transgenic roots showed that ZAR1-GFP fusion protein is co-localized with membrane dye FM4-64 in plasma membrane ( Fig 4B and 4C ) . This indicates that ZAR1 is a plasma membrane receptor-like protein kinase . To investigate the expression pattern of ZAR1 , pZAR1:NLS-YFP fusion was made and introduced into Arabidopsis . We found that ZAR1 is detected at the micropylar nuclei of the embryo sac at eight-nucleate stage ( FG5 stage ) before cellularization , and no YFP signal is observed in chalazal nuclei ( Fig 4D ) . After cellularization , the YFP signal is specifically detected in central cell and synergids in mature embryo sac ( Fig 4E ) with pZAR1:NLS-YFP transgenic plants . After fertilization , the YFP signal is detected specifically in zygote , endosperm precursor cell , and later in endosperm ( Fig 4F and 4G ) . In addition , YFP signal is also present in sperm and the vegetative cells in both pollen and pollen tube ( Fig 4H and 4I ) . Since it is very difficult in our hands to observe YFP signal in mature egg cell , we conducted RNA in situ hybridization . The result indicated that ZAR1 is expressed at high level in the egg cell ( Fig 4J ) , the central cell ( Fig 4K ) , the endosperm , and at low level in the zygote ( Fig 4L ) , While no signal was detected in the control hybridized with the sense RNA probe ( Fig 4M ) . Together , these indicated that ZAR1 is specifically expressed in the mature gametophytic cells and the product of double fertilization . To investigate whether ZAR1 interacts with CaM as suggested by the putative predicated CaM-binding motif . First , we performed protein pull-down experiment with proteins expressed in Escherichia coli . Constructs producing His-tagged ZAR1 kinase domain ( shorted as “His-kinase” in following text and the figures ) , His-kinaseK534>A with Lys534 to Ala534 mutation , and His-kinaseΔCaMBD ( kinase domain with CaMBD deletion ) , and GST-tagged CaMs , were made and expressed in bacteria . The pull-down results showed that His-kinase and His-kinaseK534>A , but not His-kinaseΔCaMBD , can be pulled down by either GST-CaM1 or GST-CaM8 ( Fig 5A and S7 Fig ) . The interaction between ZAR1 and CaM1 was independent of the concentration of Calcium , but the interaction was enhanced when the concentration of Calcium increased from 2 to 10 μmol/L ( S7 Fig ) . It shows that the ZAR1 kinase domain can physically interact with CaM1 and CaM8 via its putative CaM-binding domain in vitro , while the kinase catalytic activity is not essential for the interaction . To further confirm the above in vitro interaction between ZAR1 and CaM1 in plant cells , bimolecular fluorescent complementation ( BiFC ) and co-immunoprecipitation ( Co-IP ) assay were carried out . In BiFC experiment , Arabidopsis leaf protoplast was co-transformed with p35S:CaM1-nYFP paired with different ZAR1 constructs; meanwhile , p35S:PMS1-nYFP was introduced as a negative control . YFP fluorescence is only detected in the plasma membrane in cells co-expressing p35S:ZAR1-cYFP and p35S:CaM1-nYFP ( Fig 5B ) , indicating that the interaction between ZAR1 and CaM1 occurs at plasma membrane . These interactions were also confirmed by Co-IP assay with proteins from transformed Arabidopsis leaf protoplasts , namely the HA-CaM1 co-immunoprecipitates with either FLAG-ZAR or FLAG-ZAR K534>A , but not with FLAG-ZAR1ΔCaMBD ( Fig 5C ) . This strongly indicates that CaM1 interacts with ZAR1 via the CaMBD domain in vivo . Finally , we checked whether ZAR1 and CaM1 are co-localized in cells as CaMs are cytosolic proteins . When we observed Arabidopsis protoplasts transformed with pZAR1:ZAR1-GFP and p35S:CaM1-RFP , we found the co-localization of ZAR1-GFP and CaM1-RFP at the plasma membrane , in addition to the localization of ZAR1-GFP at plasma membrane , and most CaM1-RFP is in cytoplasm ( Fig 5D ) . Together , these data showed that ZAR1 interacts with CaM1 via its CaM-binding domain at the plasma membrane in planta . The presence of a putative glutamine-rich Gβ-binding site ( GβBS ) in ZAR1 suggests that ZAR1 may interact with Gβ although such interaction has not been demonstrated before . We performed pull-down , BiFC and Co-IP experiments to check the interaction between ZAR1 and Gβ . Both His-kinase and His-kinaseK534>A proteins , but not His-kinaseΔGβBS ( kinase domain with GβBS deletion ) , were pulled down with MBP-AGB1 ( Fig 6A ) . Consistently , such interaction was confirmed by BiFC ( Fig 6B ) and Co-IP experiments ( Fig 6C ) . YFP fluorescence was clearly detected in plasma membrane and also in endomembrane in protoplasts co-expressing ZAR1-cYFP and AGB1-nYFP . Interestingly , the YFP fluorescence is unevenly distributed on the plasma membrane ( Fig 6B ) . HA-AGB1 protein co-precipitated with FLAG-ZAR1 and FLAG-ZARK534>A , but not with FLAG-ZAR1ΔGβBS ( Fig 6C ) . These data showed that ZAR1 interacts with AGB1 via its Gβ-binding motif in vivo and in vitro . As shown above , ZAR1 is a plasma membrane protein , and AGB1 is a protein localized on plasma membrane and in nucleus [24] , we questioned whether they can meet spatially in cells . Surprisingly , ZAR1-GFP is co-localized with AGB1-mCherry in foci on the plasma membrane and also in nucleus in protoplasts co-transformed with p35S:ZAR1-GFP and p35S:AGB1-mCherry ( Fig 6D ) . Taken together , our data show that ZAR1 interacts with AGB1 indeed , and they may function together on the membrane or in the nucleus as well . The interaction between ZAR1 and AGB1 prompted us to check if zygote division is also impaired in agb1-2 . Similar to zar1-2 , abg1-2 is a fertile mutant showing endurable symmetric division of the zygote . We compared the length of apical cell and basal cell which could be measured under microscope . The length of the wild-type apical and basal cells are 16 . 45±1 . 84 μm ( n > 30 ) , and 58 . 90±5 . 12 μm ( n > 30 ) respectively ( Fig 2A , 2F and 2G ) . However , zar1-1 has longer apical cell ( 27 . 27±2 . 53 μm , n > 30 ) and basal cells ( 78 . 96±4 . 85 μm , n > 30 ) ( Fig 2B , 2F and 2G ) , zar1-2 shows a similar apical cell , but a significantly shortened basal cell ( Fig 2C , 2F and 2G ) , while the agb1-2 apical cell is slightly shortened ( Fig 2D and 2F ) but the basal cell is dramatically decreased to 44 . 84±3 . 42 μm ( n > 30 ) ( Fig 2D and 2G ) compared with the wild-type . It suggests that the zygote asymmetric division in zar1-1 , zar1-2 , and agb1-2 ovules is disrupted . Interestingly , in zar1-2 agb1-2 double mutants , the length of basal cells is 78 . 55±4 . 83 μm ( n > 30 ) ( Fig 2E and 2G ) , very approximate to that of zar1-1 basal cells , which is attributed to the non-interaction between AGB1 and ZAR1ΔK in zar1-1 ( the AGB1-binding site is deleted in the zar1-1 ) . It implies that the asymmetric division of zygote and the development of basal cell are impaired because of the mutation of ZAR1 and AGB1 . Moreover , expression pattern of cell linage-specific markers in agb1-2 is very similar to that in zar1-2 ( Fig 3I–3N ) , for example , the basal cell-specific marker , pWOX8gΔ:NLS-vYFP3 , is also ectopically expressed in embryo proper in agb1-2 ovules ( Fig 7F–7J ) . Likewise , expression of zygote and basal cell marker , pWOX9:NLS-vYFP3 is extended to the apical cell lineage in agb1-2 ovules ( Fig 7P ) compared with the wild-type ( Fig 7O ) , and pWRKY2:NLS-vYFP3 is detected unexpectedly in endosperm besides zygote and its daughter cells in agb1-2 ovules ( Fig 7K–7N ) too . It suggests that ZAR1 and AGB1 play roles in the asymmetric division of zygote and specification of apical and basal cell lineages , and they possibly function through their interaction in the same , or overlapping pathway during early embryogenesis . To investigate whether ZAR1 , CaM1 and AGB1 form complex since ZAR1 interacts with both CaM1 and AGB1 , an in vitro pull-down experiment was conducted . As shown in Fig 8A , GST-CaM1 pulls down both His-kinase and MBP-AGB1 . Similarly , MBP-AGB1 can also pull down both GST-CaM1 and His-kinase ( Fig 8A ) . This indicates that ZAR1 , CaM1 and AGB1 are able to form a complex in vitro . To check if ZAR1 , CaM1 and AGB1 can form complex in vivo , we co-express pZAR1-cYFP , pAGB1-nYFP and pCaM1-RFP in Arabidopsis protoplast . It is expected that the interaction between ZAR1 and AGB1 should restore the YFP that will be co-localized with CaM1-RFP if they form a complex . Indeed , the YFP fluorescence was restored and co-localized with CaM1-RFP on plasma membrane ( Fig 8B ) . These further indicate that ZAR1 , AGB1 and CaM1 can meet spatially and form complex in plant cells . Consistently , CaM1 and AGB1 co-precipitates with FLAG-ZAR1 in FLAG-ZAR1 transgenic plants , but not with LTi6b ( as a control ) ( Fig 8C ) . These data demonstrate that ZAR1 , CaM1 and AGB1 form complex in planta . To further investigate whether the binding of CaM1 and AGB1 affects ZAR1 kinase activity , in vitro pull-down experiment was performed . As known , calf intestinal alkaline phosphatase ( CIP ) catalyzes the removal of phosphate groups from the Ser , Thr and Tyr of phosphorylated protein non-specifically . Lys534 is the putative kinase catalytic site in ZAR1 kinase , and it is supposed to be responsible to transfer the phosphate group to Ser , Thr and Tyr of substrate . Indeed , His-kinaseK534>A loses the kinase catalytic activity . In the absence of CaM1 or AGB1 , there is no difference with the migration rate between CIP phosphatase-treated or untreated His-ZAR1kinase and His-kinaseK534>A ( Fig 8D and 8E ) . However , in the presence of CaM1 or AGB1 , the His-kinase is activated and moves slower than CIP-treated His-kinase and His-kinaseK534>A ( Fig 8D and 8E , stars ) , while CaM1 or AGB1 itself was not phosphorylated ( Fig 8D and 8E , arrowheads ) . It indicates that ZAR1’s auto-phosphorylation is promoted by the binding of CaM1 or AGB1 , but the binding is independent on the ZAR1 kinase activity , as CaM1 or AGB1 interacts with His-kinaseK534>A . These data , together , strongly suggest that ZAR1 , AGB1 and CaM1 form complex and the ZAR1 kinase activity is activated by the binding of CaM1 and/or AGB1 .
Calcium signaling has long been implicated to participate in fertilization , zygote activation , and establishment of cell polarity in animals and plants [25–27] . In mammals , Ca2+ oscillation in egg triggered by sperm is essential for egg activation and the first cleavage [28] . In plants , the Ca2+ signal seen after gamete cytoplasmic fusion in the fertilized egg may also play a role during early embryogenesis [25 , 29] . In brown algae , zygote polarity is dependent on local increase of Ca2+ concentration in pre-S phase [30] , and the first asymmetric division of the zygote is determined by the asymmetric distribution of Ca2+ and plasma membrane complexes [31] . Ca2+ signals are perceived and decoded by diverse Ca2+ sensors , such as CaM , Ca2+-dependent protein kinase ( CDPK ) , and Ca2+/CaM-dependent protein kinase ( CaMK ) , which play crucial roles in a variety of signaling networks by controlling the activities of a battery of target proteins [32–34] . Here , we demonstrate that ZAR1 interacts with CaM and its kinase activity is activated upon CaM binding . The ratio of zygote symmetric division is 22 . 6% ( n = 138 ) in pZAR1:ZAR1ΔCaMBD transgenic plants . It indicates that the interaction between ZAR1 and CaM1 is required for the asymmetric division of zygote . It is plausible to speculate that ZAR1 receptor kinase , via its interaction with CaM , is able to respond to Ca2+ increase triggered by fertilization in the zygote , and translate the signal into developmental cues during zygotic division in Arabidopsis . Therefore , it would be interesting to measure the [Ca2+]cyt changes of the egg during fertilization and to investigate the possible links between the pathway that ZAR1 mediated and the Ca2+ signaling cascade during zygote development . Heterotrimeric G proteins , another key molecular switch in cell signaling , are also involved in asymmetric divisions in animals . G proteins control asymmetric cell division by regulating localization of polarity determinants and spindle orientation [35] . Mutations in Gβγ impair asymmetric division in Drosophila [36] and C . elegans [37] . In yeasts , Gβγ interacts directly with PAK1 kinase to activate CDC42 , a critical regulator of cell polarity [38] . Unlike in animals , there are only one canonical α , one β and two γ subunits in Arabidopsis [39] , and the GEF is lost in plants during evolution [40] . In plants , heterotrimeric G proteins are mainly involved in immunity , stress response and nitrogen utilization [41–44] , their roles in development are interesting . On one hand , they seem dispensable for development since Arabidopsis plants lacking Gα or Gβ develop normally and grow well . On the other hand , they play critical roles in stress response and hormone regulation pathway . For example , Gα/COMPACT PLANT2 ( CT2 ) interacts with CLV2 to control the shoot size in maize [45]; Gβ is involved in root cell proliferation [46 , 47] and the plant-specific non-canonical Gγ ( AGG3 ) is required for organ size control in Arabidopsis [48] . In rice , Gα ( D1/RGA1 ) is involved in rice growth by modulating gibberellic acid and brassinosteroid hormone signaling [49 , 50]; COLD1 is a regulator of G protein signaling , it interacts with Gα to activate the Ca2+ channel for temperature sensing [44] . According to our data , the asymmetric division of zygote and specification of its daughter cells are impaired because of mutations of ZAR1 and AGB1 . It is suggested that the disruption of interaction between ZAR1 and AGB1 results in the abnormality of zygote division and the compensatory growth of basal cell . Consistently , the expression pattern of WOX8 and WRKY2 is impaired similarly in both zar1-2 and agb1-2 mutants . All these suggest that G proteins might function as modulators in zygote division and the determinants for cell fate of apical and basal cells . Moreover , when we transformed pZAR1:ZAR1ΔGβBS into the wild-type , the symmetric division of zygote was found in T1 generation ( 15 . 8% , n = 123 ) . These support the idea that ZAR1 and AGB1 act together to control the asymmetric division of zygote and cell fate determination of its daughter cells . As discussed above , ZAR1 interacts directly with both CaM and Gβ , and as a membrane receptor kinase , ZAR1 is able to integrate extracellular signal mediating the intracellular Ca2+ and heterotrimeric G protein pathways that control zygotic division . There are many details remain to be elucidated during this fundamental process , for example , putative ligand for ZAR1 needs to be identified . Recent studies show that secreted cysteine-rich peptides act as ligand for several LRR-RLKs to regulate asymmetric division during stomata development and cell-cell signaling in fertilization and zygote development [51–57] . It would be worthy to investigate whether such peptides are also ligands for ZAR1 during fertilization . Our data also indicate that ZAR1 plays a redundant role in controlling zygotic cell division . A stronger phenotype is seen in zar1-1+/- than in zar1-2 null allele although asymmetric division of the zygote is disrupted in both zar1-1+/- and zar1-2 plants . The strong phenotype in zar1-1+/- is most likely caused by the truncated ZAR1 protein . Moreover , in zar1-1+/- , the truncated ZAR1 protein likely functions in a dominant-negative and dosage-dependent manner . Indeed , the phenotype depends largely on the expression level of the pZAR1:ZAR1ΔK-GFP transgene ( S4B Fig ) . On the other hand , the null allele zar1-2 and zar1-3 plants are completely fertile although they show a weak zygotic division phenotype . Although both zar1-1 and zar1-2 affect zygote development , their phenotypes differ . This raises the question of how the mutations in the zar1-1 and zar1-2 alleles affect the wild type molecular function of ZAR1 . Currently , this cannot be solved easily , more in vivo studies on ZAR1 protein and its interacting partners may provide clues . Dominant negative mutations are common for LRR-RLKs and invaluable for deciphering their functions in plants . For example , all the intermediate and strong alleles of clavata1 contain missense mutations in the LRR domain and are dominant-negative , the null alleles show very weak phenotype [58] . Another example is the tomato protein Cf-9 . Dominant-negative interference by truncated Cf-9 proteins results in leaf chlorosis accompanying leaf necrosis from the base towards the apex in tomato [59 , 60] . Similarly , the disease resistant LRR-RLK Xa-21 is capable of functioning without a kinase domain [61] . In ERECTA:ERECTAΔKinase transgenic plants , the endogenous ERECTA signal is interrupted by highly stable truncated protein by dominant-negative mechanism in a dosage-dependent manner [62] . Our data show that ZAR1 and ZAR1Δkinase co-exist in zar1-1 ( S6C Fig ) . These imply that the strong phenotype in zar1-1+/- plants is most likely caused by interference of ZAR1Δkinase with endogenous ZAR1 or ZAR1-interacting partners . The weak , endurable zygotic block seen in zar1-2 homozygous plants suggests that additional proteins may take over ZAR1 function when ZAR1 is absent . Such functional redundancy is common for LRR-RLKs since the RLK/Pelle kinase family have been greatly expanded to 374–2205 members in flowering plants , of which the LRR-RLKs represent the largest subfamily with over 223 members in Arabidopsis [23 , 63] . Interestingly , there are six LRR-RLK genes that show alternative splicing for proteins of different version , including some truncated proteins [23] . This suggests that the truncated products may function as intrinsic regulators of the RLK signaling system in plants . It has been proposed that multiple receptors may act in a functionally related manner [58 , 63 , 64] . It is possible that ZAR1 is functionally redundant with related RLKs since null alleles are fertile . Therefore , it would be interesting to identify the ligand and the receptor kinases that functionally overlapped with ZAR1 in regulating the asymmetric division of zygote in Arabidopsis . Our data propose that ZAR1 plays role in the cell fate specification of basal cell and apical cell . Asymmetric cell divisions generate cells with different fate . In most angiosperms , such as Arabidopsis , zygote undergoes asymmetric division and produces a small apical cell and a large basal cell [20 , 21] . The apical and basal cell lineages have different cell fate manifested by differential expression of cell-specific markers [15 , 17] . In zar1-1+/- and zar1-2 mutants , the elongation of the zygote is very similar to that of the wild-type , but the first asymmetric division is either abolished or transited into an approximately symmetric division , configuring with the shortened basal cell . Interestingly , the basal cell-specific marker pWOX8gΔ:NLS-vYFP3 is not expressed in the elongated or symmetrically divided zygote in zar1-1+/- mutant . While in young proembryos that by-passed the arrest in zar1-2 and agb1-2 , pWOX8gΔ:NLS-vYFP3 is expressed in both the basal and apical cell lineages , indicating the mis-specification of cell fate , and the cell specification of basal cell and apical cell is disrupted . Both in zar1-2 and agb1-2 , pWRKY2:NLS-vYFP3 is much more accumulated in endosperm . These suggest that ZAR1 plays a role in defining the expression pattern of WOX8 and WRKY2 , and so far cell fate differentiation . There are over 220 LRR-RLK genes in Arabidopsis genome , 20 LRR-RLKs contain a putative CaM-binding domain and 19 contain a putative Gβ-binding motif . 15 LRR-RLKs have been experimentally proved to interact with CaM , but none of the 19 LRR-RLKs have been shown to interact with Gβ although AT5G67280 has been implicated to interact with Gβ protein genetically [42] . Our data show that ZAR1 interacts with Gβ subunit of heterotrimeric G protein , and support the bold idea that the LRR-RLKs of plants fulfill equivalent roles to GPCRs in fungi and animals in cell-cell signaling [42 , 45] . Taken together , we identified a membrane receptor kinase ZAR1 that is required for zygote asymmetric division and the cell fate of its daughter cells . Through its interaction with CaM and Gβ , ZAR1 plays a key role in integrating the intracellular Ca2+ and heterotrimeric G protein signaling with extracellular cues during early embryogenesis in Arabidopsis . Preliminarily , we found some clues that ZAR1 might participate in SSP/YODA MAPK kinase pathway ( S8 Fig ) . It is well worthy to investigate the link between ZAR1/AGB1 pathway and YODA MAPK kinase cascade signaling .
Plants were grown in greenhouse at 22°C with 50–70% humidity and under the light cycle of 16 hrs daylight/8 hrs darkness . zar1-1+/- and zar1-2 are Ds gene trap lines generated in Arabidopsis thaliana ecotype Landsberg erecta [19] . zar1-3 ( SALK_021338 ) was obtained from ABRC in ecotype A . thaliana Col-0 . agb1-2 [65] was a gift from Dr . Jirong Huang at Shanghai Institute of Plant Physiology and Ecology , CAS . The seeds of EGFP-LTi6b transgenic plant [66 , 67] were from Prof . Yurong Bi at Lanzhou University . Plant phenotype was performed as described previously [67] . To observe the embryo phenotype , seeds from zar1-1+/- siliques were mounted in Herr’s solution before observation with a Zeiss Axioskop [67] . Reciprocal crosses between the wild-type and zar1-1+/- plants were performed as reported before [68] . Total RNA was extracted with TRIzol reagent ( Invitrogen ) from different tissues of wild-type plants according to the manufacturer’s instruction . The single-stranded cDNA was transcribed by superscriptIII ( Invitrogen ) . Real time PCR was carried out using primers listed in S1 Table , and ACTIN2/8 was used as internal control . To obtain the flanking sequence of Ds insert , TAIL-PCR was performed as previous report [69] and primers used are listed in S1 Table . For construction of plasmids pZAR:ZAR1 and pZAR1:ZAR1Δk-GFP-nos , the corresponding fragments were amplified and subcloned in pCAMBIA1300 ( Clontech ) or pCAMBIA2300 ( Clontech ) with primers listed in S1 Table . Plasmids p35S:ZAR1-GFP and p35S:ZAR1-FLAG were provided by Dr . Jia Li [23] . Plasmids pZAR1:ZAR1-GFP , p35S:CaM1-RFP and p35S:AGB1-mCherry which were used for co-localization in protoplast were constructed in pBlue-SK ( Stratagene ) and pWEN57-RFP with primers shown in S1 Table as described above . Constructs of p35S:ZAR1-cYFP , p35S:CaM1-nYFP and p35S:AGB1-nYFP for BiFC , were performed as described previously [67] . The variants of point mutations were introduced into p35S:ZAR1-cYFP by site-directed mutagenesis . PMS1 ( AT4G02460 ) encoding a nucleus protein was cloned to construct as a negative control in BiFC assays . For pull-down experiments , the fragment of ZAR1-kinase was amplified with primers listed in S1 Table and cloned into pET28a ( + ) . The ZAR1-kinase mutations were introduced into the ZAR1kinase by site-directed mutagenesis . Similarly , the fragments of AGB1 was amplified and cloned into pMAL-C-2X; the full length of CaM1 and CaM8 were amplified and cloned into pGEX-4T-2 . For Co-IP experiments , full length of ZAR1 was amplified with primers listed in S1 Table , and cloned into pUC19-35S:3FLAG , the variants of ZAR1 were generated by site-directed mutagenesis . The fragments of CaM1 and AGB1 were amplified and cloned into pUC19-35S:HA as above [70] . Protoplasts from Arabidopsis leaves were prepared and transfected with 10 μg purified DNA as described previously [70] . DNA was extracted with CsCl gradient centrifugation or with EndoFree Plasmid Maxi Kit ( QIAGEN ) . Experiments for co-localization of GFP and RFP or mCherry were performed and repeated at least three times . BiFC assays were performed and repeated according to Walter [71] . FM4-64 staining was performed as described previously [72 , 73] . Arabidopsis protoplast and root cells were stained with 10 μmol/L FM4-64 and image was completed within 10 min of staining . Images were acquired with confocal LSM510 or LSM780 ( Zeiss ) . Fluorescent signals were detected with an Argon 2 laser for GFP , YFP ( excitation , 488 nm or 514 nm; emission , BP500-530 nm or BP500-560 nm emission filter ) , and chloroplast auto-fluorescence and excitation , 488nm; emission , LP 615 filter ) , and with He-Ne Laser ( excitation , 543 nm; emission BP560-600 or LP600 emission filter ) for RFP and mCherry , and He-Ne Laser ( excitation , 561nm; emission , BP570-630 emission filter ) for FM4-64 . The pull-down assay was carried out according to Xiang with slight modification [70] . Overnight culture of E . coli with different constructs was transferred to fresh medium with 1/50 dilution , then continued with 3 hrs incubation on shaker ( 250 rpm , 37°C ) . When the OD600 reaches 0 . 6 , the culture was moved to 22°C , 250 rpm for 30 min , and then IPTG was added to the final concentration of 0 . 5 mmol/L . The induction was kept for another 6 hrs at 22°C , 250 rpm . The bacterium cells were collected and resuspended with 10 ml Tris buffer ( 25 mmol/L Tris pH 7 . 5 , 50 mmol/L NaCl , 3 mmol/L DTT , 1 mmol/L PMSF , protease inhibitor Cocktails ) , and were debrised by ultrasonic . The samples were spun at 4°C , 13 , 000 rpm for 30 min . The supernatant of 100 μl was kept for input , and 50 μl of the rest supernatant was incubated with glutathione agarose ( GE Healthcare ) ( GST-tagged beads ) , in 360° shaker at 4°C for overnight . The beads were spun at 500 g for 2 min , and washed for five times with Tris buffer ( 25 mmol/L Tris pH 7 . 5 , 50 mmol/L NaCl , 3 mmol/L DTT , 1 mmol/L PMSF , protease inhibitors Cocktails , 1% Triton X-100 , 0 . 1% SDS ) . SDS loading buffer was added before Western detection . For protoplast Co-IP assay , the leaf protoplasts were co-transfected with 10 μg plasmid DNA of each construct , pAGB1-HA/pZAR1-FLAG , pAGB1-HA/pZAR1K534>A-FLAG , and pAGB1-HA/pZAR1ΔGβBS-FLAG , with aid of PEG/Calcium , and then incubated at 22°C , 45 rpm , for overnight . The protoplast protein was extracted as described previously [71 , 72] with native extraction buffer ( 20 mmol/L HEPES , pH 7 . 5 , 40 mmol/L KCl , 250 mmol/L glucose , 5 mmol/L MgCl2 , 1mmol/L PMSF , protease inhibitors Cocktails ) . The protoplast were harvested by soft spin at 100 g , then resuspended with 500 μl native extraction buffer , and vortexed for 30 s to mix . The samples were spun at 6 , 000 g for 15 min at 4°C , then supernatants were collected and the pellet were resuspended with 100 μl native extract buffer . The samples were subsequently ultrasonicated for 10 second , and Triton X-100 was added to the final concentration 1% . The samples were spun at 100 , 000 g for 10 min at 4°C , and the supernatants were collected , and diluted with native extraction buffer [70–72] . 50 μl supernatant was reserved as control for total protein . The rest was mixed with prepared anti-FLAG-M2 gels ( Sigma ) by rotating at 4°C for 1 hr . The samples were spun down at 1000 rpm for 3 min and then washed five times with washing buffer ( 20 mmol/L HEPES , pH 7 . 5 , 40 mmol/L KCl , 250 mmol/L glucose , 5 mmol/L MgCl2 , 0 . 1% Triton X-100 ) . SDS loading buffer was added to samples before detection by Western blot [68] . Co-IP assay was performed with FLAG-ZAR1 , and EGFP-LTi6b transgenic plants ( used as control for membrane protein ) [65] . The inflorescences and siliques 1–2 DAP , or seedlings of 10 DAG ( day after germination ) from transgenic plants were collected , and Co-IP assays were carried out as described previously [72 , 73] . Anti-FLAG-M2 gel was incubated with samples for 30 min , and AGB1 and CaM1 proteins were checked with Anti-AGB1 antibody ( Sigma ) and anti-CaM antibody ( Sigma ) , respectively . Sample preparation and sectioning for in situ hybridization were according to previous reports [74 , 75] with minor modification . The plants were emasculated or pollinated manually , and ovules 24 hrs after emasculation or 12–24 hrs after pollination were collected and fixed before sectioning . The ZAR1 fragment spans the nucleotide sequence from +259 nt to +1160 nt was cloned into pT-GEM , and the linearized plasmid was transcribed in vitro by T7 RNA polymerase or SP6 RNA polymerase ( Roche ) for antisense or sense probes . Sequence data from this article can be found in the GenBank/EMBL or Arabidopsis Genome Initiative database under the following accession numbers: ZAR1 ( AT2G01210 ) , CaM1 ( AT5G37780 ) , CaM8 ( AT4G14640 ) , AGB1 ( AT4G34460 ) , PMS1 ( AT4G02460 ) .
|
Flowering plants are featured as double fertilization , a process that the egg cell and the central cell of embryo sac fuse with a sperm and give rise to a diploid zygote and a triploid primary endosperm cell , respectively . The zygote develops into embryo after cell division and differentiation , and starts a new trip of next generation . Meanwhile , the primary endosperm cell proceeds nuclear division to generate a syncytium and develops into endosperm after cellularization . Embryo development initiates from asymmetric division of zygote . A small apical cell and a long basal cell are produced after the first zygotic division , which establishes the pattern of an early embryo . To unveil the molecular mechanism controlling zygote asymmetric division , we screened our Ac/Ds insertion lines for mutations controlling early embryogenesis , one of the mutations zygotic arrest 1 ( zar1 ) was reported here . In zar1 , zygote was either arrested after elongation or displayed symmetric division . The mutation also had a slight impact on the apical- and basal-cell fates manifested by the mis-expression of the cell-lineage specific markers . ZAR1 encodes a member of the RLK/Pelle kinase family , and interacts physically with Calmodulin and the heterotrimeric G protein Gβ , both in vitro and in vivo . These data suggest that ZAR1 might act as an integrator for intracellular Ca2+ and heterotrimeric G protein signaling with extracellular signals during early zygote development . Interestingly , complete loss of Gβ or ZAR1 function displayed very weak phenotype . This suggests that there might be genetic redundancy and plasticity during early embryogenesis . More studies are needed to dissect the complexity of early embryo development in plants .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"anatomy",
"keratinocytes",
"medicine",
"and",
"health",
"sciences",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"ovules",
"epithelial",
"cells",
"germ",
"cells",
"zygotes",
"endosperm",
"developmental",
"biology",
"luminescent",
"proteins",
"plant",
"science",
"yellow",
"fluorescent",
"protein",
"embryos",
"cellular",
"structures",
"and",
"organelles",
"basal",
"cells",
"embryology",
"animal",
"cells",
"proteins",
"biological",
"tissue",
"cell",
"membranes",
"biochemistry",
"cell",
"biology",
"ova",
"anatomy",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"fruit",
"and",
"seed",
"anatomy"
] |
2016
|
The Arabidopsis Receptor Kinase ZAR1 Is Required for Zygote Asymmetric Division and Its Daughter Cell Fate
|
Few models exist that accurately reproduce the complex rhythms of the thalamocortical system that are apparent in measured scalp EEG and at the same time , are suitable for large-scale simulations of brain activity . Here , we present a neural mass model of the thalamocortical system during natural non-REM sleep , which is able to generate fast sleep spindles ( 12–15 Hz ) , slow oscillations ( <1 Hz ) and K-complexes , as well as their distinct temporal relations , and response to auditory stimuli . We show that with the inclusion of detailed calcium currents , the thalamic neural mass model is able to generate different firing modes , and validate the model with EEG-data from a recent sleep study in humans , where closed-loop auditory stimulation was applied . The model output relates directly to the EEG , which makes it a useful basis to develop new stimulation protocols .
In the human electroencephalogram ( EEG ) the most prominent features of non-rapid eye movement ( NREM ) sleep are neocortical slow oscillations ( SO ) at ∼ 0 . 8 Hz [1] and fast thalamocortical spindles characterized by a waxing and waning waveform , with a frequency of 12 to 15 Hz [2] . The cellular basis of the slow oscillation is a widespread alternation of activity in neocortical networks between an active ( up ) state and a hyperpolarized silent ( down ) state [2–4] . Fast sleep spindles are generated by the interaction of inhibitory reticular thalamic ( RE ) and excitatory thalamocortical ( TC ) neurons [5] and preferentially occur during the up state of SOs [6] . Many studies indicate a functional role of slow wave sleep ( SWS ) in the formation of memory [7 , 8] . The synchronization of fast spindle activity to the depolarized up state is mediated by the thalamocortical loop . It appears to be critical for consolidation as it provides a window of opportunity and favorable conditions for plastic changes [9–11] . It has been shown that memory consolidation can be improved by transcranial electric and auditory stimulation [12 , 13] . Auditory stimulation in synchrony with the brain’s own rhythm turned out to be particularly effective . Detailed knowledge of how different stimulation modalities effect critical brain rhythms as well as the ramifications of potential stimulation parameters on their efficacy would enable an optimization of stimulation protocols and consequently an advantage for experiments in basic research and clinical applications . Mathematical models and computational approaches can yield meaningful insights into the underlying dynamics as well as provide predictions for further experiments . Multiple models based on networks of single cells have addressed brain activity during sleep in the cortex [14] , thalamus [15–17] , as well as the coupled thalamocortical system [18–21] . However , for investigations of macroscopic phenomena these models have disadvantages due to their complexity and computational load . Neural mass models have shown great success in elucidating the generation of brain rhythms and evoked responses of the awake brain [22–25] . They describe the dynamics of a large number of cells by the evolution of a single population average and provide an output which directly relates to EEG signals [26 , 27] . By still allowing an integration of physiologically motivated cell dynamics via intrinsic currents , neural mass models represent a compromise between a very detailed and an abstract modeling approach and provide insights into the dynamic repertoire of the neural populations via bifurcation analysis [28–30] . In previous work , we have shown that a cortical neural mass model equipped with an additive activity-dependent feedback current can generate a time series that closely resembles the EEG signal of sleep stages N2 and N3 , without spindles [30] . Here , we extend the cortical model by adding a thalamic module to incorporate spindle activity and investigate the underlying dynamics of the coupled system . We test the evocability of SOs and spindles by auditory stimulation during NREM sleep and validate our results with scalp EEG data from a recent sleep study in humans [13] . Our findings add further support to the dynamic mechanisms proposed in [30] .
The conductance based neural mass model employed here is derived from [31–34] . Instead of considering the evolution of high-dimensional states in a large ensemble of single neurons , the population activity can be approximated by the evolution of the mean membrane voltage of that population . The complex spiking dynamics is replaced by an empirical firing rate function Q k = Q k max 1 + exp ( - ( V k - θ ) / σ k ) , ( 1 ) with maximal firing rate Q k max , firing threshold θ and neural gain σk . It converts the average membrane voltage Vk of the population k to an output spike rate . Here , k ∈ {p , i , t , r} stands for cortical pyramidal , cortical interneuron , thalamic relay and thalamic reticular populations , respectively . The firing rate function often has a sigmoidal shape and can be interpreted as stemming from the fluctuations of neuronal states or a distribution of thresholds in the population [35] . The spike rate Qk′ of the presynaptic population k′ then elicits a postsynaptic response smk within the receiving population k . The strength of this response can be calculated by the convolution s m k = ∑ k ′ α m ⊗ N k k ′ Q k ′ , ( 2 ) of the incoming spike rate Qk′ , scaled with the averaged connectivity constant Nkk′ between the presynaptic population k′ and the postsynaptic population k and an alpha function α m = γ m 2 t exp ( - γ m t ) , ( 3 ) representing the average synaptic response to a single spike . Here , γm depicts the rate constant of the synaptic response , whereas m ∈ {e , g , r} denotes the type of synapse with e standing for excitatory AMPA and g , r for inhibitory GABA synapses in the cortex and thalamus , respectively . The evolution of the population membrane voltage Vk obeys τkV˙k=− ( Vk−ELk ) −wAMPAsek ( Vk−EAMPA ) −wGABAsgk ( Vk−EGABA ) , =−JL−JAMPA ( sek ) −JGABA ( sgk ) , ( 4 ) where w denotes a synaptic input rate that scales the input smk and E the corresponding Nernst potential . While we lean on the naming convention of Hodgkin-Huxley models to highlight the structural similarity , please note that the above quantities J and w have different units . For the sake of simplicity we have normalized the synaptic rates w to 1 , absorbing their numerical values into the connectivities Nkk′ . Intrinsic currents may additionally be included in the equation of the mean membrane voltage , given their time constant is large compared to the time constant of neuronal spiking [35 , 36] . To emphasize the connection to physiology and to better differentiate between the core neural mass model and the additional mechanisms we will denote them with I and introduce a capacitive coupling via to the neural mass via the membrane capacity Cm . The signal measured in the EEG stems mainly from the activity of pyramidal neurons [37] . We use the pyramidal membrane voltage as our model output , which is similar to other studies [38 , 39] . Populations comprising multiple clusters have been considered in [40] and lead to interesting effects . In order to keep the complexity of the model low we consider a single point source . Therefore , filtering effects by the skull/scalp can be approximated by a linear scaling and do not affect the interaction between thalamus and cortex . We use the neural mass model described in [30] , which captures essential features of the EEG during NREM sleep . In brief , it consists of an excitatory and an inhibitory neural mass , representing populations of pyramidal neurons ( p ) and interneurons ( i ) , respectively . The pyramidal population contains a slow , additive and activity-dependent firing rate adaptation , τ Na [ Na ˙ ] = α Na Q p ( V p ) - Na pump ( [ Na ] ) , I KNa = g ¯ KNa w k ( [ Na ] ) ( V p - E K ) , ( 5 ) which is believed to be the main driver for the transition to the silent ( down ) state , while the active ( up ) state is maintained by synaptic transmission [3 , 14 , 41] . The membrane potentials evolve according to τ p V ˙ p = - J L p - J AMPA ( s e p ) - J GABA ( s g p ) - C m - 1 τ p I KNa , τ i V ˙ i = - J L i - J AMPA ( s e i ) - J GABA ( s g i ) . ( 6 ) and are linked by the AMPA and GABAergic synaptic inputs . The transition from N2 to N3 is achieved by changing the neural gain , σp , and the strength of the adaptation , given by g ¯ KNa . Both parameters are known to be influenced by acetylcholine , serotonin , norepinephrine and dopamine [42–52] , whose levels change throughout the sleep-wake cycle [53] . For consistency , we maintained the original model , adjusting only the excitatory connection strengths to compensate for additional input from the thalamic module . The thalamic module comprises similarly an excitatory and an inhibitory neural mass , representing a thalamocortical ( t ) and the reticular ( r ) nucleus . As in the cortical module module they are coupled via AMPA and GABA synapses but have different synaptic time constants and only the RE population possesses a self-connection . Both populations are equipped with additional currents . The inclusion of those currents within the thalamic submodule is necessary because spindle oscillations require rebound bursts . In classical neural mass models , this kind of bursting is not possible due to the monotonic firing rate function and demands the inclusion of additional mechanisms . The same argument was used in a previous neural mass model of spindle activity [54] and a thalamocortical neural mass model of epileptic activity [55] . Finally , in [40] the authors arrive at a HH-type extension of their population model of thalamic burst activity , which has been derived from integrate-and-fire-or-burst neurons . The potassium leak current is given by I LK = g ¯ LK ( V k - E K ) , ( 7 ) as well as T-type calcium currents I T = g ¯ T m ∞ 2 h ( V k - E Ca ) , ( 8 ) which deinactivate upon hyperpolarization . They are essential for the generation of low-threshold spikes ( LTSs ) and rebound bursts . We use the description of IT given in [56] for the RE and the one in [16] for the TC population . The TC population further includes the anomalous rectifier current I h = g ¯ h ( m h 1 + g i n c m h 2 ) ( V t - E h ) , ( 9 ) responsible for the waxing and waning structure of spindle oscillations in the isolated thalamus [15] . Other currents , such as the calcium-dependent potassium currents IKCa and ICAN , are also known to play a role in spindle oscillations , but are omitted for simplicity . The thalamic module is summarized by τ t V ˙ t = - J L t - J AMPA ( s e t ) - J GABA ( s r t ) - C m - 1 τ t ( I LK t - I T t - I h ) , τ r V ˙ r = - J L r - J AMPA ( s e r ) - J GABA ( s r r ) - C m - 1 τ r ( I LK r - I T r ) . ( 10 ) Parameter settings for the currents are identical to [57] , with the exception of the deactivation function h ∞ t of the thalamic relay population , that is shifted towards more depolarized membrane voltages . The model consists of one thalamic and one cortical module . We assume the long range afferents from the cortical pyramidal population project to both populations of the thalamic nuclei , and the long range afferents of the thalamic relay population project to both populations of the cortex , as depicted in Fig 1 . The delays introduced by these long range afferents might play a crucial role in cortical dynamics [58 , 59] . However , as the axonal conduction delay between thalamus and cortex is rather small [60–63] , we approximate it by a convolution with an alpha function [64] , which can be written as ϕ ¨ k = ν 2 Q k ( V k ) - ϕ k - 2 ν ϕ ˙ k , ( 11 ) where ϕk is the resulting delayed firing rate and ν depicts the axonal rate constant of that connection . We have provided a justification for that approximation in the supporting information S2 Text . In the case of short range connections ϕk can be replaced with Qk . The parameters of the full model are given in S1 Table . We model an auditory stimulus as an elevation in background noise ϕ n ″ ( square pulse ) being gated through the thalamus . For all stimuli , we use a duration of 80 ms and 70 spikes per second . KCs and SOs were detected similar to [10] . The model output was bandpass filtered between 0 . 25 and 4 Hz . Zero-crossings were detected to extract the negative half-waves . Negative half-waves with peaks below -69 mV were considered to be KCs/SOs . Experimental data has been described in [65] . 11 Subjects were measured during two experimental nights in balanced order in a stimulation and control condition . For averages of the endogenous activity , data was taken from the control condition . See S1 Dataset . The model was implemented in C++ and run within MATLAB R2015b , using a stochastic Runge-Kutta method of 4th order [66] with a step size of 0 . 1 ms . The code can be found at github [67 , 68] .
In the isolated thalamic module , incorporation of the intrinsic currents may lead to oscillations in the spindle band . We follow closely the mechanisms established in the models by [15 , 19 , 20] . Physiologically , these oscillations are generated , through reciprocal interaction of the RE and TC populations . A LTS in the RE population causes hyperpolarization in the TC population , that deinactivates its T-type calcium current . Upon release from inhibition a rebound of activity occurs , that in turn drives the RE module to produce another LTS . Additionally , the deinactivation of the T-type calcium currents requires a strong tonic hyperpolarization by a potassium leak current [15 , 19] . As previously shown in [5 , 15 , 20 , 69] , the rhythmicity of spindle occurrence and the waxing and waning of the spindle amplitude is caused by an anomalous rectifier channel Ih . A sequence of LTS leads to the build-up of calcium , which increases the effective conductivity g ¯ h = g ¯ h ( m h 1 + g i n c m h 2 ) of Ih . The ensuing depolarization of the TC population increasingly counteracts its ability to produce a LTS and terminates the spindle oscillation [70 , 71] . Therefore , we chose g ¯ h and g ¯ LK as bifurcation parameters . A two-dimensional bifurcation analysis of the thalamic module reveals the existence of a Hopf bifurcation , as depicted in Fig 2 , which generates continuous oscillations in the spindle band due to hyperpolarization induced rebound bursts , see Fig 3-CI and 3-CII for representative time series . The torus bifurcations emerge from a blue sky catastrophe that is generated by the slow-fast interaction between the fast T-type channels and their slow modulation via Ih , which is similar to other models that exhibit switching between tonic spiking and structured bursting activity [72 , 73] . As depicted in Fig 3-SI and 3-SII this leads to spindle like oscillations in the orange shaded regions in Fig 2 . The spindles exhibit an oscillation frequency of around 13 Hz . The spindle frequency depends on the strength of the T-type calcium current g ¯ T . Importantly , spindle oscillations are initiated intrinsically . The thalamic module does not require modulatory input from external sources to initiate/terminate them . Activation of Ih is responsible for a refractory period that follows a spindle . As long as Ih activation persists , LTS generation is impeded and stronger perturbations are necessary to trigger spindle oscillations . Consequently , an increase in g ¯ h results in a larger inter-spindle interval . The left spindle regime ( SI ) is encased by the Hopf and the torus bifurcation , whereas the right spindle regime ( SII ) is constrained by two global bifurcations that are indicated by the dashed gray lines . The vertical line marks the emergence of the torus bifurcation , whereas the horizontal gray line marks the cusp bifurcation where the two saddle-nodes that accompany the left torus bifurcation vanish . Furthermore , for larger values of g ¯ LK the model transitions from high frequency spindle oscillations to low frequency delta oscillations , e . g . Fig 3-DI and 3-DII . In the coupled system , the cortex provides excitatory drive to the thalamic module , since it is predominantly in the active state . In order for the thalamic module to exhibit rhythmically occurring spindle oscillations we had to adjust g ¯ h and g ¯ LK slightly to account for that additional depolarization ( see Table 2 ) . We choose the right spindle regime , as it was suitable for reproduction of both sleep stage N2 and N3 . As can be seen in Fig 4 spindles may be triggered by KCs in the full model , but may also occur independent of KCs . During a KC the sudden drop of excitatory drive hyperpolarizes the RE and TC population , leading to deinactivation of IT . The ensuing depolarization upon the transition back to the active state triggers a LTS and a spindle sequence in turn . The spindle then projects back into the depolarizing phase of the KC . This is in good agreement with the grouping of spindles and KCs observed experimentally [6 , 74] . Although less likely the model can also give rise to KCs triggered by a spindle . This can be achieved by increasing the connection strength from the thalamic to the cortical module ( model output not shown ) . During N2 , KCs occur at a low rate . Hence , spindle initiation and termination are closely linked to the time course of Ih ( Fig 3A ) , similar to the isolated thalamic module . The parameters for the output in Fig 4 are given in Table 2 . Given the parameter setting in Table 2 , the cortical module is within a stable focus , close to a Hopf bifurcation accompanied by a canard explosion . This leads to noise driven medium amplitude background oscillations around the stable focus , that are interrupted by large amplitude deflections ( KCs ) . In good agreement with experimental findings , KCs also appear within the isolated cortex , although they may be initiated through thalamic input . On the transition to sleep stage N3 the canard phenomenon vanishes in a cusp bifurcation and only a high amplitude limit cycle remains . SOs are noise driven oscillations around a stable focus , close to a Hopf bifurcation [30] . In contrast to sleep stage N2 spindle initiation and termination are now dominated by the modulatory input from the cortical module , that overrules the Ih rhythm . Rather than occurring rhythmically spindles are time-locked to the depolarized phase of a SO . In Fig 5 an example time series is shown . Importantly , not every SO is able to trigger a spindle , as can be seen in Fig 5 ( 9–12 s , 13–15 s ) . We observed that in a sequence of SOs the first triggers a spindle , which leads to an activation of Ih . This reduces spindle amplitude or even inhibits spindle initiation by the following SO . To further validate the model , we determined averages of the generated EEG signal and fast spindle power time-locked to the negative peak of the endogenous KCs/SOs during N2 and N3 . This method is often used to illustrate the grouping of spindles by SOs and morphological features of SOs , e . g . in [6 , 13 , 65] . Model output and data for N2 and N3 is depicted in Fig 6 . As can be seen in Fig 6 , the grouping of spindles by SOs is present in the model . Spindle power is highest during the positive half-wave following the negative peak . However , there are some notable differences . Compared to the experimental data the initial depolarization preceding the transition to the down state is less prominent , leading to a shallower slope of the transition to the down state . In the thalamocortical model the transition to the depolarized up state occurs considerably earlier with a time to peak of 300 ms , compared to 440 ms in the data . This stems from strong depolarizing input by thalamic spindle bursts , which start directly after the negative peak of a KC/SO and push the cortex further into the depolarized state . However , this is still in line with other experimental studies , that find different timings of spindles for the supplementary motor area of the cortex [75] . In the following we show the ability of the model to reproduce data from a recent experiment in humans performing auditory closed-loop stimulation during NREM sleep [13] . The stimulation protocol is as follows: After the negative peak of a SO was detected , two auditory stimuli were applied phase-locked to the following positive peak of the depolarized up phase of the detected and the subsequent SO . In the experimental study the delay time between the negative peak and the ensuing positive half-wave peak was determined for every subject independently . The second stimulus followed after a fixed interval of 1075 ms . Detection was then paused for 2 . 5 s . We accordingly determined the delay time from the model output , resulting in a delay of 450 ms for the N3 parameter setting . The second stimulus was chosen to occur 1075 ms after the first one and we also paused detection for 2 . 5 s . Stimuli are given as elevations in mean background noise of the thalamic relay population for a duration of 80 ms . Fig 7 shows the averaged EEG signal and model output time-locked to the first stimulus ( t = 0 ) . There is a good agreement between model output and the experimental data . Especially the large amplitude , late components of the ERP are very close to the original waveform . The early component of the evoked potential , the P200 , can be seen in the experimental data after each stimulus , but it is more pronounced in the model output . In addition , the evoked spindle responses of model and data also have similar time courses . In both cases spindle power is systematically increased during the depolarized up phases induced by the stimuli . However , the strong increase in spindle power seen in the data after the first stimulus is not visible in the model . We hypothesize this to stem from a recruitment effect , where the stimulus activates a larger fraction of the thalamus than the endogenous slow oscillation would . As our thalamic module is a point model without any spatial extent , these effects are excluded by construction . Interestingly , in the experimental data there is a drop in spindle power after the second stimulus is applied . This seems to be a refractoriness of the thalamus after the second slow oscillation , which has also been observed in [76] . Despite the model showing such a refractory period in the isolated thalamus ( Fig 3A ) , as well as during trains of endogenous SOs in the full model ( Fig 8A ) , it lacks it upon stimulation ( Fig 8B ) . This happens because stimulation disturbs the Ih mediated spindle termination mechanism . As the stimulation depolarizes the TC population , the calcium concentration drops , because calcium influx through the IT current stops and calcium leaks out with a time constant of 10ms . Without the elevated calcium concentration , Ih deactivates back to baseline levels and immediately allows for a new full fledged spindle .
The model emphasizes the role of IT and Ih currents in the generation of thalamocortical rhythms as they were sufficient to reproduce the investigated EEG phenomena . It reproduces the grouping of spindles and KCs/SOs , observed in human EEG [6] , that is thought to play a crucial role in the consolidation of memory [77 , 78] . Additionally , it exhibits refractoriness of spindle oscillations , i . e . not every SO in a train of endogenous SOs triggers a spindle . Although adding extra currents increases dimensionality and parameter space , the model still preserves the overall simplicity and computational efficacy common to neural mass models . Relative to the negative deflection of a KC , spindles consistently start earlier than in the data . Consistently , the depolarizing up phase of endogenous KCs and SOs arrives earlier in the model than in the data . A comparison with the results from the isolated cortical module shows , that this is mostly due to strong depolarizing input from the thalamus . Yet , there is no clear explanation for the difference between model and experiment . It might be due to the simplification of the intrinsic mechanisms , e . g . firing rate adaptation in cortex and spindle dynamics in thalamus . On the other hand it could also be that finer details , e . g . spatial extension or the layered structure of the cortex are important for its temporal dynamics . Also the way conduction delays between cortex and thalamus were implemented , namely via an extra convolution with an alpha function , might play a role . A recent experimental study suggests that the refractoriness of thalamic spindles is a limiting factor for the impact of auditory stimulation upon memory consolidation [76] . They found , that longer trains of stimuli do not provide any benefit in memory consolidation compared to the two stimulus protocol . Remarkably , the first stimulus triggers a strong spindle , whereas the following stimuli show a diminished spindle response . This clearly indicates the importance of the grouping of spindles and SOs for the consolidation of memory . In contrast to these experimental findings , auditory stimulation in the model alleviates the refractoriness of the thalamic module , leading to spindle oscillations with similar amplitude following every stimulus . This is because strong depolarization of the thalamic populations by the stimulus interrupts the thalamic Ih rhythm . We see this as a challenge for the understanding of how auditory stimulation is processed during sleep and how it interacts with spindle generation . Recently , Cona et al . also developed a neural mass model to describe the sleeping thalamocortical system [79] . They combined two distinct firing modes via the activation of the T-type calcium current , showing that this multiplicative change in firing rate can lead to periodic spindle-like oscillations . However , in this study we include the currents directly into the equation of the membrane voltage , similar to [30 , 54] . Our model relates directly to scalp EEG signals during natural sleep and auditory stimulation . In our model , we induce the transition between the different sleep stages by changes of the three key parameters ( gKNa and σp in the cortex and g ¯ LK in the thalamus ) , that are directly linked to the action of neuromodulators [30 , 80–82] . These parameters are known to be affected by neuromodulators , such as noradrenalin , serotonin and acetylcholine [44 , 46 , 50 , 51 , 83] , whose concentrations vary over the night . Regulation of neuromodulator concentrations arises through complex interactions within different sleep regulatory networks [53 , 84] . Recently there has been progress in the mathematical description of sleep regulatory networks [85–89] . However , as we focus on the different dynamical modes the thalamocortical system can exhibit and how thalamus and cortex interact , we do not include sleep regulation in this manuscript . The waveform of a KC has been described as being biphasic , consisting of a large negative deflection ( down state ) followed by a pronounced depolarization ( up state ) —or triphasic , comprising an initial positive bump followed by a down state and an up state . Menicucci et al . [90] analyzed the shapes of KCs in N2 and N3 and found that on average a triphasic pattern , up-down-up , is present in both sleep stages . Our model does not show this sequence for sleep stage N2 . In vivo , sleep stage N2 is rarely stationary and spans varying depths of sleep , as well as transitions to other sleep stages . In contrast , our model depicts idealized N2 at a single point in time , to separate it from wakefulness and N3 . Choosing a parameter setting closer to N3 will naturally give rise to a depolarization preceding the down state . We predict that biphasic KCs should be found mostly in early N2 or very late N2 , as in the second half of the night after the major SWS episodes . The model is consistent with the observation that during N2 and N3 of natural sleep the cortex is mostly in the active state [91] . We adopt the view of [30] , where KCs were characterized as transient events—reversed spikes—initiated by a canard explosion . Consequently the down state is never stable in our model . This may seem counterintuitive as many intracellular recordings support the notion of bistability . However , neural mass models represent population averages , whereas intracellular recordings only sample individual members of a population , leaving open this alternative interpretation derived from stereotypical graphoelements in the EEG .
|
Sleep plays a pivotal role for the consolidation of memory . While REM sleep had originally been the focus of research due to its similarity with wakefulness , more recent studies suggest that different sleep stages are responsible for the consolidation of different types of memory . To better understand the changes in neuronal dynamics between the different sleep stages , neural mass models are a valuable tool , as they relate directly to the large-scale dynamics measured by an EEG . Here , we present a model of the sleeping thalamocortical system . We first show that the isolated thalamic submodule is able to generate different oscillatory behavior found in vivo . Furthermore , the full thalamocortical model reproduces the EEG of sleep stages N2 and N3 and preserves the temporal relationship between cortical K-complexes/slow oscillations and thalamic sleep spindles . A comparison with event related potential data from a recent sleep study in humans demonstrates its possible application in predicting the outcome of external stimulation on EEG rhythms . Our study shows , that a neural mass model incorporating few key mechanisms is sufficient to reproduce the complex brain dynamics observed during sleep .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"sleep",
"depolarization",
"membrane",
"potential",
"brain",
"electrophysiology",
"brain",
"electrophysiology",
"neuroscience",
"learning",
"and",
"memory",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"physiological",
"processes",
"functional",
"electrical",
"stimulation",
"cognition",
"brain",
"mapping",
"memory",
"bioassays",
"and",
"physiological",
"analysis",
"neurotransmitters",
"neuroimaging",
"electroencephalography",
"research",
"and",
"analysis",
"methods",
"imaging",
"techniques",
"clinical",
"neurophysiology",
"memory",
"consolidation",
"electrophysiological",
"techniques",
"thalamus",
"gamma-aminobutyric",
"acid",
"biochemistry",
"diagnostic",
"medicine",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"cognitive",
"science",
"neurophysiology"
] |
2016
|
A Thalamocortical Neural Mass Model of the EEG during NREM Sleep and Its Response to Auditory Stimulation
|
Distinct classes of neurons and glial cells in the developing spinal cord arise at specific times and in specific quantities from spatially discrete neural progenitor domains . Thus , adjacent domains can exhibit marked differences in their proliferative potential and timing of differentiation . However , remarkably little is known about the mechanisms that account for this regional control . Here , we show that the transcription factor Promyelocytic Leukemia Zinc Finger ( PLZF ) plays a critical role shaping patterns of neuronal differentiation by gating the expression of Fibroblast Growth Factor ( FGF ) Receptor 3 and responsiveness of progenitors to FGFs . PLZF elevation increases FGFR3 expression and STAT3 pathway activity , suppresses neurogenesis , and biases progenitors towards glial cell production . In contrast , PLZF loss reduces FGFR3 levels , leading to premature neuronal differentiation . Together , these findings reveal a novel transcriptional strategy for spatially tuning the responsiveness of distinct neural progenitor groups to broadly distributed mitogenic signals in the embryonic environment .
The formation of neural circuits within the developing central nervous system ( CNS ) depends upon the spatially and temporally ordered generation of distinct classes of neurons and glia from multipotent neural stem and progenitor cells ( NPCs ) . An essential feature of this progression is the ability of NPCs to self-renew in a manner that permits early-born cells such as neurons to form while maintaining a sufficient progenitor pool to generate later-born cell types such as glia . At the heart of this process is the interplay between mitogenic signals from the extracellular environment and cell intrinsic factors , which integrate this information to permit either progression through the cell cycle or the onset of terminal differentiation [1] . At early stages of development , NPCs are broadly responsive to mitogenic stimulation . However , this responsiveness markedly changes over time and often becomes region-specific such that some groups of cells proliferate for protracted time periods while others rapidly differentiate [2] , [3] . While important for determining the size and shape of the developing CNS , the mechanisms underlying these differences in mitogen sensitivity remain poorly defined . These features of NPCs are exemplified in the developing spinal cord , where many extrinsic and intrinsic factors regulating progenitor maintenance and differentiation have been characterized . In the early neural plate and tube , NPCs are organized in a proliferative neuroepithelial sheet and sustained by the mitogenic actions of several growth factors , particularly Fibroblast Growth Factors ( FGFs ) . FGFs are broadly present in neural tissues and the surrounding mesoderm and act through receptor tyrosine kinases ( FGFRs ) expressed by NPCs throughout the course of neural development [4]–[6] . Ligand binding to FGFRs activates multiple downstream signaling cascades such as the MAPK/ERK , PI3K/AKT , PLCγ , and STAT3 pathways to both promote cell division and inhibit neuronal differentiation [7] . Among the many targets of FGF signaling are members of the SOXB1 family of transcription factors , which play key roles , first , sustaining neuroepithelial progenitor properties and , second , blocking the expression and activity of proneural basic helix-loop-helix ( bHLH ) proteins that promote cell cycle exit and neuronal differentiation [8]–[12] . As development proceeds , NPCs become increasingly poised to undergo terminal differentiation through the actions of retinoids , which activate the expression of homeodomain and bHLH transcription factors such as PAX6 and OLIG2 . These factors participate in the dorsoventral patterning of NPCs and promote the accumulation of proneural bHLH proteins needed to trigger cell cycle exit and neuronal differentiation [13] . These activities are counterbalanced by the mitogenic actions of FGFs acting in concert with NOTCH receptors and their downstream effectors , the HES proteins [5] , [14] . Mutual inhibition between proneural bHLH and HES proteins sets up a dynamic equilibrium between self-renewal and terminal differentiation [15] that must be resolved in a progenitor domain-specific manner . The mechanism by which this resolution is achieved has remained unclear . One possibility is that further intrinsic or extrinsic factors regulate this equilibrium by regionally altering the sensitivity of NPCs to mitogens , such as the FGFs . The period of neurogenesis in the spinal cord is relatively brief , lasting for only a few days in chick and mouse development , after which time undifferentiated NPCs up-regulate expression of the pro-glial transcription factors SOX9 and NF1A , and begin to give rise to astrocyte and oligodendrocyte precursors [16] , [17] . During this transition , NPC maintenance remains dependent on FGF signaling [18]–[20] . Moreover , the expression of FGFR3 becomes particularly enriched in astrocyte progenitors [21] , suggesting that differential FGF signaling might play a role in the specification or expansion of astroglial cells over others . Despite its importance for progenitor maintenance and gliogenesis , very little is known about the mechanisms through which FGFR expression and activity is regulated in specific populations of NPCs . To identify the regulatory factors that influence NPC maintenance in the spinal cord , we carried out expression profiling experiments to define the genes that are deregulated in the spinal cord of Olig2 mutant mice . Olig2+ NPCs exhibit limited capacity for self-renewal , suggesting that Olig2 represses genes that promote NPC proliferation [22]–[24] . Through these studies , we identified the gene Zbtb16 , which encodes the Promyelocytic Leukemia Zinc Finger ( PLZF ) transcription factor , as one of the most prominently elevated genes in the absence of Olig2 function ( Figure S1B–C ) . PLZF is a member of the BTB/POZ family of transcription factors known to regulate progenitor maintenance in multiple tissues [25] . PLZF was initially identified as a protein whose functions are subverted through chromosomal rearrangements resulting in acute promyelocytic leukemia [26] . It has subsequently been found to be a key stem cell maintenance factor in both the hematopoietic system and male germline [26] . PLZF also exhibits a highly dynamic expression pattern in the developing rodent forebrain and hindbrain [27] , and is associated with neural rosette formation in differentiated embryonic stem cell cultures [28] . More recently , PLZF was found to suppress the earliest steps in neurogenesis in developing zebrafish [29] , though its mechanism of action and role at later stages of development have not been resolved . In this study , we identify a novel role for PLZF preserving a population of NPCs in the central region of the spinal cord from early development through to the onset of astrogliogenesis . Loss of PLZF compromises progenitor maintenance , leading to premature neuronal differentiation . Conversely , its elevation is sufficient to repress neurogenesis and enhance glial cell production . These phenotypes result from the ability of PLZF to promote the expression of FGFR3 in NPCs , which then acts though the STAT3 pathway to gate the response of NPCs to FGF mitogens present in the neural tube . This mechanism permits PLZF-expressing progenitors in the central spinal cord to differentiate at a slower pace than neighboring cells and expand the population of cells available for astrocyte production . Together , these data indicate that PLZF provides a critical link between the transcriptional programs and mitogenic signals that regulate the balance between NPC proliferation and differentiation .
To explore the function of PLZF in neural development , we first mapped its expression relative to other markers of NPCs and differentiated neurons in the chick spinal cord . PLZF is first detected at e2 [Hamburger Hamilton ( HH ) stage 10] in a subset of SOX2+ NPCs in the open neural plate before the onset of neurogenesis , and then becomes broadly expressed by NPCs in the neural tube as neurogenesis commences between e2 . 5 ( HH stage 15 ) and e3 ( HH stage 17 ) ( Figure 1A–D; unpublished data ) [9] , [30]–[33] . Between e4 to e5 ( HH stages 21–28 ) , the peak period of spinal cord neurogenesis [9] , [31]–[33] , PLZF becomes restricted to pdI6-p2 progenitors in the central region that express high levels of IRX3 and PAX6 , and bounded by progenitors expressing MSX1/2 dorsally and OLIG2 ventrally ( Figure 1E–F , I–J , M; Figure S2B–C , G–H ) . These PLZF+ progenitors are thus fated to give rise to interneurons early in development and astrocytes at later times [34] , [35] . A very similar pattern of expression is observed in the developing mouse spinal cord ( Figure S1 ) , suggesting an evolutionarily conserved role for PLZF in spinal cord development . Within the central progenitor domain , PLZF is prominently expressed by dividing SOX2+ progenitors and is down-regulated as cells express the proneural transcription factor NEUROG2 and begin to form differentiated interneurons marked by LHX1/5 and NEUN staining ( Figure 1F , K; unpublished data ) . This pattern of PLZF expression in NPCs is distinct from that seen in the dorsal spinal cord , where PLZF is excluded from the ventricular zone ( VZ ) and instead expressed by differentiated interneurons throughout the course of development ( Figure 1K , L; unpublished data ) . For the remainder of this study , we will focus solely on the actions of PLZF in the central progenitor populations . From e5–e7 ( HH stages 25–30 ) , progenitors in the intermediate spinal cord up-regulate the expression of the early glial fate determinants SOX9 and NF1A , and begin to transform into astrocyte progenitors [16] , [17] . During this time , PLZF expression in the VZ overlaps with SOX9 and NF1A , but then declines by e9 ( HH stage 35 ) , the time at which astrocyte progenitors migrate into the grey matter and differentiate ( Figure 1G , H , L , N; Figure S2A–J ) [16] , [17] . PLZF is undetectable within migratory astrocyte progenitors by e10 ( HH stage 36 ) and later stages ( unpublished data ) . PLZF was also excluded from early and late SOX9+ OLIG2+ oligodendrocyte progenitors , consistent with its down-regulation from the OLIG2+ motor neuron progenitors from which many oligodendrocyte progenitors emerge ( Figure 1L; Figure S1O ) . Together , these data indicate that PLZF is associated with the maintenance of a central population of spinal NPCs during the progression from neurogenesis to gliogenesis , and its extinction in both cases coincides with the onset of cellular differentiation ( Figure 1M–N ) . Since PLZF is associated with stem and progenitor cell maintenance in other tissues , we sought to examine whether its function plays a comparable role in the developing spinal cord . We first investigated the consequences of elevating PLZF activity on NPC maintenance and neuronal differentiation . Expression vectors encoding PLZF and an IRES-nuclear Enhanced Green Fluorescent Protein ( nEGFP ) reporter cassette under the control of the cytomegalovirus enhancer-chick beta actin promoter were electroporated into the chicken spinal cord , and embryos collected 2 d later to assess changes in neuronal differentiation . In spinal cords electroporated with the empty vector , ∼60% of transfected cells expressed NPC markers such as SOX2 while the remaining ∼40% expressed neuronal markers such as NEUN ( Figure 2A , E–F , J ) . When PLZF was misexpressed , the fraction of transfected cells expressing SOX2 increased by ∼26% relative to empty vector controls and the proportion giving rise to neurons was reduced by ∼39% ( Figure 2B , E , G , J ) . Thus , ectopic PLZF expression can restrict neuronal differentiation and sustain cells in a progenitor state . To determine the basis of these changes , we examined the impact of PLZF misexpression on proneural bHLH transcription factors . In the spinal cord , three proteins—ASCL1 , NEUROG1 , and NEUROG2—play a critical role promoting cell cycle exit and neuronal differentiation in different regions [36] . Where PLZF was elevated , we observed a ∼14% reduction in the expression of both ASCL1 and NEUROG1 mRNA and a ∼28% decrease in the number of NPCs expressing NEUROG2 protein relative to spinal cords transfected with the control vector ( Figure 2K–L , N–O , Q–R , T–V ) . We further investigated whether these changes resulted from increased expression of HES genes , which are well-described inhibitors of proneural bHLH gene expression [14] , [36] . Despite clear changes in proneural gene expression , there was no apparent effect of PLZF misexpression on the two primary HES genes expressed in the chick spinal cord , HAIRY1 and HES5-2 ( Figure S3A–B ) . PLZF misexpression also did not lead to the precocious onset or ectopic expression of early glial progenitor markers such as SOX9 and NF1A ( Figure S2K–R ) . PLZF thus appears to be capable of suppressing the expression of multiple proneural genes , blocking neuronal differentiation , and promoting NPC maintenance in a HES gene-independent manner . Previous studies have shown that many of PLZF's functions in embryonic development and tumor progression are related to its function as a transcriptional repressor , an activity mediated by the binding of cofactors such as histone deacetylases to PLZF's N-terminal BTB domain [37]–[40] . However , more recent studies have demonstrated that PLZF can work as a transcriptional activator in some instances [41] , [42] . To determine the mechanistic basis by which PLZF acts in neural progenitors , we generated constitutive repressor and activator forms of PLZF and measured their activity when electroporated into the developing spinal cord . These modified forms of PLZF were generated by appending either a potent transcriptional repression domain from the Drosophila Engrailed protein ( EnR-PLZF ) [43] or an activation domain from Herpes virus 16 ( VP16-PLZF ) [44] to the C-terminal DNA binding portion of PLZF . When misexpressed in the chick spinal cord , the EnR-PLZF fusion recapitulated all of the features of full-length PLZF misexpression , with the electroporated cells displaying high levels of neural progenitor markers such as SOX2 , reduced expression of proneural bHLH genes and proteins , and decreased propensity for neuronal differentiation ( Figure 2C , E , H , J , M , P , S–V ) . By contrast , VP16-PLZF had the opposite effect , directing most of the electroporated cells to undergo neuronal differentiation and lateral migration into the mantle zone of the spinal cord ( Figure 2D–E , I–J ) . Thus , the ability of PLZF to both promote progenitor maintenance and block neurogenesis appears to stem from its transcriptional repressor activities . Since ectopic PLZF is sufficient to enhance NPC maintenance ( Figure 2 ) , we next investigated whether PLZF is required for continued progenitor proliferation . Towards this end , we generated a plasmid vector encoding two short-hairpin RNAs to target the chick PLZF transcript ( shPLZF ) along with an IRES-nEGFP reporter cassette to identify the transfected cells . Electroporation of this construct into the spinal cord reduced PLZF levels to nearly background staining levels ( Figure 3A , G; Figure S4A–C ) , and led to substantial changes in NPC maintenance . The overall area of the VZ decreased by ∼20% , and the average expression level of SOX2 within the remaining transfected progenitors declined by ∼11% ( Figure 3B , H , M–N ) . These changes were further accompanied by reduced expression of other genes associated with NPC maintenance including HES5-2 and ID2 ( Figure 3C–D , I–J , O ) . As these progenitor features were lost , early differentiation markers such as NEUROG2 were correspondingly elevated ( Figure 3E , K , P ) . Despite these changes , we did not observe significant changes in the total number of NEUN+ or TUJ1+ neurons formed after PLZF knockdown ( unpublished data ) . Perhaps accounting for this result , we found that PLZF loss was associated with a ∼2-fold increase in the frequency of cells undergoing apoptotic cell death measured by activated CASPASE3 staining ( Figure 3F , L , Q ) . Importantly , defects in progenitor maintenance and cell death observed when PLZF was knocked down were rescued by the inclusion of an expression plasmid encoding human PLZF that lacks the shRNA target sequences ( Figure S4D–N ) , confirming the specificity of these manipulations . To complement this analysis , we investigated the effect of PLZF loss on progenitor maintenance and interneuron differentiation in Green's Luxoid mice ( Zbtb16Lu/Lu ) , which possess a nonsense mutation in the PLZF coding sequence that ablates its DNA binding function [45] . In contrast to the acute loss of PLZF function in the chick , we did not observe any overt signs of elevated cell death in Zbtb16Lu/Lu mutant mice ( unpublished data ) . However , using a panel of lineage-restricted makers on the spinal cords from Zbtb16Lu/Lu mutant and control littermates , we found that differentiation was increased among some interneurons whose progenitors normally express high levels of PLZF ( pdI6-p2; Figure 1I–J , M; Figure S2 ) . Specifically , we observed the number of cells in the p1 ( Dbx1− , Nkx6 . 1− , Sox11− ) and p2 ( Nkx6 . 1+ , Olig2− , Sox11− ) central progenitor domains were reduced by ∼12% in the Zbtb16Lu/Lu mutants , and this decrease was mirrored by a >14% rise in the number of dI6 , V1 , V2a , and V2b interneurons distinguished by their expression of specific transcription factors including Bhlhe22 ( Bhlhb5 ) , Foxp2 , Vsx2 ( Chx10 ) , and Gata3 ( Figure S5A–L , N–Q , T–W , Y–AA ) [24] , [34] . In contrast , no alterations were observed in the numbers of dorsal Isl1+ dI3 interneurons or ventral Isl1+ motor neurons , which respectively derive from Pax7+ dorsal progenitors and Olig2+ ventral progenitors that do not sustain high levels of PLZF expression under normal conditions ( Figure S5A–C , G–I , M , R , S , X , Y–AA; unpublished data ) . Together , these data suggest that PLZF function is required to maintain a population of progenitors within the intermediate spinal cord and restrict their differentiation into spinal interneurons . We next sought to determine the long-term consequences of manipulating PLZF activity on cell fate . Do the observed reductions in neuronal differentiation and enhanced progenitor maintenance associated with elevated PLZF expression ultimately result in increased glial production or continued expansion of neuroepithelial progenitors ? To discriminate between these outcomes , we used the Tol2 transposon-mediated gene transfer system [46] to stably transfect chick NPCs in ovo with either an IRES-EGFP or PLZF-IRES-EGFP expression cassette at e3 , and analyzed the effects on neuronal and glial development 12 d later at e15 . Since SOX2 is expressed by both NPCs and glial-restricted progenitors at this time , we used antibody staining for NESTIN as a marker for uncommitted neural progenitors along with NEUN and SOX9 to respectively distinguish differentiated neurons and glial progenitors . At the e15 time point , the majority of transfected cells had initiated lineage-specific differentiation irrespective of PLZF misexpression , reflected in a low frequency of NESTIN staining in both control ( ∼7% ) or PLZF-transfected ( ∼5% ) embryos ( Figure 4L; unpublished data ) . However , the differentiated fates of the transfected cells were markedly different . Whereas ∼27% of control transfected cells expressed NEUN , PLZF expression reduced this frequency to ∼9% ( Figure 4A , F , L ) . Instead , the majority ( ∼86% ) of the PLZF-transfected cells expressed glial progenitor markers such as SOX9 compared to ∼66% in the control population ( Figure 4B , G , L ) . Since the earlier expression of PLZF did not result in the precocious onset or ectopic expression of SOX9 , the astrocyte progenitor marker NF1A , or the definitive astrocyte differentiation marker GFAP ( Figure S2K–V ) , we infer that the transition of the PLZF-expressing progenitors towards gliogenesis proceeds along a normal developmental schedule . Despite its normal exclusion from OLIG2+ cells ( Figure 1I; Figure S1 ) , ectopic PLZF expression resulted in a ∼2-fold increase in the number of SOX9+ OLIG2+ oligodendrocyte progenitors , as well as a comparable increase in the expression of FGFR3 , which is commonly associated with astrocyte progenitors [21] , and a ∼2- to 3-fold increase in the number of cells expressing GFAP+ ( Figure 4C–E , H–J , M ) . Interestingly , these ectopic glial progenitors and glia were not uniformly distributed throughout the spinal cord but instead clustered adjacent to the VZ as if the cells were impaired in their differentiation or migration . Collectively , these data provide evidence that PLZF plays an important role preserving a pool of progenitors available for gliogenesis at the later stages of embryonic development , but its function must ultimately be silenced for glial cell maturation ( Figure 4N ) . We next set out to identify the mechanism by which PLZF maintains specific NPCs in an undifferentiated state . Since PLZF misexpression did not appear to elevate the expression of NOTCH-responsive HES genes ( Figure S3A–B ) , we considered other pathways known to block neurogenesis . Several observations suggested that the effects of PLZF on differentiation could be mediated by the FGF signaling pathway . First , the FGF pathway is crucial for the establishment , preservation , and proliferation of NPCs both in vivo and in vitro [7] . Second , FGF signaling positively regulates SOX2 expression and blocks differentiation [47] , [48] , much like PLZF misexpression . Third , in both the chick and mouse spinal cord we observed a striking coincidence between the expression patterns of PLZF and FGFR3 , one of the principal receptors that mediates FGF signaling during the period of neurogenesis under consideration in this study ( Figure 5A–F; Figure S1P–U; Figure S6A–F ) [6] . Taken together , these findings raised the possibility that PLZF promotes NPC maintenance by up-regulating FGFR3 expression and thereby enhancing the responsiveness of NPCs to FGFs in the embryonic environment . Supporting this model , ectopic expression of PLZF was able to expand FGFR3 expression , whereas PLZF knockdown decreased FGFR3 ( Figure 5G–K ) . These alterations in FGFR3 occurred without significant changes in the homeodomain proteins associated with dorsoventral patterning such as IRX3 , PAX3 , PAX6 , and PAX7 ( Figure S3E–H ) , suggesting that these effects were not simply due to alterations in NPC identity . Moreover , the effects were specific to FGFR3 , as PLZF manipulations did not alter the expression of either FGFR1 or FGFR2 ( Figure S6G–I ) . If changing the level of FGFR3 expression accounts for the actions of PLZF on NPC maintenance and differentiation , then directly elevating FGFR3 levels or blocking its receptor kinase activity should , respectively , recapitulate the effects of PLZF misexpression and knockdown . To test this prediction , we electroporated spinal cords with expression vectors encoding either full-length FGFR3 or a truncated , dominant-negative ( dn ) form of FGFR1 that forms nonfunctional heterodimers with FGFR3 and other FGFRs and blocks their downstream signaling activity [49] . Embryos transfected with FGFR3 displayed a strikingly similar phenotype to that observed after PLZF misexpression . In both cases , there was a ∼23% increase in SOX2 expression and a corresponding reduction in the formation of NEUN+ neurons within the transfected cells ( Figure 2A–B , E–G , J; Figure 5L–M , P–Q , T ) . In contrast , when endogenous FGFR3 activity was disrupted by dnFGFR misexpression , the fraction of transfected cells expressing SOX2 dropped by ∼25% , suggesting that loss of FGF signaling compromises progenitor maintenance in a manner similar to that seen following PLZF knockdown ( Figure 3H , M–N; Figure 5N , R , T ) . If PLZF acts by promoting FGFR3 expression , then the activity of FGFR3 should be epistatic to that of PLZF ( Figure 5U ) . In this case , the pro-progenitor activity of ectopic PLZF would be dependent upon FGFR function , while direct elevation of FGFR3 levels should , in turn , overcome the loss of NPCs seen after PLZF knockdown ( Figure 6AB ) . To examine this possibility , spinal cords were first concomitantly electroporated with expression vectors encoding both PLZF and dnFGFR . Supporting the hypothesis , the increase in progenitor maintenance associated with PLZF elevation was blocked , and cells instead differentiated precociously as observed with dnFGFR misexpression alone ( Figure 6A–C , E–G , I–K , Y ) . In the converse experiment , the coelectroporation of FGFR3 with shPLZF rescued the changes in SOX2 protein staining levels , size of the progenitor pool , and numbers of cells expressing NEUROG2+ cells seen after electroporation with shPLZF alone ( Figure 6M–O , Q–S , U–W , Z–AA , and unpublished data ) . Together , these experiments suggest that PLZF does indeed act upstream of FGFR3 . We next assessed how manipulations of PLZF and FGFR were reflected in the activity of the second messenger effectors of FGF signaling . Early in chick development , FGF stimulation is associated with increased phosphorylation of ERK1/2 and expression of the ETS domain transcription factor ETV1 ( ER81 ) and ETV4 ( PEA3 ) , as well as feedback inhibitors of the pathway such as SPRY1 and SPRY2 [50] , [51] . Surprisingly , we were unable to detect changes in any of these effectors in the chicken spinal cord , even under conditions in which embryos had been electroporated with constructs encoding constitutively activated forms of FGFR1 and FGFR3 [52] that potently blocked neuronal differentiation and expanded the progenitor pool ( Figure S6J–K , M–N; Figure S7A–D; unpublished data ) . These results indicate either that the available reagents are insufficient to report pathway activity at the stages of development examined or that PLZF and FGFR3 act through an alternative signaling pathway . STAT3 is a noncanonical effector of FGF signaling [53] , [54] that has been implicated in blocking neurogenesis and promoting either NPC maintenance or astrogliogenesis in various systems [55]–[57] . We also confirmed that STAT3 is expressed broadly throughout the VZ of the spinal cord at the time of our experiments ( Figure S7E ) . To test whether PLZF and/or FGFR3 regulate STAT3 activity in the spinal cord , we co-expressed either PLZF or FGFR3 with a STAT3 transcriptional reporter construct capable of measuring pathway activity in the chick embryo [58] . In both cases , the activity of the STAT3 reporter was elevated ∼2- to 5-fold ( Figure S7F–I ) . Consistent with this result , we found that electroporation with a plasmid encoding a constitutively activated form of STAT3 ( STAT3-C ) [59] promoted progenitor maintenance and blocked neuronal differentiation in a manner that was nearly identical to the results seen with PLZF and FGFR3 misexpression ( Figure 2A–B , E–G; Figure 5L–M , O–Q , S–U ) . Moreover , expression of a dominant-negative mutant form of STAT3 was sufficient to counter the progenitor promoting activity of PLZF ( Figure 6A–B , D–F , H–J , L , Y , AB ) while STAT3-C overcame progenitor loss associated with PLZF knockdown ( Figure 6M–N , P–R , T–V , X , Z–AB ) . Thus , the actions of PLZF and FGFR3 appear to be mediated by the STAT3 arm of the FGF signaling pathway rather than the ERK/MAPK pathway typically associated with FGFR activity . These data are consistent with a role for PLZF in gating both the abundance of FGFR3 on NPCs and the activity of its downstream effectors to sustain the NPC pool in the spinal cord . The observations that an increase in FGFR3 is sufficient to expand the progenitor pool and impede differentiation suggest that NPC maintenance in the spinal cord might be principally constrained by the amount of FGFRs present on the cells rather than availability of FGF ligands in the environment . Indeed , previous studies have shown that FGF2 and FGF8 , two of the preferred ligands for FGFR3 , are broadly expressed throughout the VZ of the developing spinal cord and present in the cerebrospinal fluid , and thus unlikely to provide spatial control over NPC expansion ( Figure S4L ) [5] , [60]–[63] . To test whether FGFR3 levels are limiting , we reasoned that ectopic expression of FGFs throughout the spinal cord should elicit progenitor maintenance responses in a regional manner , with stronger effects seen in the PLZF+ FGFR3high intermediate region of the spinal cord compared to the PLZF− FGFR3low dorsal spinal cord . For this analysis , PAX6 was used to monitor NPCs in place of SOX2 . The extent of PAX6 expression completely overlaps with SOX2 , and the high versus low levels of PAX6 in the intermediate and dorsal spinal cord served as a convenient proxy for assessing the presence or absence of PLZF ( Figure S3C–D ) . Consistent with our prediction , transfection of the spinal cord with an expression vector for FGF8 led to a ∼20% increase in the expression of PAX6 and the fraction of cells incorporating BrdU in the dorsal spinal cord , compared to a ∼75% enhancement in the intermediate spinal cord ( Figure 7A–E , H–L , V–W ) . Moreover , FGF8 misexpression did not significantly change NEUROG2 expression in the dorsal spinal cord , whereas it reduced NEUROG2 by >20% in the intermediate spinal cord ( Figure 7A , F–H , M–N , X ) . Thus , PLZF+ progenitors appear to be more responsive to FGF ligand stimulation than adjacent PLZF− domains . Based on these results , we tested whether PLZF misexpression could enhance the response of NPCs to ectopically expressed FGFs . In regions of the spinal cord where PLZF and FGF8 were cotransfected , the VZ became dramatically enlarged and disorganized , with a ∼2-fold increase in the number of PAX6+ and BrdU+ cells , and a ∼20–25% reduction in the proportion of those progenitors undergoing neurogenesis ( Figure 7O–U , X; Figure S8A–O ) . These effects were distinct from the relatively mild expansion of NPCs seen after ectopic expression of PLZF or FGF8 alone , yet remarkably similar to the effects seen after electroporation with constitutively activated FGFR3 plasmids ( Figure S8P–R ) . Collectively , these experiments reveal regional differences in the sensitivity of spinal cord NPCs to FGF mitogen stimulation that correlates with their relative expression of PLZF and FGFR3 . Moreover , elevating the level of PLZF has the capacity to render cells hyperresponsive to FGF stimulation .
PLZF is first expressed throughout the VZ during the early phase of NPC expansion , but then becomes strikingly restricted to a central domain of progenitors fated to give rise to ventral interneurons early in development and astrocytes at later times . By manipulating PLZF expression in both chicken and mouse embryos , we found that its function is both necessary and sufficient to suppress neuronal differentiation and permit the emergence of glial progenitors . Although PLZF exhibits pro-glial activity , our data suggest that this function is most likely indirect and related to its effects on progenitor maintenance as neither misexpression nor removal of PLZF function appeared to significantly alter the onset of expression for the early glial fate determinants SOX9 and NF1A ( Figure S2K–V; unpublished data ) . Moreover , PLZF misexpression led to a marked increase in the numbers of both astrocyte and oligodendrocyte progenitors , even though PLZF is not normally present in oligodendrocyte progenitors . Lastly , PLZF levels notably decline as astrocyte progenitors begin to differentiate , and the sustained expression of PLZF appears to impede glial cell maturation . Taken together , these data suggest that the primary role for PLZF is to preserve the progenitor pool over the course of neurogenesis such that it can acquire competence to give rise to glial cells at later stages of development . It is notable that reducing PLZF activity resulted in consistent but partial phenotypes . This lack of an absolute necessity for PLZF may stem from functional redundancy among genes of the BTB/POZ family . Indeed , we observed a greater suppression of progenitor maintenance following electroporation with an activator form of PLZF ( VP16-PLZF ) that can override the endogenous transcriptional repressor functions of PLZF and potentially other BTB/POZ proteins than with shRNA constructs selectively targeting PLZF . To date , we have identified five additional family members with expression in the developing spinal cord , suggesting there may be complementary functions with PLZF ( Z . B . G . and B . G . N . unpublished observations ) . A comparable situation exists within the hematopoietic lineage where the lack of a prominent phenotype in either PLZF-null mice or in cell lines transfected with PLZF targeting shRNA is attributed to the presence of other BTB/POZ proteins such as the closely related FAZF [25] . The FGF pathway also receives inputs from other signaling networks and has extensive feedback regulatory mechanisms [7] , [65] . Thus , the absence of PLZF and reduced expression of FGFR3 could be compensated over time by changes in these modulatory components . Alternatively , the rather mild loss of function phenotypes seen in the nervous system may reflect the subtlety by which PLZF and FGFR3 act to keep cells in a proliferative state . Rather than constituting a simple on/off switch for progenitor maintenance , PLZF and FGFR3 finely sculpt the timing of neuronal differentiation and proportions of neurons formed to shape the functionality of neural circuits . The growth and morphogenesis of the nervous system depends upon the ability of the FGFs to promote the rapid proliferation of NPCs and block neuronal differentiation [4] , [66] . FGF8 is initially expressed throughout the neural plate but then becomes progressively restricted to the adjacent paraxial mesoderm and notochord [5] , [61] , [62] . FGF2 is also expressed first in low levels by the notochord , but is ultimately present throughout the VZ of the spinal cord , and within the embryonic cerebrospinal fluid [60] , [62] , [63] . Despite the broad distribution of these FGF mitogens , NPCs in the spinal cord exhibit spatially distinct proliferative responses [2] , [67] . Our findings suggest the differential effects of FGFs may stem , in part , from the regional control of FGFR3 expression by PLZF . When FGFR3 levels were increased by misexpression of either PLZF or FGFR3 , NPCs continued to proliferate and neuronal differentiation was accordingly blocked . These findings strongly suggest that receptor availability is a limiting factor in NPC proliferation and maintenance ( Figure 7Y ) . This conclusion is further supported by the observation that the FGFR3high NPCs in the intermediate spinal cord display a heightened response to ectopically expressed FGF8 compared to their FGFR3low dorsal counterparts ( Figure 7Y ) . Regional restriction of FGFR3 expression may also be relevant for ventral progenitors . OLIG2+ motor neuron progenitors express low levels of PLZF and FGFR3 and , perhaps as a consequence , differentiate earlier than many other progenitor populations in the spinal cord ( Figure 7Y ) [68] , [69] . The limited expression of FGFR3 within OLIG2+ cells may also explain why these cells exhibit limited stem cell capacities when grown in vitro [22] , [70] , since the culture conditions used for NPC expansion typically rely upon FGFs as the primary mitogenic signal . While PLZF exhibits a positive effect on FGFR3 expression , the target of this interaction remains unresolved . Our data suggest that PLZF carries out its functions as a transcriptional repressor as seen in many other systems [37]–[40] . Thus , PLZF may be acting by repressing an inhibitor of FGFR3 expression . Such an inhibitor may have more general roles controlling proliferation; therefore , the identification of the direct targets of PLZF and the cofactors that it associates with in the developing CNS will be an important area for future investigation . Within the CNS , FGF signaling is implicated in many steps in neuronal development including neural induction , regional patterning , progenitor expansion , axon outgrowth and guidance , and synaptogenesis [7] . This broad range of activities raises the question of how such distinct outcomes may be achieved from a common signal ? In vertebrates , some of the diversity in response stems from the varying affinities of the 22 FGF ligands for four FGFRs , which exist in multiple splice isoforms , as well as interactions between FGFs and FGFRs with particular heparin sulfate proteoglycans present in the extracellular matrix [7] . By selectively promoting the expression of FGFR3 , PLZF could render the central spinal cord particularly sensitive to particular ligands or bias the selection of downstream signaling effectors . Upon ligand binding , FGFRs dimerize and phosphorylate a number of secondary messengers that feed into the ERK/MAPK , AKT/PI3K , PLCγ/PKC , and/or STAT pathways [7] . It is currently unclear whether the diversity in cellular responses to FGF exposure can be explained simply by the differential activation of one or more of these signaling pathways . Nevertheless , it is clear that cellular responses to FGF are strongly influenced by the presence of particular intrinsic factors and most likely crosstalk with other environmental signals . For example , in the developing brain , FGF8 exposure can drive cells to adopt a forebrain or midbrain identity depending on whether the cells express the homeodomain transcription factors SIX3 or IRX3 [71] . The situation in the spinal cord is likely similar , with transcription factors such as IRX3 and PLZF not only influencing levels of FGFR expression , but potentially also the manner in which FGF signals are interpreted . Our data , together with previous studies , further suggest that FGF signaling may utilize distinct transduction pathways at different times in development . During the processes of neural induction , neural tube formation , and early progenitor patterning , FGFs are associated with robust activation of the ERK/MAPK pathway ( Z . B . G . unpublished observations ) [47] , [51] , [66] . However , during the peak period of neurogenesis in the spinal cord and transition towards gliogenesis , we were unable to detect signs of ERK/MAPK activity even under conditions where constitutively activated FGFR1 or FGFR3 were expressed ( Figure S7A–D ) . Rather , FGFR activation appeared to stimulate the STAT3 pathway . STAT3 forms a prominent node in multiple receptor tyrosine kinase and cytokine signaling pathways , and its activation can result in a wide range of effects including NPC maintenance and gliogenesis [72] . During early CNS development , STAT3 promotes SOX2 expression , and disruption of its activity can impair the emergence of NESTIN+ NPCs from embryonic stem cells differentiated in vitro [56] . Our data suggest that the ability of STAT3 to regulate SOX2 might be similarly utilized by PLZF and FGFR3 in the spinal cord . Later in development , STAT3 activity falls under the control of additional inputs , most notably the CNTF signaling pathway , and its function plays a critical role in regulating the onset of astrocyte differentiation [55] . It is notable that the PLZF-expressing progenitors in the intermediate spinal cord are ultimately fated to give rise to astrocytes , raising the possibility that the early employment of STAT3 for progenitor maintenance may predispose those progenitors to assume an astroglial fate at later time through the continued activation of the STAT3 pathway . Thus , PLZF regulates a downstream response to FGFs signaling distinct from the earlier role of FGFs promoting the rapid proliferation of the neural tube . This result suggests more nuance in FGF signaling than previously appreciated , and is reminiscent of recent studies showing that specific second messenger effectors mediate the diverse activities of the BMPs while establishing dorsal spinal circuitry [73]–[75] . In summary , PLZF and FGFR3 work in parallel with other FGFR , mitogen signaling pathways , and , most likely , other members of the BTB/POZ family , to modulate proliferation in the spinal cord and thereby permit NPCs to differentiate at characteristic rates and times in development . PLZF facilitates the mitogenic activity of the FGFs , which act in a STAT3-dependent manner to maintain a specific population of NPCs in a proliferative state and ensure that the necessary number of progenitors is available for the transition from neurogenesis to gliogenesis . Aberrant activation of FGFR3 and STAT3 has been observed in a multitude of human cancers [54] , [76] , [77] . The identification of PLZF as a critical regulator of FGFR3 and STAT3 activity thus provides important new insights into the mechanisms by which such tumors could arise and offers a novel therapeutic target .
Plasmid expression vectors were generated by cloning cDNAs of interest into a Gateway cloning-compatible variant of the vector pCIG [24] , [78] as follows: PLZF , full-length chick clone isolated by PCR from e4 chick cDNA; EnR-PLZF and VP16-PLZF were created by respectively fusing either the Drosophila Engrailed repressor domain [43] or the herpes simplex VP16 transactivation domain [44] to aa 300–665 of chick PLZF; FGFR3 , WT form of the human FGFR3 [79]; caFGFR3 , myristoylated and constitutively activated ( K650E ) form of the human FGFR3 cytoplasmic domain ( aa 399–806 ) [52]; STAT3-C , mouse STAT3 containing two activating mutations ( A662C , N664C ) [59] obtained from Addgene; and dnSTAT3 was created by incorporating into the mouse STAT3 nonphosphorylatable Y705F mutant [80] , obtained from Addgene , an additional H332Y mutation that disrupts DNA binding [81] . Sustained misexpression vectors were created using the Tol2kit system [82] . Briefly , Multi-Site Gateway Technology ( Invitrogen ) was used to transfer the CMV enhancer/β-actin promoter , the gene of interest , and an IRES-GFP reporter into the pDestTol2pA2 vector , which contains recognition sites for Tol2 transposase that permits stable integration into the chick genome [46] . The following expression vectors were also used in the experiments: RCAS-activated FGFR1 [61] , [83] , pCMX-FGF8 [61] , [84] , and pCAGGS-T2P2 ( Tol2 transposase ) [46] . PLZF shRNA vectors were created by subcloning target sequences against the chick PLZF transcript ( 5′-cgcagctgagatcctagaaata-3′ and 5′-ttcagcctgaagcaccagctgg-3′ ) into the plasmid pCIG-shRNA [9] , [24] . STAT3 activity was measured by transfection of the reporter vector BGZA-4m67-STAT3 containing four STAT3 binding sites driving the expression of a LacZ reporter [58] . Fertilized chicken eggs were acquired from AA Lab Eggs , Inc . and McIntyre Poultry and Fertile Eggs . Eggs were incubated at 37°C and 60% humidity , staged , and electroporated with plasmid vectors as previously described [9] , [69] . Unless otherwise indicated , embryos were electroporated at e3 ( HH 17 ) and collected at e5 ( HH 25 ) . Olig2Cre [85] and Olig2GFP [22] knock-in mice were maintained as previously described and interbred to produce Olig2 mutant embryos . The Luxoid mouse strain deficient for PLZF was rederived from cryopreserved embryos purchased from the Jackson Laboratory ( Strain Name B6 . C4-Zbtb16Lu/J ) . Mice were maintained in accordance with the guidelines specified by the UCLA Chancellor's Animal Research Committee . Tissue was collected , fixed , and cryosectioned prior to immunohistochemical staining or in situ hybridization as described previously [61] , [69] . Specific antibodies and in situ probes are described in Tables S1 and S2 . Dissected tissue was briefly fixed in 4% paraformaldehyde at 4°C , rinsed repeatedly in PBS containing 2 mM MgCl2 , equilibrated overnight in 30% sucrose , frozen on crushed dry ice in OCT mounting media ( Sakura Tissue-Tek ) , and cryosectioned . Prior to staining , slides were fixed in 4% paraformaldehyde for an additional 10 min at 4°C and then rinsed twice in PBS containing 2 mM MgCl2 , for 10 min per wash . Slides were overlaid with 1 mL of X-Gal Staining Buffer ( 1 mg/mL X-GAL [5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside] , 35 mM potassium ferrocyanide , 35 mM potassium ferricyanide , 0 . 02% NP-40 , 2 mM MgCl2 , in PBS ) and placed in a humidified chamber at 37°C for several hours to overnight . Once signal had developed , slides were repeatedly rinsed in PBS with 2 mM MgCl2 , coverslipped , and imaged using brightfield microscopy . All images were collected using either a Zeiss Observer D1 microscope equipped with an Apotome optical imaging system or a Zeiss LSM5 Exciter confocal imaging system . Images were processed using Zeiss Axiovision and LSM Exciter software suites and Adobe Photoshop . Pixel intensity analysis of mRNA and protein expression was performed using NIH ImageJ software . Cell counts were performed manually and in most cases represented as the mean value of multiple images of tissue sections collected from several independent specimens as indicated in the figure legends . For in ovo electroporation experiments , all transfected cells were counted for each image analyzed , with , on average , 100 cells per image in experiments focusing on the ventral region of the spinal cord at e4 ( HH 21 ) , more than 350 transfected cells per image in experiments with an e5 ( HH 25 ) endpoint , and over 900 cells per image in experiments with an e15 ( HH 41 ) endpoint . Unless stated otherwise , results are expressed as fractional changes normalized to electroporation with empty vector controls set to a value of 1 . 0 . The statistical significance of differences observed between experimental and control groups were assessed with the Student's t test using GraphPad Prism 5 . 0 software .
|
The embryonic spinal cord is organized into an array of discrete neural progenitor domains along the dorsoventral axis . Most of these domains undergo two periods of differentiation , first producing specific classes of neurons and then generating distinct populations of glial cells at later times . In addition , each of these progenitors pools exhibit marked differences in their proliferative capacities and propensity to differentiate to produce the appropriate numbers and diversity of neurons and glia needed to form functional neural circuits . The mechanisms behind this regional control of neural progenitor behavior , however , remain unclear . In this study , we identify the transcription factor Promyelocytic Leukemia Zinc Finger ( PLZF ) as a critical regulator of this process in the chick spinal cord . We show that PLZF is initially expressed by all spinal cord progenitors and then becomes restricted to a central domain , where it helps to limit the rate of neuronal differentiation and to preserve the progenitor pool for subsequent glial production . We also demonstrate that PLZF acts by promoting the expression of Fibroblast Growth Factor ( FGF ) Receptor 3 , thereby enhancing the proliferative response of neural progenitors to FGFs present in developing embryos . Together , these findings reveal a novel developmental strategy for spatially controlling neural progenitor behavior by tuning their responsiveness to broadly distributed growth-promoting signals in the embryonic environment .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
PLZF Regulates Fibroblast Growth Factor Responsiveness and Maintenance of Neural Progenitors
|
Cysteine proteinases of Fasciola hepatica are important candidates for vaccine antigens because of their role in fluke biology and host-parasite relationships . In our previous experiments , we found that a recombinant cysteine proteinase cloned from adult F . hepatica ( CPFhW ) can protect rats against liver fluke infections when it is administered intramuscularly or intranasally in the form of cDNA . We also observed considerable protection upon challenge following mucosal vaccination with inclusion bodies containing recombinant CPFhW produced in Escherichia coli . In this study , we explore oral vaccination , which may be the desired method of delivery and is potentially capable of preventing infections at the site of helminth entry . To provide antigen encapsulation and to protect the vaccine antigen from degradation in the intestinal tract , transgenic plant-based systems are used . In the present study , we aimed to evaluate the protective ability of mucosal vaccinations of 12-week-old rats with CPFhW produced in a transgenic-plant-based system . To avoid inducing tolerance and to maximise the immune response induced by oral immunisation , we used the hepatitis B virus ( HBV ) core protein ( HBcAg ) as a carrier . Animals were immunised with two doses of the antigen and challenged with 25 or 30 metacercariae of F . hepatica . We obtained substantial protection after oral administration of the plant-produced hybrids of CPFhW and HBcAg . The highest level of protection ( 65 . 4% ) was observed in animals immunised with transgenic plants expressing the mature CPFhW enzyme flanked by Gly-rich linkers and inserted into c/e1 epitope of truncated HBcAg . The immunised rats showed clear IgG1 and IgM responses to CPFhW for 4 consecutive weeks after the challenge .
Infection with Fasciola hepatica , a liver fluke , is one of the most significant veterinary problems due to the worldwide distribution of this parasite and a wide spectrum of host organisms [1] . Fasciolosis causes economic losses of US$3 billion annually due to its impact on livestock production , thereby affecting the food industry worldwide [2 , 3] . In recent years , the number of F . hepatica infections has dramatically risen , a trend that has been attributed to climate change [4 , 5] . The prevalence of fasciolosis has increased by up to 12-fold in the EU member states during recent years [6] . Human fasciolosis caused by F . hepatica is recognised by WHO as an important emerging but neglected tropical disease , with estimates of 2 . 4–17 million people infected worldwide , and approximately 180 million living at risk of infection [7 , 8] . Large endemic areas have been described in Peru [9 , 10] , Egypt [11] , Iran [12 , 13] , North America [14] , Pakistan [15] and other regions , with prevalences of 72–100% in the Bolivian Altiplano [16] . Currently , the treatment of fasciolosis is based primarily on the use of chemotherapy [17] . As F . hepatica drug resistance becomes more frequent , it is possible that there will be more cases of infection in humans with drug-resistant F . hepatica , which poses a real problem for the treatment of human fasciolosis [18] . The emergence of drug-resistant parasites [19–23] , combined with the growing consumer concern over chemical residues in food and their passage into the environment , have prompted the need for novel means of disease control [24] . The most effective method of parasite control is vaccination [3 , 25] . It is strongly believed that the control of the human infection would greatly benefit from vaccines targeting the animal infection [26] . A number of experimental , parenterally administered vaccines against fasciolosis have shown that the development of a successful commercial vaccine still remains a challenge [26–29] . There are two primary issues that need to be addressed when developing a vaccine . The first is the selection of an appropriate vaccine antigen . Several immunogenic fluke antigens have been identified; among them , the most promising appears to be a cysteine proteinase ( cathepsin L ) [27 , 30–32] . We previously showed that the cDNA of this F . hepatica cysteine proteinase , administered intramuscularly or intranasally , induces a protective immune response when delivered prior to the infection with fluke metacercariae ( mc ) [33 , 34] . We also observed considerable protection and reduced pathology upon challenge following parenteral and mucosal vaccination with inclusion bodies containing recombinant cysteine proteinase cloned from adult F . hepatica ( CPFhW ) and produced in Escherichia coli [35 , 36] . Another important aspect of vaccine development is the route of antigen administration . As the intestinal tract is the location where invasion of the fluke begins , it is postulated that the host protective response should occur in the intestinal mucosa-associated lymphoid tissue . It has previously been shown that challenge infections in immunised rats are rejected at the level of the gut and peritoneum in the first few days after infection [37–39] . Once the fluke reaches the liver and bile ducts , it seems to be impervious to protective immune mechanisms [40] . Therefore , oral vaccination is the desired route of delivery to prevent infections at the site of pathogen entry . Unfortunately , vaccines administered into the intestinal tract may be ineffective due to the rapid degradation of the vaccine antigens by intestinal proteases . It has been speculated that the plant cell wall delays the digestion of plant-produced and delivered antigens [41] . Therefore , oral vaccination is the desired route of antigen delivery , and plant-based delivery vehicles may increase the amount of antigen presented to the gut-associated lymphoid tissue [42] . Additionally , production of antigen in plants for oral vaccination eliminates the need for purification , cold storage , transportation and sterile delivery . In our previous study [43] , we reported that feeding mice with lettuce expressing the cysteine proteinase from F . hepatica is effective in inducing a specific antibody response against this antigen . The aim of the present study was to evaluate the immunogenicity and protective ability of various modifications of plant-produced F . hepatica cysteine proteinase in oral vaccination . As a carrier for the F . hepatica antigen , we used hepatitis B virus ( HBV ) core protein ( HBcAg ) . The carrier was applied to avoid tolerance and to potentiate the immune response induced by oral immunisation . We showed that substantial protection can be obtained when the plant-expressed hybrid proteins are orally administered .
All experimental protocols were approved by the III Local Animal Experimentation Ethics Committee of Warsaw University of Life Sciences , approval number: 39/2003 . The experiments were performed according to the guidelines of European Communities Council Directive ( 86/609/EEC ) . All efforts were made to minimise animal suffering and to reduce the number of animals used . F . hepatica mc were obtained from experimentally infected intermediate hosts , a laboratory strain of Galba truncatula snails reared at W . Stefański Institute of Parasitology . Miracidia were cultured from the F . hepatica eggs obtained from the gall bladders of naturally infected cattle slaughtered in the slaughterhouse . Liver and gall bladders were classified as waste because of the presence of pathological changes caused by the fluke invasion . There was no need to apply for permission to use them . Each snail was exposed to two recently hatched miracidia for approximately 24 h at room temperature . After exposure , the snails were maintained in Petri dishes and fed Oscillatoria algae . Mc were collected starting from day 70 after the exposure of the snails to miracidia . To collect the mc , the infected snails were placed in Petri dishes lined with transparent cellophane and exposed to light for up to 2 h . The mc were stored in water at 4°C for at least 2 weeks before use . Inbred 12-week-old Sprague-Dawley ( SPRD/Mol/Lod ) male rats were used in the experiments . The animals were housed in groups and randomly assigned to the treatment or control groups . The rats were acclimated for 1 week before the experiments were initiated . They were provided with food and water ad libitum until 16 h before being fed the antigen and challenged with mc , during which period they were deprived of food [39] . Three separate experiments were carried out using various CPFhW-based antigens expressed in plants ( Table 1 ) . In all these experiments , two doses of transgenic lettuce were intragastrically administered to the vaccinated rats in 4-week intervals . For each vaccination , 1 g of lyophilised lettuce was used , which corresponds to 10 μg of the antigen [43] . In each experiment , the group of control rats was mock-immunised with the lyophilised control lettuce according to the same schedule as that used for the experimental animals . Twenty-eight days after the second antigen dose , each rat was orally challenged with mc of F . hepatica in 1 ml of water . After administration , the syringe and cannula were flushed with 0 . 5 ml of water to recover any remaining mc . In Experiment I , 25 mc were used . In Experiments II and III , the viability of the metacercariae was lower . Therefore , in order to maintain the experimental conditions , the number of metacercariae was increased proportionally to the number of that observed for in vitro excystment of this F . hepatica isolate . In Experiments I and II , all rats were euthanised and dissected five weeks after the challenge infection . In Experiment III , four rats from each group were euthanised and dissected on days 7 , 35 and 63 after the challenge . Livers were removed for the evaluation of macroscopic lesions namely the hepatic fibrosis caused by the migration of F . hepatica through the liver parenchyma . Hepatic damage due to the invading parasite was evaluated subjectively by observing macroscopic alterations in the organ based on a number of criteria , including the following: colour change to greyish-white , increase in size , change in consistency , dilatation and thickening of bile ducts , and formation of surface scars [44] . The degree of lesions observed ( index of liver damage ) was summarised semiquantitatively using the following scale to express the intensity and extent of the alteration ( tissue necrosis or liver nodules ) observed: “0” , no visible sign of tissue necrosis or liver nodules; “1” , mild liver necrosis; “2” , moderately mild liver damage of up to 15% of the liver surface; “3” , moderate liver damage—approximately 30% of the liver surface; “4” , intense liver damage of up to 50% of the liver surface; “5” , severe liver necrosis with >50% of the liver surface showing pathological changes . Index of liver damage and the intensity of the fluke invasion ( no . of flukes found during the dissection ) was estimated on days 35 and 63 after the challenge . To recover flukes from the parenchyma and biliary tree , the livers were stored at 37°C in separate Petri dishes containing RPMI-1640 cell culture medium . Blood , peritoneal fluid samples , and mesenteric lymph nodes were also collected in Experiment III to compare the cellular and antibody responses of vaccinated and control rats during the challenge infection . The peritoneal and mesenteric lymph nodes from each rat were collected aseptically . Each tissue was immersed in RPMI-1640 cell culture medium ( 4°C ) supplemented with 2% heat-inactivated foetal bovine serum . The tissues were cut into 1 mm3 pieces with sterile scalpels . The tissue fragments were placed in a Petri dish containing sterile wash medium on ice . The suspended cells were then washed and quantified using a haemocytometer . Cell viability was determined by trypan blue exclusion . The counted cells were centrifuged for 5 min at 1200 rpm and 4°C , followed by resuspension in PBS containing 0 . 05% sodium azide and incubation on ice for 15–30 min . The cells were pelleted by centrifugation and resuspended . For these and peripheral blood samples , the quantities of eosinophils and monocytes and the phenotype of the T cells ( CD4+ and CD8+ ) were investigated by using a panel of monoclonal anti-rat antibodies ( BD Pharmingen ) . Monoclonal antibodies against rat CD4 ( clone: OX-38 ) receptors were labelled with phycoerythrin , and the CD8 antibodies ( clone: OX-8 ) were labelled with fluorescein isothiocyanate . The cells incubated with corresponding isotype control ( BD Pharmingen ) mouse IgG2a , κ labelled with phycoerythrin ( clone: G155-178 ) and mouse IgG1 , κ labelled with fluorescein isothiocyanate ( clone: MOPC-31C ) antibodies were used as controls for nonspecific antibody binding to the cells . For FACS analysis , single-cell suspensions ( 50 μl ) were incubated with mAbs and washed . Subsequently , red blood cells were lysed in FACS Lysing Buffer ( Becton Dickinson ) , and leukocytes were analysed using a FACSCalibur flow cytometer ( Becton Dickinson ) with an argon excitation source . Data acquisition was performed using CellQuest software ( Becton Dickinson ) . The results were expressed as the percentage of total mononuclear cells from a designated region . Leukocytes were identified by their characteristic appearance on a dot plot of FSC versus SSC and electronically gated to exclude platelets and dead-cell debris . Eosinophils are autofluorescent , and this property was used to identify them . Lymphocytes were selected using fluorescence-labelled antibody specific for the antigen as well as their phenotypic and morphometric features . Microtiter plates ( MaxiSorb ) were coated with fluke CPFhW [35] or ES [34] antigen ( 15 μg/ml of 0 . 05 M carbonate-bicarbonate buffer at pH 9 . 6 ) . The plates were incubated overnight at 4°C and subsequently washed four times with 10 mM Tris/0 . 15 M NaCl at pH 7 . 4 ( TBS ) containing 0 . 05% Tween 40 . The excess binding sites were blocked by washing with 100 μl/well of TBS and 4% skimmed milk for 2 h at room temperature . The serum samples ( diluted 1:100 ) were added to each well , and the plates were incubated for 30 min at 37°C . Each serum sample was tested in triplicate . Following three washes , HRP-conjugated anti-rat IgG1 , IgA , and IgM monospecific antisera ( Bio-Rad formerly AbD Serotec ) were added to each well , and the plates were incubated for 30 min at 37°C . The plates were washed three times , and the binding of the conjugates was visualised with 3 , 3 , 5’ , 5’-tetramethylbenzidine in 0 . 1 M sodium citrate , pH 4 . 5 , containing 0 . 03% H2O2 . The reaction was stopped , and the absorbance was measured at 405 nm using an MRX ELISA Reader ( Dynatech Laboratories ) . For the detection of the IgE antibodies , the sera were diluted 10-fold , and a monoclonal mouse antibody specific to rat IgE ( ICN Immunobiologicals ) was used [45] . Peroxidase-labelled anti-mouse Ig ( ICN Immunobiologicals ) was used as the secondary antibody . The data are expressed as the mean ± the standard deviation for each experimental group . One-way ANOVA with parametric F-test was used to compare the results of the cell count and antibody levels between groups after the challenge in Experiment III . Mann-Whitney U-tests were used to compare the number of flukes recovered from vaccinated and challenge control rats . p-values < 0 . 05 were considered statistically significant . The percent protection for vaccinated animals was calculated as ( 1 − V/C ) × 100 , where C is the mean burden of the control animals challenged with mc , and V is the mean burden of the immunised rats challenged with mc [46] . Genetic sequences used in this study are deposited in GenBank with accession numbers: CPFhW , AY277628; HBcAg , Z35716; Ubiquitin , X05731 .
Three experiments were conducted to compare the efficacy of oral administration of lyophilised lettuce containing the antigen variants . The worm burdens and indexes of liver damage following the challenge are shown in Table 2 . In all conducted experiments ( I , II , III ) , a profound reduction in the fluke burden was found in the vaccinated rats . In addition , the livers of the vaccinated rats were not as damaged as those of the challenge controls . Among the used antigens , the mature CPFhW enzyme fused to HBcAg ( T ) was found to provide the highest degree of protection . In Experiment I , the highest degree of protection ( 64% ) was observed in rats orally fed lyophilised lettuce expressing the sequence encoding the mature CPFhW enzyme fused to HBcAg ( T ) ( Group 1 ) . Rats immunised with lettuce expressing the propeptide of CPFhW enzyme ( Group 2 ) showed significantly ( p<0 . 05 ) lower protection than observed in the group immunised with the mature CPFhW ( Group 1 ) . In Experiment II , 65 . 4% of protection was observed in the rats vaccinated with lettuce expressing HBcAg ( T ) with an insertion encoding the mature CPFhW enzyme flanked by Gly-rich linkers ( mCPFhW::G::C ) . In the next Experiment ( Experiment III ) , where a new batch of plants expressing the same antigen ( mCPFhW::G::C ) was used , the level of protection was slightly lower ( 62 . 5% ) than in Experiment II . Lower protection was observed when the rats were immunised with lettuce transformed with mature CPFhW enzyme not fused to HBcAg ( T ) ( Group U ) . When the rats were immunised with lettuce transformed with the U::mCPFhW construct ( Experiment III , Group U ) , the observed protection level was lower than when the construct containing mature CPFhW fused with HBcAg ( T ) was used ( Experiment I Group 1 , Experiment II Group 1 and Experiment III Group G ) . The humoural response to F . hepatica antigens was examined by comparing the antibody OD values of the control and vaccinated groups using one-way ANOVA . The levels of antibodies against CPFhW and ES antigens were measured in the sera of experimental rats after the challenge infection . All serum samples were tested in triplicate . The results are presented in Figs 2 and 3 . Rats in group G showed a significantly higher IgG1 response than the rats in groups U and C at weeks 1 and 5 after the challenge infection ( Fig 2A ) . After the challenge , the CPFhW-specific IgM antibodies in both vaccinated groups showed significantly ( p<0 . 05 ) higher OD values than the antibodies in the challenge controls ( Fig 2B ) . No statistically significant difference between the vaccinated and challenge control groups were observed with respect to the CPFhW-specific IgA and IgE levels ( Fig 2C and 2D , respectively ) . IgM and IgA targeting ES antigens were present at significantly ( p<0 . 05 ) higher levels in the vaccinated rats than in the challenge controls ( Fig 3B and 3C ) . The challenge with F . hepatica mc caused remarkable changes in the numbers of eosinophils , monocytes , CD4+ cells and CD8+ cells in both the vaccinated and challenge control rats ( Figs 4–6 ) . Significantly ( p<0 . 05 ) increased numbers of blood eosinophils were observed in the vaccinated rats and challenge controls at 5 weeks after challenge and in groups G and C at 9 weeks after challenge ( Fig 4A ) . Similar cellular response dynamics were observed for the peritoneal fluid ( Fig 5A ) . Blood monocytes were present at similar levels in the vaccinated and control rats at weeks 1 and 5 after challenge , but at the end of the experiment , the rats in group C had the highest numbers of these cells ( Fig 4B ) . In the peritoneal fluid , the numbers of these cells were significantly ( p<0 . 05 ) higher than that in the challenge controls at the first and fifth weeks after the challenge infection ( Fig 5B ) . CD4+ and CD8+ lymphocytes appeared in the blood of the vaccinated rats in significantly ( p<0 . 05 ) higher numbers than in controls on day 35 after infection . In contrast , at week 9 , the highest level of these cells was found in the animals of group C ( Fig 4C and 4D ) . In the peritoneal fluid , the highest number of CD4+ cells was found in rats in group G on day 7 post-infection; however , the number of CD8+ cells at the same time was significantly ( p<0 . 05 ) lower than in group U ( Fig 5C and 5D ) . At the end of the experiment , the opposite situation was observed . The peritoneal fluid from both vaccinated and challenge control rats contained significantly fewer CD4+ and CD8+ cells than the peritoneal fluid from uninfected control rats on day 35 post-infection ( Fig 5C and 5D ) . Mesenteric lymph nodes showed no CD4+ response on day 7 after the challenge ( Fig 6A ) , whereas the CD8+ cells reached the highest level in group U at that time ( Fig 6B ) .
Since the early 1990s , edible vaccines based on transgenic plants have been investigated . Transgenic plants offer several features that make them an attractive platform for the delivery of oral vaccine antigens [48] . When an antigen is produced in edible plants or parts thereof , the production cost of a vaccine can be reduced considerably , thus increasing its availability and usability for both humans and animals [42] . Furthermore , the production of vaccines in plants eliminates the risk of contamination with animal pathogens such as viruses and prion proteins [49] . Transgenic plants serve as bioreactors for antigen production and provide a natural built-in system for antigen encapsulation [50] . The data from the literature suggest that when administered orally , the antigen immunogenicity and biological activities are preserved in the harsh environment of the gastrointestinal tract due to their natural bioencapsulation in the plant cell . It has been demonstrated that bioencapsulation influences the prolonged release and presentation of the antigen to immune responsive sites [51] . Once these vaccines pass through the gastric environment and reach the small intestine , they are taken up by the M cells for the induction of mucosal and systemic immune responses [52 , 53] . The results presented in this study clearly indicate that substantial protection can be obtained when plant-produced heterologous proteins ( hybrids of CPFhW and HBcAg ( T ) ) are administered with food . The administration of durable powdered lyophilised plant tissue made possible to concentrate and standardise the plant-bioencapsulated antigen content in the vaccine doses . The HBV core protein ( HBcAg ( T ) ) was used as a carrier to enhance the immune response . This protein forms symmetrical structures , contains potent T helper epitopes and is able to activate B cell production of anti-HBcAg immunoglobulins both in the presence and absence of antigen-specific T cells . Therefore , the HBcAg ( T ) has gained interest as a carrier system to potentiate humoural and cellular immune responses against heterologous epitopes fused to this protein [47 , 54] . Foreign heterologous epitopes can be inserted at different positions within the HBcAg protein . It has been shown that it is possible to fuse HBcAg with large inserts without disrupting its ability to confer immunogenicity [55] . Ravin et al . [56] used HBcAg as a carrier for the ectodomain of influenza virus matrix protein 2 ( M2e ) . In their experiment , the M2e peptide was fused to the N-terminus of the HBcAg protein , and the chimeric protein was expressed in plants . This plant-produced antigen was highly immunogenic in mice , and the mice were protected against lethal influenza challenges . Fusion to both the N-terminus and the C-terminus is compatible with the assembly of the HBcAg particle and preserves its native antigenicity and immunogenicity . However , fusion to an immunodominant internal site of HBcAg ( c/e1 epitope ) reduces its antigenicity and immunogenicity , while dramatically enhancing the immunogenicity of the inserted foreign epitope [57] . In the construct mCPFhW::G::C , the DNA sequence of mature F . hepatica CPFhW enzyme was inserted into the c/e1 epitope of HBcAg ( T ) , which is located around amino acid 80 [58] . The original core protein amino acids were removed , including Asp-78 , Pro-79 and Ala 80 . To minimise steric constraints , Gly-rich linkers were added between the mature CPFhW enzyme and the remaining regions of HBcAg ( T ) . Gly-rich linkers were used because glycine has low preference to form an α-helix , and its lack of sidechain maximises the freedom of the polypeptide backbone conformation [59] . We observed the highest level of protection ( 65 . 4% ) after the administration of transgenic plants expressing this hybrid protein; however , in rats vaccinated with hybrid protein ( mCPFhW::HBcAg ( T ) ) lacking the Gly-rich linkers , the protection level was only marginally lower ( 64% ) . The mature cysteine proteinase is an enzyme that is secreted by F . hepatica and with which the host organism has contact . This contact can explain why immunisation with this antigen resulted in the best protection in the challenged animals , especially in comparison to immunisation with the propeptide form of the CPFhW protein . Although the propeptides of cathepsin L-like proteins have been predicted to contain important B cell epitopes [60] , immunisation with this antigen variant resulted in lower protection values than immunisation with the mature protein . Obtaining a high expression level of recombinant proteins in transgenic plants is a well-known challenge . Therefore , attaining a high expression level of the recombinant protein is one of the primary design objectives when preparing DNA constructs . The data from the literature suggest that the presence of the ubiquitin promoter or ubiquitin gene can significantly augment the expression of the gene fused to ubiquitin [61] . The yield enhancement has been proposed to be due to a chaperone effect exerted by the ubiquitin . In our experiments , we used the U::mCPFhW construct , consisting of the ubiquitin sequence ligated to the 5' end of the sequence encoding the mature CPFhW , to potentiate the expression of the mature CPFhW . As a result , we obtained the expression of mature CPFhW as a translational fusion with ubiquitin . In our study , the expression of the mature CPFhW fused to ubiquitin allowed for the accumulation of the mature CPFhW to approximately 20 μg/g wet weight . This was the highest CPFhW expression level obtained among all the DNA constructs presented in this study . In eukaryotic organisms , ubiquitin is cleaved out from the fused protein in vivo [62]; hence the antigen administered in the plant material that was transformed with the U::mCPFhW construct was the mature CPFhW alone . In the experiments where CPFhW was used for immunisation , we observed a 50% protection . It is a lower level of protection in comparison with that obtained in experiments where the mature CPFhW fused to HBcAg was used . These results support the use of HBcAg as a carrier for foreign antigens to potentiate protection in the vaccinated animals . In Experiment III , we observed that the oral vaccination of Sprague-Dawley rats with the lyophilised transgenic lettuce expressing the mCPFhW::G::C or the U::mCPFhW led primarily to a Th2 antibody response against the metacercarial challenge . These enterally immunised rats showed clear IgG1 and IgM responses to CPFhW for 4 consecutive weeks after the challenge , with an increase in the level of IgG1 specific to CPFhW . Clear IgG1 , IgM and IgA antibody responses against the ES antigens were visible in all the challenged rats , but the IgM and IgA antibody levels were higher in the vaccinated animals . Also in our previous study [63] , we observed significant IgM and IgA antibody responses in immunised rats . Tliba et al . [64] have reported the persistence of a high level of IgM on the teguments of F . hepatica isolated from rat livers between weeks 1 and 8 post-infection . They speculated that the IgM deposition on the fluke’s tegument might inhibit eosinophil access to the parasite , which would enable it to avoid the antibody-dependent cell-mediated cytotoxicity . Tliba et al . [64] had also observed the presence of IgA and IgE antibodies on the flukes and in the liver tissue surrounding parasites , albeit to a smaller extent . Van Milligen et al . [39 , 65] have reported that the protective immune responses in immunised rats kill juvenile flukes within the gut wall and peritoneal cavity , and a much smaller percentage of fluke survive gut wall penetration and migration through the peritoneal cavities of immune rats when compared to naïve rats . It has also been shown that in resistant rats , newly excysted juveniles entering the peritoneal cavity are coated with antibodies as well as eosinophils , neutrophils , macrophages and mast cells [66] . This suggests that resistance to flukes in rats may involve both an effective antibody response and lymphoid cells attacking juvenile flukes , with eosinophils and IgG1 correlating with protection at the gut wall [64 , 66] . We investigated the effectiveness of oral vaccination with plant-produced F . hepatica CPFhW in rats . Rats are often used as a model to study immunity in cattle due to the fact that the course of infection is similar in both of these animals . In rats , as well as in cattle , infection with F . hepatica mc causes partial protection to a challenge infection [67–69] . Therefore , the results reported in this study on the oral vaccination with plant-produced F . hepatica CPFhW are the basis for the development of similar vaccines against F . hepatica infection in large ruminants . However , some important issue to consider are how the differences in the ruminant gut might affect the vaccine and whether the plant-bioencapsulated antigen would still be viable once the plant material reaches the intestine . Although there are numerous studies reporting immunogenicity of orally delivered plant-made vaccines in humans and small animal models , only a few have successfully demonstrated their efficacy in ruminants [70–74] . The concept of transgenic plant-based vaccines has been successfully employed by Pelosi et al . in their study on antigen-specific antibody responses against a model antigen ( the B subunit of the heat labile toxin of enterotoxigenic E . coli ) in sheep following oral immunisation with plant-made and delivered vaccines . The delivery resulted in antigen-specific immune responses in mucosal secretions of the abomasum , small intestine and mesenteric lymph nodes . These findings suggest that the orally administered plant-bioencapsulated antigens are still viable after the passage through the rumen and elicit mucosal and systemic immune responses in sheep [70] . It has been suggested that the loss of production in cattle is significant once the Fasciola infection levels establish above 30–40 flukes [75 , 76] . In sheep , production losses are observed with fluke burdens ranging from 30 to 54 flukes [75 , 77] . Taking these data into account , it appears that a fluke vaccine would need to reduce the fluke burden below these thresholds to ensure sustainable production benefits . This indicates that the vaccine efficacy required to reduce fluke burdens in sheep and cattle below the threshold varies from approximately 50% to 80% [29] . It has been suggested that due to the emergence of drug-resistant parasites [19–23] , combined with the growing consumer concern over chemical residues in food and their passage into the environment , it may be commercially feasible to introduce a fluke vaccine with suboptimal but reasonable efficacy ( 50% ) . Such a vaccine would provide economic benefits for dairy producers because it would not compromise fluke control during lactation , and it would leave no residues in the milk [29] . In light of these findings , oral vaccination against F . hepatica infection with plant-produced antigens is a viable approach due to its convenience and high efficacy without adjuvants .
|
Infection with Fasciola hepatica , a liver fluke , is one of the most significant veterinary problems due to the worldwide distribution of this parasite , a wide spectrum of host organisms and the resulting economic loss . Human fasciolosis caused by F . hepatica is recognised by the World Health Organization as an important emerging but neglected tropical disease . Development of an effective vaccine against this disease is becoming a priority , especially as the appearance of drug-resistant strains undermine the currently employed drug-based treatments . The two primary issues when developing a vaccine are the selection of an appropriate vaccine antigen and the route of antigen administration . In our studies , we use one of the F . hepatica cysteine proteinases , which are promising antigens for vaccine construction . We evaluate the immunogenicity and protective ability of various modifications of this cysteine proteinase produced in plants . We show that substantial protection can be obtained when plant-expressed hybrid proteins are administered orally .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"chemical",
"compounds",
"enzymes",
"immunology",
"enzymology",
"organic",
"compounds",
"vaccines",
"preventive",
"medicine",
"amino",
"acids",
"infectious",
"disease",
"control",
"antibodies",
"plants",
"flowering",
"plants",
"vaccination",
"and",
"immunization",
"lettuce",
"cysteine",
"public",
"and",
"occupational",
"health",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"vegetables",
"proteins",
"antigens",
"chemistry",
"immune",
"response",
"sulfur",
"containing",
"amino",
"acids",
"biochemistry",
"organic",
"chemistry",
"physiology",
"biology",
"and",
"life",
"sciences",
"proteases",
"physical",
"sciences",
"organisms"
] |
2017
|
Immune response of rats vaccinated orally with various plant-expressed recombinant cysteine proteinase constructs when challenged with Fasciola hepatica metacercariae
|
Experimental studies have shown that one key factor in driving the emergence of drug resistance in solid tumors is tumor hypoxia , which leads to the formation of localized environmental niches where drug-resistant cell populations can evolve and survive . Hypoxia-activated prodrugs ( HAPs ) are compounds designed to penetrate to hypoxic regions of a tumor and release cytotoxic or cytostatic agents; several of these HAPs are currently in clinical trial . However , preliminary results have not shown a survival benefit in several of these trials . We hypothesize that the efficacy of treatments involving these prodrugs depends heavily on identifying the correct treatment schedule , and that mathematical modeling can be used to help design potential therapeutic strategies combining HAPs with standard therapies to achieve long-term tumor control or eradication . We develop this framework in the specific context of EGFR-driven non-small cell lung cancer , which is commonly treated with the tyrosine kinase inhibitor erlotinib . We develop a stochastic mathematical model , parametrized using clinical and experimental data , to explore a spectrum of treatment regimens combining a HAP , evofosfamide , with erlotinib . We design combination toxicity constraint models and optimize treatment strategies over the space of tolerated schedules to identify specific combination schedules that lead to optimal tumor control . We find that ( i ) combining these therapies delays resistance longer than any monotherapy schedule with either evofosfamide or erlotinib alone , ( ii ) sequentially alternating single doses of each drug leads to minimal tumor burden and maximal reduction in probability of developing resistance , and ( iii ) strategies minimizing the length of time after an evofosfamide dose and before erlotinib confer further benefits in reduction of tumor burden . These results provide insights into how hypoxia-activated prodrugs may be used to enhance therapeutic effectiveness in the clinic .
Solid tumor vasculature is characterized by a disorganized , aberrant network structure of tortuous , hyperpermeable blood vessels [1] . These characteristics lead to nonuniform spatial distributions of drug and oxygen ( as well as other nutrients and growth factors ) throughout tumors , which in turn have been implicated in the emergence and evolution of resistance [2–7] . Indeed , several recent studies have demonstrated that the presence of spatial gradients of drug in an environment can accelerate the emergence of antibiotic resistance in bacteria [8 , 9] . One explanation for this phenomenon is that regions of low drug concentration generate local niches where sustained cell proliferation drives the production of new genetic variants . These spatial regions often coincide with hypoxic ( low oxygen ) conditions where drug-resistant variants may possess a survival advantage over drug-sensitive cells [2 , 10–13] , thus enabling the establishment of stable pockets of drug resistance in tumor regions not easily accessible by drugs . In light of these observations , one strategy proposed is to design therapy regimens that exploit the interaction between tumor cell populations and their environments to achieve long-term tumor eradication or control . Hypoxia is defined as reduced levels of molecular oxygen ( typically less than 1% ) in tissue . In contrast , ambient air exists at approximately 21% and most human organs have oxygen levels in the range of 2% to 9% [14] . The prevalence of hypoxic regions in solid tumors has led to the development of hypoxia-activated prodrugs ( HAPs ) , which are compounds designed to metabolize into active drugs upon entry into hypoxic environments [15–18] . For example , one such compound , evofosfamide , consists of a radical anion linked to a potent DNA-alkylating agent which penetrates effectively through tissues under normoxic ( normal oxygen ) conditions . Under hypoxic conditions , however , the radical anion undergoes irreversible fragmentation and releases the activated drug into the tumor [18 , 19] . This type of novel action allows evofosfamide to penetrate and target cancer cells within hypoxic region of a tumor , unlike standard therapies whose range is often confined to well-vascularized , normoxic regions . Currently , several HAPs are in clinical trials [20] . Tirapazamine was the first HAP to be tested in the clinic , in combination with both chemotherapy and radiotherapy; however , results did not show any significant therapeutic benefit . It was thought that off-target toxicity and insufficient tissue penetration were the primary contributing factors to this result . More recently , evofosfamide , which has a superior tissue-penetration ability , underwent Phase III testing in combination with chemotherapies for pancreatic cancer and soft tissue sarcoma . Neither of these trials demonstrated a significant survival benefit . However , given the response kinetics of hypoxic cancer cells to these therapies in preclinical studies , we hypothesize that the potential of HAPs has not yet been fully realized in previous clinical trials , and that mathematical modeling may be beneficial in identifying the combination treatment strategies that lead to survival benefit . We have demonstrated in previous work that identifying the right dosing schedule is important in improving cancer treatment outcomes , and further that mathematical modeling is an effective tool to help identify optimal administration strategies [21–23] . In particular , we demonstrated in [23] that altering the timing of treatment periods in sequential combination therapies may prevent or delay the emergence of drug resistance . This work builds upon a large body of literature on evolutionary modeling of drug resistance in cancer ( see , e . g . the review [24] and references therein ) . Here we discuss a few recent contributions to modeling combination therapies in cancer . For example , Komarova et . al . [25] utilized a stochastic birth-death process model to study the impact of combination therapies in Chronic Myeloid Leukemia , finding that a combination of two but not three drugs should be used in the prevention of drug resistance . In a later work by the same authors , Katouli et . al . [26] designed a general algorithm to compare combination treatment protocols in Chronic Myeloid Leukemia according to their cross-resistance properties , and to identify the protocols with the highest probability of treatment success . Most recently , Bozic et al . [27] developed a mathematical model that predicts responses to combinations of targeted inhibitors in melanoma patients . Using this model , the author predicted that combinations of two or three drugs will be far more effective than sequential treatment with the same agents , with the potential for complete cure . Here we will focus our modeling efforts on designing HAP-targeted combination treatment strategies for non-small cell lung cancer ( NSCLC ) . Erlotinib is a tyrosine kinase inhibitor commonly used to treat EGFR-mutant non-small cell lung cancer [28–30] . However , most patients develop resistance and disease progression within 12–18 months of starting treatment [31] . Consequently , novel approaches to prevent , or at least delay , the onset of resistance to erlotinib are of great clinical importance . A significant amount of research has been dedicated to improving treatment of non-small cell lung cancer . Several studies have shown that it may be beneficial to continue therapy with tyrosine kinase inhibitors such as erlotinib even after the point of disease progression [32–34] . Previous work has focused on the use of mathematical models of tumor growth and resistance during erlotinib treatment to optimize therapeutic strategies and minimize a patient’s risk of resistance [21 , 22] . Another approach that has been extensively studied is the use of combination therapy to mitigate resistance to erlotinib [23 , 35 , 36] . However , these studies lack a consideration for the heterogeneous oxygen and drug distributions throughout the tumor , and in particular their role in mediating tumor response to therapy and the emergence of drug resistance . Recently , we have demonstrated through modeling efforts that the consideration of heterogeneous tumor oxygenation and drug concentration reveals dramatically different treatment outcomes and evolutionary responses to therapy when compared to models under homogeneous environmental conditions [2] . In this work , we investigate the potential benefits of using a HAP in combination with standard therapy to prevent the emergence of resistance to erlotinib in non-small cell lung cancer . We design a stochastic mathematical model , with parameters informed by experimental data , to describe the evolutionary dynamics of a cancer cell population within a heterogeneous tumor microenvironment during treatment with erlotinib and evofosfamide . Using this model , we show that a combination treatment strategy results in treatment outcomes preferable to those resulting from monotherapy with either of these drugs alone . We also use a novel approach to define toxicity constraints for this combination therapy , which allows us to optimize treatment strategies over the space of tolerated dosing schedules using both drugs in order to minimize tumor burden and probability of resistance . Determining toxicity constraint profiles for combination therapies could have significant clinical implications as severe toxicity issues is one of the major reasons that HAP combinations have not been successful in clinical trials .
In order to model tumor evolution within an environment with heterogeneous oxygen and drug concentrations , we consider a stochastic population dynamic process in which the cell population is distributed amongst a series of habitats with varying oxygen and drug profiles . Oxygen concentration within the tumor decays exponentially as a function of distance from the nearest blood vessel . This decay rate is parameterized in the model based on estimates of the half-length away from the blood vessel [2] . This is used to define the oxygen concentration in each microenvironmental compartment; hence every compartment corresponds to a volume some distance from the nearest blood vessel . To estimate the relative contributions of each of these compartments to the tumor microenvironment , we utilize experimental data capturing relative frequencies of a spectrum of oxygen partial pressures throughout solid tumors [37] . We consider a total of 32 environmental compartments in which compartment i has an oxygen partial pressure of 2 . 5 ⋅ i mmHg to mirror this data . We then construct a mixture model of compartments in which the weighting of each compartment is determined based on the relative frequency of its corresponding oxygen partial pressure in the experimental profile . A schematic of this process is depicted in Fig 1 . Within each compartment , we use a multi-type , non-homogeneous , continuous-time birth-death process to model the population of cancer cells during treatment . We assume for now that the evolutionary dynamics within each microenvironmental compartment are independent . ( This assumption will be relaxed later , see section on Migration in S1 Text . ) The number of erlotinib-sensitive cells in compartment i at time t is denoted by Xi ( t ) , and the number of erlotinib-resistant cells in compartment i at time t is given by Yi ( t ) . The joint process Xi ( t ) = ( Xi ( t ) , Yi ( t ) ) represents the combined state of the sensitive and resistant cell populations in compartment i at time t . In compartment i , erlotinib-sensitive cells proliferate and die with rates λX , i ( t ) and μX , i ( t ) , respectively , while erlotinib-resistant cells proliferate and die with rates λY , i ( t ) and μY , i ( t ) . These birth and death rates reflect the effect of treatment on the cancer cell population in an environmental compartment i , and therefore depend on the concentrations of oxygen and both drugs found in that compartment at time t . During every sensitive cell division , a mutation may arise with some small probability u , giving rise to a new resistant cell . This mutation rate falls in the range between 10−8 and 10−6 [21 , 38 , 39] . Here we use u = 10−7 . We consider an initial population of M = 1 . 6 ⋅ 106 erlotinib-sensitive cancer cells and zero resistant cells . The number of sensitive cells Mi initially in compartment i is calculated using the relative compartment weights . The evolutionary dynamics within each microenvironmental compartment can be described using analytic approximations for the probability of resistance and means of the sensitive and resistant cell populations . The derivations of these analytic expressions are outlined elsewhere [40] . Because u is very small , we approximate 1 − u ≈ 1 in the following . Then the population of sensitive cells Xi ( t ) in compartment i can be described using a simple birth-death process . Hence the mean number of sensitive cells at time t in this compartment is E [ X i ( t ) ] = M i exp ∫ 0 t λ X , i ( τ ) - μ X , i ( τ ) d τ . ( 1 ) The mean number of resistant cells in this compartment at time t is E [ Y i ( t ) ] = ∫ 0 t b i ( τ ) exp ∫ 0 t - τ λ Y , i ( τ + η ) - μ Y , i ( τ + η ) d η d τ , ( 2 ) where bi ( t ) , the rate of production at time t of the resistant cells from the sensitive cell population in compartment i , is given by the formula b i ( t ) = M i exp ∫ 0 t λ X , i ( τ ) - μ X , i ( τ ) d τ λ X , i ( t ) u . Lastly , the probability of resistance in compartment i at time t is P [ Y i ( t ) > 0 ] = 1 - exp ∫ 0 t - b i ( T ) 1 - P i e x t ( T , t ) d T . ( 3 ) P i e x t ( T , t ) represents the probability that a group of resistant cells originating from a single resistant cell produced at time T in compartment i is completely extinct by time t , and is given by the formula P i e x t ( T , t ) = ∫ 0 t - T μ Y , i ( τ + T ) ω i ( τ , T ) d τ 1 + ∫ 0 t - T μ Y , i ( τ + T ) ω i ( τ , T ) d τ , where ω i ( τ , T ) = exp ∫ 0 τ μ Y , i ( η + T ) - λ Y , i ( η + T ) d η . Now we calculate the probability of resistance and means of the sensitive and resistant cell populations in the entire tumor at time t . We can obtain the means at time t of the sensitive cell population X ( t ) and the resistant cell population Y ( t ) in the entire tumor by summing over all compartments: E [ X ( t ) ] = ∑ i E [ X i ( t ) ] , E [ Y ( t ) ] = ∑ i E [ Y i ( t ) ] . Here , E [ X i ( t ) ] and E [ Y i ( t ) ] are given by Eqs ( 1 ) and ( 2 ) , respectively . The mean tumor size at time t is given by E [ X ( t ) + Y ( t ) ] = E [ X ( t ) ] + E [ Y ( t ) ] . The probability that there exists one or more resistant cells in compartment i at time t is P [ Y i ( t ) > 0 ] . Then 1 - P [ Y i ( t ) > 0 ] is the probability of having zero resistant cells in this compartment at time t . Since we assume independence of the microenvironmental compartments , this implies the probability of having no resistant cells in the entire tumor at time t is given by ∏ i ( 1 - P [ Y i ( t ) > 0 ] ) . Therefore , the probability of resistance at time t is P [ Y ( t ) > 0 ] = 1 - ∏ i 1 - P [ Y i ( t ) > 0 ] , where P [ Y i ( t ) > 0 ] is given by Eq ( 3 ) . The evolutionary dynamics of the tumor cell population depend on the birth and death rates of sensitive and resistant cells in each microenvironmental compartment . These rates , in turn , vary as the concentrations of both drugs change over time . To reflect this variation , we first define distinct functions describing the individual effects of erlotinib and evofosfamide on these birth and death rates . The growth kinetics of the cancer cell population during treatment by each of these drugs are estimated using a combination of pharmacokinetic and experimental cell viability data . In the following sections , functions pertaining to erlotinib are denoted with 1’s and functions pertaining to evofosfamide are denoted using 2’s . We note that the birth and death rates used in this parameterization are taken from in vitro data . Thus the specific time scale and cell population sizes of the model predictions are relevant to the in vitro setting .
We defined the space of all tolerated single-agent and combination therapy dosing schedules using clinical trial data on drug tolerability . For each single-agent treatment we defined a toxicity constraint curve representing the relationship between frequency of drug administration and maximum tolerated dose . In addition , we analyzed the overlapping toxicities between the two drugs as well as each drug elimination rate to determine the necessary conditions for safely administering both drugs in succession . This combination therapy constraint , together with the toxicity constraint curves corresponding to each of the monotherapies , defines the space of all tolerated dosing schedules . To investigate the potential of combination therapies , we first compare treatment outcomes resulting from several combination therapies with the monotherapy schedules currently in clinical use . The standard dosing schedule for erlotinib is 150 mg/day . Two evofosfamide dosing schedules have been tested in a clinical trial and designated as maximum tolerated dosing schedules: 670 mg/m2 given every 3 weeks and 575 mg/m2 given weekly . We consider combination schedules which are clinically feasible and satisfy the toxicity constraints described in the previous section . Table 3 provides an overview of all of these dosing schedules . Schedule A is the standard erlotinib dosing schedule , and schedules B and C are the two evofosfamide dosing schedules . The remaining schedules ( D through J ) represent all combination therapies considered in this analysis . Since the toxicity constraints for erlotinib and evofosfamide are formulated in terms of the number of doses administered in a 3-week period , we define these schedules based on 3-week cycles , and hence only show dosing protocols for the first 21 days since this pattern repeats every 3 weeks for each schedule . Entries in Table 3 represent doses of either erlotinib in mg ( subscript 1 ) or evofosfamide in mg/m2 ( subscript 2 ) . For a fixed schedule ( column ) and day ( row ) , a single entry represents the one dose scheduled for that day , and the lack of an entry indicates that no drugs are administered on that day . Two entries on a single day for a given schedule represent the scheduling of two doses on the same day . For each of the ten dosing schedules in Table 3 , the mean tumor size and probability of resistance over the course of treatment is predicted using the model . The results of these calculations up to recurrence time ( the time at which the cancer cell population reaches its initial size once again ) are plotted in Fig 4A and 4B , respectively . The red curves show the evolutionary dynamics of a tumor during treatment with erlotinib alone , the blue curves show the dynamics during monotherapy with evofosfamide , and the green curves show the evolutionary dynamics of the cancer cell population during combination therapy . The label on each curve indicates which dosing schedule from Table 3 corresponds to those results . The means of the sensitive and resistant cell populations are shown separately in Fig 4C for one of each type of dosing schedule: erlotinib alone , evofosfamide alone , and combination therapy . Fig 4D shows the mean tumor size for the combination schedules , averaged only over those that develop resistance . We find that all the combination therapies considered produce treatment outcomes superior to the standard monotherapy schedules . Fig 4A demonstrates that the combination schedules result in lower average tumor sizes over the course of treatment than those resulting from either of the monotherapies . Even more significantly , Fig 4B shows that the probability of developing resistance decreases dramatically with the use of combination therapy . Under monotherapy with either drug , the probability of resistance eventually reaches one ( in agreement with clinical results ) ; this is due to the fact that sensitive cell division is not sufficiently inhibited by therapy to prevent the emergence of resistance before eradication of the tumor . However the model predicts that for a significant fraction of patients tumor eradication is possible under combination therapy . The breakdown of sensitive and resistant cell populations under therapy is shown in Fig 4C . Erlotinib monotherapy yields a steady but slow decline of the sensitive cell population , due to the fact that the tumor oxygen distribution consists primarily of hypoxic regions and erlotinib does not penetrate well to these areas . On the other hand , treatment with evofosfamide alone targets hypoxic regions which comprise the majority of the cell population , leading to an initial steep decline of the sensitive cell population . However , due to the toxicity constraint during the subsequent break in treatment the mean of the sensitive cells quickly surpasses the initial population size and drives the production of erlotinib-resistant mutants . During combination therapy , however , the cancer cell population demonstrates an initial steep decline due to evofosfamide , followed by a long-term controlled phase due to the combination of evofosfamide and erlotinib . This tight control over the sensitive cell population during combination therapy is possible due to the fact that cancer cells close to blood vessels are receiving lethal concentrations of erlotinib while cancer cells in hypoxic regions are targeted by evofosfamide . Fig 4D demonstrates that for the patients who develop resistance , the average length of time until tumor recurrence is longer for all of the combination therapy dosing schedules than it is for either of the standard monotherapies . For example , it takes patients who develop resistance 40 . 54% longer to rebound on Schedule I compared with standard erlotinib therapy . However , the length of time until recurrence as well as the overall probability of resistance varies between specific combination schedules; this serves as motivation for identifying the optimal timing and dosage sequence for combination schedules in the following section . Finally , we note that the combination strategies were also compared to optimized monotherapies ( subject to the toxicity constraints developed in the previous section ) , and we found that no tolerated monotherapy schedule could outperform combinations in delaying or preventing resistance . We next utilize the mathematical model to optimize over the space of tolerated combination treatment strategies ( constrained by toxicity constraints derived in the previous section ) to minimize the probability of developing resistance or maximally delay recurrence . We consider three distinct classes of combination therapies . Class 1 investigates schedules created by systematically combining standard erlotinib monotherapy with a variety of evofosfamide dosing schedules . However , the amount of time between administration of different drugs can play an important role in the degree to which therapy affects the cancer cell population . Thus the rationale for defining the other two classes are to investigate schedules decreasing the amount of time after erlotinib but before evofosfamide dosing ( Class 2 ) , and after evofosfamide but before erlotinib dosing ( Class 3 ) . For each class , we start with a base erlotinib dosing schedule complying with the monotherapy toxicity constraint curve in Fig 3A . Modifications to this schedule are then made to incorporate n doses of evofosfamide , where n varies from 0 ( corresponding to erlotinib monotherapy ) to a maximal value N ( corresponding to evofosfamide monotherapy ) , in a three-week period . The dose of evofosfamide is determined by the toxicity constraint curve in Fig 3B . Whenever necessary , the minimum number of erlotinib doses are removed to comply with the combination toxicity constraint described in the previous section . To define each combination schedule , we begin by defining a single cycle of length L = 504/n hours , consisting of the base erlotinib dosing schedule and a single dose of evofosfamide . This is done using a 4-step process: ( i ) calculate the evofosfamide dose , ( ii ) place the evofosfamide dose at either t = L − 24 or t = L − 6 depending on the class , ( iii ) fill the remaining time in the cycle with the base erlotinib dosing schedule , and ( iv ) remove any necessary erlotinib doses to comply with the combination toxicity constraint . For a detailed description of how to define a single cycle of treatment for all combination dosing schedules , see S1 Text . These cycles are then repeated some finite number of times to form complete dosing schedules . In Class 1 , we use the standard erlotinib monotherapy schedule of 150 mg/day , and the evofosfamide dose in each cycle is given 24 hours before the start of the next cycle . In Class 2 , we use a low-dose erlotinib schedule of 7 mg twice daily , which allows for a shorter waiting period after erlotinib doses and before evofosfamide doses . Class 3 uses the same standard erlotinib monotherapy as in Class 1; however , the evofosfamide dose in each cycle is given 6 hours before the start of the next cycle instead of 24 , which decreases the amount of time after evofosfamide doses and before erlotinib doses . Fig 5 shows an example depicting dose schedule definition for one cycle of treatment for all three optimization classes when n = 3 . For the reasons stated in the above paragraph , a cycle in Class 1 or 3 contains a standard erlotinib dosing schedule , whereas a cycle in Class 2 contains a low-dose erlotinib schedule . When n = 3 , each cycle is one week . So for Classes 1 and 2 , the evofosfamide dose in each cycle is given 24 hours before the end of the week , and for Class 3 the evofosfamide dose in each cycle is given 6 hours before the end of the week . This information is all depicted in step 1 ( the top row ) of Fig 5 . In step 2 ( the bottom row ) , the necessary number of erlotinib doses leading up to the evofosfamide infusion is removed in order to satisfy the combination toxicity constraint . In Classes 1 and 3 , the last dose of erlotinib shown in step 1 is removed in step 2 to allow the erlotinib concentration to fall sufficiently low before the evofosfamide infusion . Note that in Class 2 the removal of erlotinib is not necessary due to the already low erlotinib concentration . Combination schedules outperform monotherapy endpoints . We calculated the means of the sensitive and resistant cells as well as the probability of resistance , after nine weeks of treatment for every dosing schedule in each optimization class . Table 4 demonstrates that the tumor population size under n = 0 or N ( i . e . the monotherapies ) after 9 weeks is O ( 1010 ) for any optimization class or either drug type , whereas Fig 6 demonstrates that each of the combination therapies considered yields a total population size of O ( 107 ) or less . In fact , when comparing Table 4 to Fig 6 we observe that all of the combination therapies we have investigated yield a considerable benefit ( in terms of tumor population sizes as well as probability of resistance ) over any of the optimal monotherapies in each dosing class . Thus we do not include the endpoints corresponding to monotherapies in Fig 6 due to the large disparity in the sizes of these results . Furthermore , at the end of treatment with evofosfamide , the tumor primarily consists of sensitive cells , which agrees with our previous observation that evofosfamide is unable to control the sensitive cell population without erlotinib . Note that some of the means are unrealistically high for an in vivo setting ( > 1012 ) , due to the fact that the model is parameterized using in vitro growth rates . Means of the sensitive and resistant cell populations for n = 1 … N − 1 ( true combination therapies ) are plotted in Fig 6A and 6B , respectively . The sums of these means , or the mean tumor sizes , are plotted in Fig 6C . Note the presence of two local minima in these three figures . This phenomenon is explained in S1 Text . The probability of resistance for each dosing schedule is plotted in Fig 6D . All four panels include results for Class 1 in blue , results for Class 2 in red , and results for Class 3 in yellow . Note that in these figures , every integer on the x-axis corresponds to a combination dosing schedule given by the number of evofosfamide doses in a 3-week period , as defined previously . Minimize treatment break after evofosfamide dosing . Next consider the results due to combination therapy shown in Fig 6 . We note that all three classes lead to similar results , which suggests that our findings regarding the characteristics of optimal combination dosing strategies are quite robust . Upon closer examination , Fig 6 demonstrates that the tumor population sizes under the Class 3 schedules are generally less than half the population sizes under Class 2 , and also less than those under Class 1 . This suggests that designing schedules that minimize the amount of time after a dose of evofosfamide and before a dose of erlotinib may lead to better control of the tumor population . This finding is in agreement with our previous observations that the tumor population response to evofosfamide is strong but short-lived; hence quickly intervening in the subsequent population growth phase is important . As we move along the spectrum of combination densities ( horizontal axis ) from monotherapy with erlotinib to monotherapy with evofosfamide , there is a clear region in the interior ( approximately n = 9–17 evofosfamide doses ) where the tumor size and sensitive and resistant population size are minimized . This region also contains the region minimizing the probability of developing resistance . To investigate this further , Table 5 shows the specfic n which optimizes the given characteristic ( means of sensitive , resistant , and total cancer cells as well as probability of resistance are all minimized ) . In addition , the bottom row of Table 5 indicates the best overall dosing schedule among all three classes which minimizes the particular value that column represents . Sequential alternating sequences are optimal . From Table 5 we observe that all optimal dosing schedules correspond to values of n between 12 and 17 . This implies that combination therapies incorporating more frequent , smaller doses of evofosfamide result in better treatment outcomes . Even more interestingly , all values of n in Table 5 correspond to the same type of dosing schedule . Besides n = 12 in Class 2 , every other optimal n corresponds to a dosing schedule which alternates between a single dose of erlotinib and a single dose of evofosfamide . We call these alternating dosing schedules . The dosing schedule corresponding to n = 12 in Class 2 consists of two low doses of erlotinib for every dose of evofosfamide , which is still quite similar to the alternating dosing schedules . Thus , even though the optimization ranged over a full spectrum of treatment schedules incorporating variable dose densities for each drug , the optimal therapies were those that utilized close to an equal number of doses of evofosfamide and erlotinib in a sequential alternating fashion .
In this work we considered an approach to investigate the use of hypoxia-activated prodrugs ( HAPs ) to enhance the effectiveness of targeted therapies and in particular , prevent the ( usually inevitable ) emergence of drug resistance . To this end we developed a model reflecting the heterogeneity of oxygen and drug concentrations throughout a tumor to describe the evolutionary dynamics of resistance emerging under combination HAP-targeted therapy strategies . The model was parametrized using experimental and clinical pharmacokinetic data to investigate potential combinations of the HAP evofosfamide with the targeted tyrosine kinase inhibitor erlotinib against EGFR-activated non small cell lung cancer . Our model predictions are useful in comparing the outcomes of a spectrum of dosing schedules in the in vitro setting , and provide a means to use available experimental data to help inform and guide future clinical studies in vivo . We investigated combinations in which doses were not given simultaneously ( to avoid toxicities ) and our model predicted that the complementary action of evofosfamide and erlotinib results in a combined ability to control the tumor’s evolution and growth . In particular: These alternating dosing schedules ( and other similar dosing schedules ) are likely the most effective because the constant switching between erlotinib and evofosfamide allows the strengths of these drugs to complement one another . Too much time spent taking erlotinib without evofosfamide allows the sensitive cell population to remain quite substantial for a long period of time ( due to the lack of targeting the hypoxic regions ) , which , in turn , leads to a high probability of a resistance mutation arising . On the other hand , too much time spent on evofosfamide without erlotinib allows the sensitive cell population to expand drastically since evofosfamide is unable to control its long-term growth . Alternating between these two drugs allows each one to provide the necessary control over the cancer cell population the other one is lacking . In addition , it is important to consider the subpopulation of cancer cells each drug acts on . Erlotinib acts primarily on portions of the tumor microenvironment close to blood vessels , whereas evofosfamide acts primarily on hypoxic regions that are further from the blood stream . Because of this , alternating frequently between the two drugs allows the entire population of cancer cells in the tumor microenvironment to be constantly controlled by the drugs . This same phenomenon has recently been observed with a different combination therapy utilizing evofosfamide in neuroblastoma and rhabdomyosarcoma preclinical models [48] . These results demonstrate that incorporating HAPs in combination with targeted therapies may be an effective tool in preventing resistance , and suggest an alternative use for HAPs . Current clinical trials have combined HAPs not with targeted therapies but with chemotherapies to control tumor growth . In addition , these trials used dosing strategies involving simultaneous drug administration rather than sequential administration , as is used in our model . It is difficult to draw conclusions about the outcome of these clinical trials using our model since this would require growth rate parametrizations and pharmacokinetics for the chemotherapies utilized in the combination treatment ( gemcitabine and doxorubicin ) in different cell types ( pancreatic cancers and soft tissue sarcomas ) . However , we did observe that the reduction in probability of developing resistance is dependent on the exact timing and sequence of the combination therapy . This highlights the importance of using mathematical modeling to predict treatment outcomes and inform decisions regarding schedules to be tested in clinical trials . In addition to its promising clinical implications , this work provides insight into the biological factors which can cause a treatment strategy to either succeed or fail . Specifically , analysis and comparison of the tumor evolutionary dynamics during single-agent and combination therapy suggests that erlotinib and evofosfamide may be effective together because they target separate subpopulations within the tumor microenvironment and on much different scales of time with differing degrees of strength . This theory can be generalized to predict which types of drugs have the potential to be strong partners in combination therapy; specifically , this methodology can be applied to determine the biological and pharmacokinetic parameters that may lead to treatment success or failure with monotherapy or combination therapy . These findings highlight the importance of designing combination therapies with drugs whose strengths complement each other in order to maximize the therapeutic benefits . Another important implication of this work , and something to consider when designing combination dosing regimens using two or more drugs , is the role that variability in timing between the dosing of different drugs plays in treatment outcomes . This work gave rise to multiple promising improvements that could be made in the treatment of non-small cell lung cancer . However , the dosing strategies proposed here need to be tested in vivo to verify these model predictions . In addition , this work provided a novel framework for defining drug toxicity constraints , which is sufficiently general to be extended to any drug or combination of drugs . One planned extension of this work is to further study the implications of cellular migration within the microenvironment on the evolutionary dynamics of the tumor . An initial investigation into this impact is shown in Section 7 of S1 Text . For this study , experimental work investigating the details of the migration patterns and quantification of migration rates in this system are necessary . Other extensions of this work include considering the possibility of pre-existing resistance as well as modeling the bystander effect , which refers to the idea that evofosfamide , once activated in a hypoxic region of the tumor , diffuses outward and affects cancer cells in normoxic regions as well [16 , 19] . In addition , it would be useful to explore the effect of HAPs other than evofosfamide on the probability of developing resistance in order to determine whether the results presented here are specific to evofosfamide or rather are a general phenomenon of HAPs used in combination with tyrosine kinase inhibitors . Since evofosfamide is hypoxia-activated and birth and death rates due to erlotinib are microenvironment-dependent , there is good reason to suspect that alterations to the tumor microenvironment would have a large impact on treatment outcomes with both single-agent and combination therapy .
|
It has been suggested that one key factor driving the emergence of drug resistance is the spatial heterogeneity in the distribution of drug and oxygen throughout a tumor due to disorganized tumor vasculatures . Researchers have developed a class of novel drugs that penetrate to hypoxic regions where they are activated to kill tumor cells . The inclusion of these drugs , called hypoxia-activated prodrugs ( HAPs ) alongside standard therapies in combination may be the key to long-term tumor control or eradication . However , identifying the right timing and administration sequence of combination therapies is an extremely difficult task , and the time and human costs of clinical trials to investigate even a few options is often prohibitive . In this work we design a mathematical model based upon evolutionary principles to investigate the potential of combining HAPs with standard targeted therapy for a specific example in non-small cell lung cancer . We formulate novel toxicity constraints from existing clinical data to estimate the shape of the tolerated drug combination treatment space . We find that ( i ) combining these therapies delays resistance longer than any monotherapy schedule with either evofosfamide or erlotinib alone , and ( ii ) the best strategy for combination involves single doses of each drug sequentially administered in an alternating sequence . These model predictions of tumor dynamics during treatment provide insight into the role of the tumor microenvironment in combination therapy and identify treatment hypotheses for further experimental and clinical testing .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"cardiovascular",
"anatomy",
"oxygen",
"cancer",
"treatment",
"cell",
"processes",
"toxicology",
"oncology",
"toxicity",
"pharmaceutics",
"drug",
"administration",
"blood",
"vessels",
"chemistry",
"blood",
"plasma",
"hematology",
"blood",
"cell",
"biology",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"drug",
"therapy",
"chemical",
"elements"
] |
2016
|
Leveraging Hypoxia-Activated Prodrugs to Prevent Drug Resistance in Solid Tumors
|
Polyomavirus BK ( BKPyV ) frequently reactivates in immunosuppressed renal transplant recipients ( RTRs ) and may lead to graft loss due to BKPyV-induced interstitial nephritis ( BKVN ) . Little is known on the differentiation of CD8+ T cells targeting BKPyV in RTRs . Here we investigated whether BKPyV-specific CD8+ T cell differentiation differs in RTRs with varying degrees of BKPyV reactivation and/or BKVN . Using combinatorial encoding with tetramers carrying BKPyV major capsid protein ( VP1 ) and large T antigen protein ( LTAG ) epitopes , we investigated CD8+ T cell responses to BKPyV in longitudinally obtained PBMC samples from 46 HLA-A02-positive RTRs and 20 healthy adults . We were also able to isolate BKPyV-specific CD8+ T cells from five renal allografts , two of which were affected by BKVN . Before transplantation , BKPyV-specific CD8+ T cells targeting VP1 and LTAG epitopes appeared predominantly as central-memory and CD27+/CD28+ effector-memory ( TEM ) , and naïve-like PD-1-expressing cells , respectively . After viral reactivation , BKPyV-specific CD8+ T cells assumed CD28− TEM and TEMRA states in patients who were able to control BKPyV , whereas differentiation lagged behind in patients with severe viral reactivation or BKVN . Furthermore , VP1-specific CD69+/CD103+ tissue-resident memory ( TRM ) cells accumulated in BKVN-affected allografts but lacked signs of effector differentiation . In contrast , granzyme B-expressing effector cells were detected in allografts not affected by BKVN . In conclusion , effector-memory differentiation of BKPyV-specific CD8+ T cells in patients with high viral load or BKVN is impaired . Further characterization of the specific mechanisms behind this altered cellular differentiation is necessary to develop therapies that can prevent the emergence of BKVN .
Polyomavirus BK ( BKPyV ) establishes a mode of latent infection in the vast majority of the general , immunocompetent population [1 , 2] . However , in immunosuppressed renal transplant recipients ( RTRs ) , BKPyV can escape the weakened immunological response leading to reactivation in up to 60% of the patients . In as much as 10% of these reactivations , the virus causes a severe interstitial nephritis ( BKVN ) in the allograft that is associated with graft loss [3 , 4] . Until now , the only effective treatment option for BKPyV reactivation following renal transplantation involves tapering of the immunosuppressive drug therapy , allowing the patient’s immune system to recover and overcome the virus . However , this also increases the chance on allograft rejection [3 , 4] . For these reasons , effective and more specific treatment strategies are urgently needed . It is here that modern immunotherapies , such as adoptive transfer of virus-specific T cells , come into view . Recently , it was shown that BKPyV reactivation occurs concomitantly to a loss of polyfunctional T cells specifically targeting BKPyV epitopes , emphasizing the importance of T cells for effective immunological control of this virus [5–7] . T cell populations specific for BKPyV can be expanded in vitro and may then theoretically be used to treat BKPyV reactivation [8] . However , because each human virus triggers the formation of a specialized subset of T cells , carrying a distinct armamentarium to combat the respective virus [9] , it is essential to understand what type of T cells confers protection against BKPyV . Previously , we used BKPyV virion protein 1 ( VP1 ) peptide-loaded HLA A02-restricted tetramers to determine the phenotype and function of VP1-specific CD8+ T cells in the circulation of healthy individuals . We found that these cells largely exist in a central-memory ( TCM ) or early-differentiated state [10] , a phenotype that was recently associated with stem cell-like properties [11] . However , in healthy individuals BKPyV-specific T cells may seldom encounter their cognate antigen [12] , whereas in RTRs BKPyV frequently reactivates , thus exposing the host’s T cells to substantial amounts of antigen and inflammation . Because of their specific capacity to detect and control intracellular pathology , as caused by viruses , we here investigated the phenotypic and functional differentiation of BKPyV VP1- and large T antigen ( LTAG ) -specific CD8+ T cells in the circulation of RTRs suffering from various degrees of BKPyV reactivation over the course of transplantation . In addition , we characterized BKPyV-specific T cells obtained from the allograft of some patients . Using this approach we aimed to identify whether differences in clinical outcome of BKPyV-infection are associated with altered differentiation pathways and/or effector functions of CD8+ T cells targeting this virus . Using combinatorial encoding with six different HLA A02-restricted tetramers we confirmed that VP1-specific cells before transplantation mainly exist in a central-memory ( TCM ) or early-differentiated effector-memory ( TEM ) state , whereas LTAG-specific CD8+ T cells unexpectedly exhibit a naïve-like phenotype with frequent expression of PD-1 . After transplantation , both VP1 and LTAG-specific cells showed CD28− TEM differentiation , sometimes with CD45RA re-expression ( TEMRA ) . This mainly occurred in RTRs with low or undetectable viral load but not in patients with high viral load and/or BKVN . Within the renal allograft of two BKVN patients , we detected a high frequency of CD69/CD103-expressing tissue-resident BKPyV VP1-specific memory cells that , in contrast to the CD69/CD103-negative recirculating BKPyV-specific cells in kidneys from non-BKVN-affected patients , did not express granzyme B .
We included longitudinally obtained samples from 46 HLA-A02-positive RTRs: 21 in whom BKPyV replication had not been observed in the first year after transplantation ( not-reactivating or NR patients ) , 11 RTRs in whom BKPyV had reactivated with a peak viral load below 1*104 copies/ml ( Rlow patients ) , 6 RTRs showing BKPyV reactivation with a peak viral load higher than 104 copies/ml ( Rhigh patients ) , and 8 RTRs with peak viral load higher than 104 copies/ml and biopsy-proven BKVN ( BKVN patients ) . Samples from 20 HLA A02-positive healthy individuals served as a control . When comparing all study groups containing RTRs , there was a statistically significant difference in overall HLA mismatches that derived from a high total number of mismatches in the NR patients . Also , donor age was greater in the Rlow patients when compared to BKVN patients . Finally , estimated glomerular filtration rates were significantly lower in Rhigh patients when compared to NR patients ( Table 1 ) . From five other patients who underwent a graft biopsy because of deterioration in renal allograft function during active BKPyV-infection , we obtained graft-eluted cells . Histological examination revealed BKVN in 2 of them , and no BKPyV infection in the other three patients . All grafts contained various degrees of interstitial fibrosis , tubular atrophy and cellular infiltrates . Serological assessment showed the presence of anti-BKPyV antibodies in all patients before transplantation . Antibody titres increased significantly in the first year after transplantation in all RTRs in whom BK viremia was detected , but not in the NR patients ( Fig 1A ) . Therefore , the rise in antibody titres is a reflection of viral reactivation as measured in the circulation but does not necessarily seems to prevent the reactivation as was shown previously [5 , 13] . Peak viral load in RTRs were detected most often in the second and third quarter of the first year post transplantation ( Fig 1B ) . The viral load in the Rlow patients had dropped close to the quantifiable detection threshold of 1000 copies/ml already at the ≤6 months post peak viral load points . In the Rhigh patients , this did not occur until somewhere in between the ≤1 year and ≤ 2 year post peak viral load time points . BKVN patients did not drop below this threshold during follow-up ( Fig 1B ) . In response to detection of BKPyV viremia , the dosage of immunosuppressive drugs was carefully diminished , aimed at decreasing the BKPyV-load and preserving renal allograft function . First , the dose of mycophenolate mofetil was tapered in steps of 250 to 500 mg per 2 weeks , followed by decreasing the dose of tacrolimus by 0 . 5 to 2 mg per 2 weeks . Previously , BKPyV-specific CD8+ T cells were shown to be present in the circulation of both healthy individuals and RTRs at extremely low frequencies [10 , 15–18] . To enhance the sensitivity and specificity of detection of BKPyV-specific CD8+ T cells , we here used combinatorial encoding of HLA-A02 tetramers loaded with two different immunodominant BKPyV VP1 peptides and one immunodominant LTAG peptide ( S1A Fig ) . Using this technique , and staining a large number of PBMCs ( up to 12*106 PBMCs per sample ) , we detected BKPyV VP1-specific CD8+ T cells in 6 out of 20 healthy individuals , and in 2 of 21 NR patients; 8 of 11 Rlow patients; 6 of 6 Rhigh patients; and in 5 of 8 BKVN patients at some time point ( s ) during follow-up . We detected LTAG-specific cells in 12 of 20 healthy individuals , and in 4 of 21 NR patients; in 2 of 11 Rlow patients , in 2 of 6 Rhigh patients and in 4 of 8 BKVN patients during follow-up ( S2 Fig ) . In RTRs , both VP1 and LTAG-specific cells were detected more frequently during viremia . Expansion of BKPyV-specific CD8+ T cell populations occurred in some individuals after transplantation , but not in all patients ( Fig 1C ) . This was corroborated by a rise in Ki-67 expression after transplantation , particularly by the VP1-specific cells , indicating active cell proliferation . Ki-67+ expressing cells were detected neither in the samples from the NR patients , nor in those from the healthy individuals ( Fig 1D ) . Using multichannel flowcytometry , we determined the expression of various molecules characteristic for T cell differentiation and function ( S1B Fig ) . Previously , we found that circulating BKPyV VP1-specific CD8+ T cells in healthy individuals were predominantly TCM cells ( CD45RA−CCR7+CD27+ ) or early-differentiated TEM ( CD45RA−CCR7−CD27+ ) cells [10] . In the current study , adding the expression of CD28 to the classification , we confirmed these findings ( Fig 2A ) . Both LTAG and VP1-specific CD8+ T cells circulating in RTRs before transplantation showed similar phenotypes as in healthy individuals ( Fig 2A ) . Comparison of LTAG and VP1-specific CD8+ T cells , however , revealed substantial differences in both healthy individuals and RTRs , with the LTAG-specific CD8+ T cells displaying a predominant CD45RA+CCR7+CD28+CD27+ surface phenotype ( Fig 2A ) . This phenotype may define antigen-inexperienced T cells , but also a subset of very early differentiated antigen-experienced CD8+ T cells with stem-cell-like traits , that , amongst others , is defined by expression of the tumour necrosis factor receptor family member CD95 ( FAS receptor ) [19 , 20] . However , only about 16% of LTAG-specific CD8+ T cells with a “naïve” CD45RA+CCR7+CD28+CD27+ phenotype expressed CD95 , which equalled the CD95 expression on the total population of CD45RA+CCR7+CD28+CD27+ CD8+ T cells ( S3 Fig ) . Thus , based on this surface marker , only a fraction of LTAG-specific cells could be assigned as typical stem-cell memory cells . Importantly , the LTAG-specific cells were significantly enriched for the expression of PD-1 when compared to the total naïve CD8+ T cell pool ( S3 Fig ) suggesting that they have indeed been stimulated by antigen . In addition , no major differences were found between the BKPyV-specific CD8+ T cells of patients just before renal transplantation and healthy control individuals regarding other immunological characteristics of BKPyV-specific cells like their T-bet- or Eomes expression; expression of granzyme B or granzyme K , and IL-7Rα ( CD127 ) , PD1 , or CD95 ( Fig 2 ) . During BKPyV reactivation , the composition of both VP1- and LTAG-specific CD8+ T cell populations changed , as shown in Fig 3A and S4 Fig . The most profound changes were noted in the Rlow patients , in whom substantial proportions of normally cytotoxic intermediately-differentiated ( CD45RA−CCR7−CD28−CD27+ ) , CD45RA− effector-type ( CD45RA−CCR7−CD28−CD27− ) TEM and TEMRA ( CD45RA+CCR7−CD28−CD27 ) CD8+ T cell subsets specific for either VP1 or LTAG became detectable during and after the time point of peak viral load . In the Rhigh and BKVN group , these subsets were also formed amongst the VP1-specific CD8+ T cells but later in time and in smaller proportions . CD28− TEM subsets also emerged amongst LTAG-specific populations , but primarily at the moment of peak viral load in the Rhigh , after which their sizes diminished during the later time points . CD28− TEM differentiation was seldom observed in the BKVN patients . Differentiation had also occurred within the LTAG-specific cell-populations from NR patients at one year post-transplantation ( Fig 3 and S4 Fig ) . Recently , we found that the expression levels of T-bet and Eomes , master transcriptional regulators of type 1 ( cytotoxic ) T cell differentiation , are strong indicators of the degree of CD8+ T cell differentiation [21] . We also showed that BKPyV VP1-specific CD8+ T cells circulating in healthy individuals mostly express low or intermediate levels of T-bet , whereas they lack expression of Eomes [10] . Here , we studied whether the expression of T-bet and Eomes was influenced by the BK viremia occurring in RTRs . Fig 4A shows that at all time points and in each patient group , VP1- and LTAG-specific cells expressed significantly more T-bet than Eomes . The frequency of T-bet- and Eomes-expressing VP1-specific cells was comparable between the different study groups . Although referring to data from only six patients , the frequency of both T-bet and Eomes-expressing LTAG-specific CD8+ T cells appeared to be higher in the Rlow patients than in the other study groups . This is also illustrated by Fig 4B , which shows two representative patients from the Rlow- , respectively BKVN group . Remarkably , despite the clear CD28− TEM differentiation detected in the LTAG-specific CD8+ T cells from NR patients around the first year after transplantation ( Fig 3A ) , these populations did not contain increased frequencies of T-bet- and Eomes- expression at that time point ( Fig 4A ) . The cytokine IL-7 is important for T cell homeostasis in the absence of antigen and inflammation and IL-7Rα expression is rapidly lost following T cell receptor-dependent activation [22] . As described previously , nearly all VP1-specific cells in healthy individuals expressed IL-7Rα , further suggesting that these cells infrequently encounter their antigen ( Fig 2 ) [10] . As shown above , we found similar data for the LTAG-specific cells in healthy individuals and in patients just before renal transplantation ( Fig 2B ) . IL-7Rα was also expressed on the majority of BKPyV-specific CD8+ T cells in NR-patients , Rhigh and BKVN patients ( Fig 5A ) . In sharp contrast , IL-7Rα expression in the Rlow patients was clearly downregulated during BKPyV-reactivation , especially on the LTAG-specific cell populations , as is also illustrated by two representative patients from the Rlow— , respectively the BKVN group ( Fig 5B ) . Next , we studied functional properties of BKPyV-specific CD8+ T cells , viz . their cytotoxic capacity as judged by both the presence of the serine proteases granzyme K and granzyme B , expression of the degranulation marker CD107a and their cytokine-producing capacity . Previously , we found that a small number of BKPyV VP1-specific CD8+ T cells in healthy individuals expressed granzyme K and/or B , which we confirmed in the present study ( Fig 2 ) [10] . Despite the CD28− TEM differentiation occurring after BKPyV reactivation , particularly in the Rlow group , no clear differences in granzyme expression were observed at any time-point between these and other patients ( Fig 6 ) . As a marker for degranulation , we studied the surface expression of CD107a on BKPyV-specific CD8+ T cells after stimulation in vitro . Fig 7 shows in all groups at all time points a rather low frequency of CD107a+ cells , suggesting minimal degranulation of these cells , at least in the peripheral circulation . The cytokine production capacity of the different BKPyV-specific CD8+ T cell populations was tested by stimulating PBMC with PMA/ionomycin , followed by visualization of the BKPyV-specific CD8+ T cells using combinatorial encoding with tetramers . This approach is hindered by downregulation of the T cell receptor upon T cell activation . For unknown reasons , this particularly affected the LTAG-specific cells in the Rhigh and BKVN patients . As such , we were unable to detect sufficient LTAG-specific cells in these patient groups for analysis . In the Rlow group , where LTAG-specific cells were still detectable after stimulation , we observed that a modest proportion produced IL-2 , TNFα and INFγ ( Fig 7 and S5 Fig ) . Previously , we found that the majority of VP1-specific CD8+ T cells in healthy individuals produced combinations of three cytokines , most commonly IL-2 , INFγ and TNFα [10] . This was confirmed in the present study , and was also observed in patients before renal transplantation and thereafter , irrespective of detectable BKPyV reactivation ( Fig 7 ) . No major differences in cytokine production capacity of VP1-specific cells were observed during follow-up . Because BKPyV nephropathy is the final consequence of uncontrolled BKPyV-replication in the kidney allograft , we studied the presence of BKPyV-specific CD8+ T cells within the graft of two patients and compared them to their peripheral blood counterparts . As a control , we studied graft-eluted cells from three RTRs without BKVN . In the two BKVN grafts , we detected only VP1-specific CD8+ T cells , whereas in the three non-BKVN-affected grafts we detected one VP1- and two LTAG-specific populations . We found that in the BKVN grafts the VP1-specific T cells were about 10e5 times enriched when compared to the peripheral blood compartment . In contrast , the frequencies of the one VP1-specific population and two LTAG-specific populations that we detected in the non-BKVN allografts , were similar to those in the paired peripheral blood samples ( Fig 8A ) . Tissue-resident memory T-cells ( TRM ) are characterized by expression of CD69 and CD103 [23 , 24] , both molecules ensuring that TRM populations are retained in the respective tissue and that they do not re-enter the circulation [25] . In the two patients with BKVN , most of the graft-eluted VP1-specific CD8+ T cells expressed both CD69 and CD103 , designating them as TRM cells ( Fig 8B ) . In contrast , in patients without BKVN , a minority of the BKPyV-specific CD8 T cells stained double-positive for these markers . The graft-eluted VP1-specific CD8+ T cells from the two BKVN patients were comparable to those in peripheral blood , showing a CD45RA−CD27+/− TEM phenotype ( Fig 8C ) . Both VP1- and LTAG-specific graft-eluted cells from patients without BKVN were also quite similar to their PB counterparts but showed a more advanced differentiation state , bearing a CD27− TEM or TEMRA phenotype . The accumulated VP1-specific CD8+ T cells in the two BKVN-affected kidneys contained very few granzyme B positive cells . In contrast , a considerable proportion of the BKPyV-specific CD8+ T cells in the non-BKVN-kidneys expressed this serine protease , although the percentage was lower than in the PB compartment ( Fig 8D ) .
Here , we document that in renal transplant patients with high viral load and/or BKVN , the effector-memory differentiation of circulatory BKPyV VP1- and LTAG-specific CD8+ T cells is distinct from that in patients with low viral load . VP1-specific CD8+ T cells collected before transplantation started off with a TCM or early-differentiated TEM phenotype , whereas the LTAG-specific cells curiously primarily displayed a naïve-like phenotype . Nevertheless , following transplantation and viral reactivation in the Rlow patients , both VP1- and LTAG-specific populations differentiated into CD28− TEM cells , with LTAG-specific cells even acquiring the TEMRA state . In the Rhigh and BKVN patients , VP1- and LTAG-specific CD8+ T cells instead generally persisted in their TCM and CD28+CD27+ TEM differentiation state . In line with this , the frequency of circulating T-bet and Eomes-expressing LTAG-specific cells was highest in patients with low viral replication . Furthermore , the BKPyV-specific CD8+ T cells in Rlow patients downregulated their expression of IL-7Rα , emphasizing the activation of these cells . Despite these dissimilarities in differentiation patterns , the BKPyV-specific cells in the distinct patient groups expressed similar but low levels of granzyme K and B . Also , we did not find any difference between the groups in cytokine production by the BKPyV-specific CD8+ T cells , which were polyfunctional as we showed before . Because we found no differences in properties of BKPyV specific CD8+ T cells between healthy individuals and patients shortly before transplantation , possible effects exerted by the uremic state or by any drug medication at present or in the past seem not to be involved . When compared to human cytomegalovirus ( hCMV ) or Epstein-Barr virus ( EBV ) -specific CD8+ T cells , the frequencies of BKPyV-specific cells in the circulation are very low , making them difficult to detect [10 , 15–18 , 26] . Schachtner et al . used in vitro stimulation with overlapping BKPyV peptide pools in an Interferon-γ Elispot assay , and showed that the overall BKPyV-specific CD4+ and CD8+ T cell response was significantly delayed in patients who developed BKVN [5] . The same group recently demonstrated that this delay concerns mainly the T cell response targeting LTAG epitopes [6] , which is in line with the data presented here . Schaenman et al . recently also reported impaired BKPyV-specific CD8+ and CD4+ T cell responses in patients with severe reactivation of BKPyV . However , in contrast to the findings presented in the current manuscript , this group detected a particularly frequent expression of CD107a by BKPyV-specific CD8+ T cells [7] . We think that this results from much longer T cell stimulation in vitro , which was applied to assay T cell cytokine/CD107a expression capacity . We would like to emphasize that in vitro stimulation with peptide and co-stimulation for several hours will significantly alter the phenotype of T cells . For example CCR7 , IL-7Rα and CD62L expression rapidly disappears from the cell surface after stimulation [22 , 27 , 28] . Also , in vitro T cell activation results in the induction of T-bet and Eomes expression , which directly induce expression of molecules like granzyme B , interferon-γ , CD122 and IL-15Rα [29–36] . This does not occur when using tetramers to isolate T cells if done in the proper conditions , viz . brief period of staining in the absence of co-stimulation , at a low temperature . The low expression frequency of CD107a , as detected in our current study , is in line with the low expression frequency of granzyme B by the BKPyV-specific CD8+ T cells , since both molecules are located in the same cytotoxic granules [37] . However , whereas tetramers are well fit to determine the phenotype and functional properties of antigen-specific T cells directly ex-vivo , they cannot visualize the total number of virus-specific T cells active against a given antigen , as can be done with overlapping peptide pool stimulation assays , owing to the epitope restrictions of the tetramers . The naïve-like LTAG cells detected prior to transplantation expressed PD-1 significantly more often when compared to the total CD45RA/CCR7/CD28/CD27 CD8+ T cell population . Apart from being a marker of functional exhaustion , PD-1 is also recruited into the immunological synapse upon T cell activation [38 , 39] . Therefore , this naïve-like state may represent a subset of antigen-experienced T cells in a very early differentiation state , close to the CD95-expressing naïve-like population of stem-cell memory cells that was described recently [19 , 20] . In view of the low percentage of granzyme-expressing cells , it may therefore well be that the normal immunological control of BKPyV by CD8+ T cells is not exerted by granzyme K or B . For example , human CMV-specific CD8+ T cells highly express granzyme B and T-bet . Instead , CD8+ T cells targeting EBV epitopes , primarily express granzyme K and Eomes , suggesting that each virus is controlled by a distinct type of CD8+ T cell equipped with a specific armamentarium [21 , 40–44] . Therefore , CD8+ T cells may also have adopted a distinct strategy to control BKPyV , especially considering the long relationship between man and this virus [45] . Given the polyfunctionality with regard to cytokine production , BKPyV-specific CD8+ T cells may rely much more on production of typical cytokines to control BKPyV proliferation than on exerting cytotoxicity against infected cells . It is important to mention that we only investigated the immunodominant HLA-A02-restricted T cell response in this study . Whilst this was done because this is the most abundant HLA class I molecule expressed by the general Western population , immunodominant BKPyV T cell responses indeed also occur via other HLA class I molecules as shown recently by Cioni et al . [46] . Furthermore , different viral proteins can trigger different types of T cells , amongst which possibly cells with immunomodulatory function . One should also consider that the mechanism by which viral control is executed , may not be reflected by T cells located in the peripheral blood compartment . Indeed , the epicentre of BKPyV infection and inflammation is located within the renal allograft and not in the circulation . In the two patients with BKVN , from whom we obtained graft-eluted cells , the frequency of VP1-specific CD8+ T cells in the graft was indeed much higher than in their paired peripheral blood samples , suggesting sequestration of virus specific cells within the allograft . The majority of these graft-eluted cells consisted of CD69/CD103 double-positive TRM cells [23] . Surprisingly , also here only very few of these cells expressed granzyme B . Considering the immunopathology in BKVN grafts , as evidenced by histological damage and deteriorated graft function , this large TRM population appeared not capable to control the viral infection . In contrast , in patients without BKVN , only few BKPyV specific TRM cells were detected in the graft , that apparently possibly contributed to local control of the virus . Although more granzyme B-expressing cells were present than in the BKVN patients , they were mainly CD103-negative and their frequency was still lower than in the peripheral blood compartment . Probably , the intragraft BKPyV-specific CD8+ T-cells with the CD103-negative effector phenotype , are recirculating cells . In fact , in paired peripheral blood samples , similar phenotypes were found . Whether the TRM cells originate from in situ differentiation of these recirculating effector cells , or vice versa , is unknown . Neither do we understand why so few effector cells were detected in the BKVN-allografts , and why the large population of TRM cells in the BKVN-allografts failed to contain the infection . This situation is reminiscent of so-called tumour-infiltrating lymphocytes ( TILs ) , which are in general dysfunctional [47] . By analogy with that , we suppose to name these cells as Virus-specific Tissue-Infiltrating Lymphocytes ( V-TILs ) . Given the small sample size in the current study , further research into ( BKPyV-specific ) kidney-resident T cell memory populations is required . Specific reasons for the impaired effector-memory differentiation of circulating BKPyV-specific CD8+ T cells in the patients with high viral loads / BKVN require further research . One possibility is that differentiation did occur , but was not measurable in the peripheral blood compartment due to retention of these cells in the tissue . Considering truly impaired differentiation , this may be the consequence of defective CD4+ helper cell function , insufficient costimulation , individual differences in susceptibility to immunosuppressive medication , or differences in the virulence of various BKPyV sub- or quasispecies . More knowledge on these possibilities , also on BKPyV-specific CD4+ T cell differentiation in these patients , is needed to better understand the disease process in order to develop effective BKPyV-directed immunotherapy in the future . In conclusion , our findings show an impaired effector-memory differentiation program of BKPyV-specific CD8+ T cells in patients with severe BKPyV reactivation and/or BKVN . This offers an explanation for the pathogenesis of this clinical entity in RTRs , as well as a rationale for the potential effectiveness of immunotherapies to treat BKPyV reactivation in the future .
From the cohort of renal transplant recipients ( RTRs ) who were transplanted at the Academic Medical Center ( AMC , Amsterdam , The Netherlands ) between 2008 and 2013 , we selected 25 HLA-A02-positive patients , who experienced a reactivation of BKPyV-infection as demonstrated by a positive DNA real-time quantitative PCR ( qPCR ) in plasma within the first two years after a first transplantation . We included only HLA-A02-positive individuals in this study because this is the most ubiquitously expressed HLA subclass ( ~50% ) by the Western population . BKPyV DNA was quantified before and at regular intervals of 3 months after transplantation , and more frequently when qPCR had become positive , or earlier when BKPyV reactivation was clinically suspected . Peripheral blood samples were collected at the same time points; mononuclear cells ( PBMC ) and sera or plasma samples were frozen . Time points chosen for analyses comprise: pre-transplantation ( pre Tx ) ; the period prior to detection of the peak viral load ( pre-peak ) ; the moment of peak viral load; the period of the first 6 months after detection of the peak viral load ( ≤6 months post peak ) ; the period from month 6 to month 12 after detection of the peak viral load ( ≤ 1 year post peak ) ; and the period between the first year and the second year after detection of the peak viral load ( ≤2 years post peak ) . Data points of individual patients shown and analysed were the ones collected closest to t = 6 months post peak , t = 12 months post peak and t = 24 months post peak . The pre-peak time point was defined as the number of months from transplantation to peak viral load divided by two . For obvious reasons , these restrictions did not apply to the pre-transplantation samples and the peak viral load samples as these concerned single sampling moments . Each time frame holds no more than one data point from an individual patient . All other data points collected and measured during follow-up were excluded from the analyses and the graphs shown in this manuscript . Immunosuppressive treatment included induction with CD25mAb ( Basiliximab ) , and maintenance therapy , consisting of corticosteroids 10 mg/day orally , mycophenolate mofetil 2 gram/day and tacrolimus aimed at serum trough levels of 6–10 ng/ml . Exclusion criteria comprised previous transplantation , PRA > 5% , inadequate viral load monitoring frequency , inadequate sampling frequency and/or treatment with immunosuppressive medication other than the agents described above . From the same cohort of RTRs , 21 HLA-A02-positive patients were included in whom no BKPyV reactivation occurred . These patients were treated , monitored and sampled according to the same protocol . In addition , we isolated mononuclear cells from renal allograft tissue and paired peripheral blood of 5 RTRs . Two patients who underwent a graft biopsy because of deterioration in renal allograft function during active BKPyV-infection were diagnosed to have BKVN based on histological analysis and a positive SV40 staining . BKPyV was not actively replicating in the three other RTRs and histological signs of BKPyV infection were lacking . All grafts contained various degrees of interstitial fibrosis , tubular atrophy and cellular infiltrates . As a control , we also included PBMC isolated from 20 HLA-A02-positive buffy coats from healthy blood donors ranging between 18 and 64 years of age ( Sanquin , Blood Supply , Amsterdam , the Netherlands , Table 1 ) . For these latter subjects we could not obtain serum samples . We chose a viral load of 10e4 copies/ml as cut-off value between Rlow and Rhigh patients , because it was previously proposed as a critical threshold for developing BKVN [48] . However , as opposed to the BKVN patients , we were unable to detect BKVN in the Rhigh patients by immunohistochemistry of their allograft biopsies . The study was approved by the Medical Ethical Committee of the AMC , and written-informed consent was obtained from all patients in accordance with the Declaration of Helsinki . PBMC were obtained using standard density gradient centrifugation and subsequently cryopreserved until the day of analysis [49] . Samples of human renal cortex were obtained from transplantectomies and renal allograft biopsies . Kidney mononuclear cells were isolated using mechanical disruption and enzymatic digestion . Renal cortex tissue was cut into small pieces , washed thoroughly with PBS to remove blood and incubated with collagenase type IV ( 150 U/ml , Worthington , Lakewood , NJ , USA ) and DNase I type IV ( 50 U/ml ) in HBSS + 2% fetal calf serum ( FCS ) + 0 . 6% bovine serum albumin ( BSA ) for 20’ at 37°C . The tissue pieces were washed and processed through a single-cell strainer . Renal biopsy eluates were analyzed directly . Isolates of larger kidney samples underwent density gradient centrifugation and were cryopreserved . Viral DNA was isolated from 200 ul plasma sample by Magnapure96 isolation ( Roche applied Science ) using the total nucleic acid isolation kit according to the instructions of the manufacturer . Subsequently , isolated DNA was amplified by an internally controlled quantitative realtime TaqMan PCR targeting the Large T-antigen Gene . Quantification was based on standard curves using quantified plasmid DNA containing the target sequence . Values over 1000 copies/ml were considered to be positive . Serum samples were analysed by Luminex for IgG reactivity against the BKPyV-genotype Ib1 major capsid protein 1 ( VP1 ) according to a published protocol [50] . Glutathione—casein ( GC ) coupled Bio-Plex polystyrene beads ( Bio-Rad Laboratories , Hercules , CA , USA ) containing a combination of fluorescent dyes were coupled to either GST-BKPyV VP1 . tag or GST . tag . For each antigen , 3 , 000 GC-coupled beads per sample were loaded with crude bacterial lysates containing relevant GST-fusion protein . Samples were preincubated with GST . tag containing bacterial crude lysates ( 2 mg/mL ) in blocking buffer to reduce nonspecific GST binding . The antigen-coated bead mixtures were incubated with serum diluted 1:100 . For detection of bound serum antibodies , beads were incubated with goat anti-human total immunoglobulin G—biotin ( 1:1 , 000 dilution; Jackson ImmunoResearch Laboratories Inc . , West Grove , PA , USA ) , streptavidin R—phycoerythrin ( 1:1 , 000 dilution; Invitrogen ) , and washed . Beads were analyzed in a Bio-Plex 100 analyzer ( Bio-Rad Laboratories ) . Results are presented as median fluorescent intensity ( MFI ) units . For each sample , antigen-specific binding was obtained by subtracting the MFI for beads coated with GST alone from those of beads coated with GST VP1 . The cut-off value to determine BKPyV-seropositivity was based on sera of healthy children aged 10–15 months old , as described [51] . For the detection of BKPyV-specific CD8+ T cells we utilized combinatorial encoding with six HLA-A02 tetramers loaded with different immunodominant BKPyV peptides . With this technique we generate unique two-colour codes for the parallel detection of three different BKV-specific CD8+ T cells populations . As described previously , this technique significantly increases the sensitivity in comparison to single multimer staining and allows for a detection limit as low as 0 . 002% of total CD8+ T cells in large sample sizes ( S1B Fig ) [52] . To achieve a large enough sample size , we stained up to twelve million PBMC with the tetramers per experiment and determined the presence of BKPyV VP1 and LTAG-specific CD8+ T cell populations as well as their expression of various surface and intracellular markers by multichannel flowcytometry ( S1C Fig ) . As advised previously , we used a pre-defined inclusion cut-off value of at least 10 double-positive tetramer events ( S1B Fig and S1 Table ) [52] . Tetrameric complexes were obtained from Sanquin ( Amsterdam , Netherlands ) and from the NIH Tetramer Core Facility . Three different and previously tested immunodominant epitopes , shared by the majority of BKPyV strains were selected [10 , 15 , 16 , 26] . This concerned two BKPyV capsid protein VP1 epitopes: BKPyV VP1-derived AITEVECFL ( VP1 p44 ) and BKPyV VP1 LLMWEAVTV ( VP1 p108 ) ; and one large T antigen protein ( LTAg ) epitopes: BKPyV LTAg LLLIWFRPV ( LTAg p579 ) . These were incorporated in phycoerythrin ( PE , Sanquin ) , allophycocyanin ( APC ) and Brilliant Violet™ 421-labeled HLA-A02 tetrameric complexes ( NIH ) . PBMC were washed in phosphate-buffered saline containing 0 . 01% ( wt/vol ) NaN3 and 0 . 5% ( wt/vol ) bovine serum albumin . Samples were split into aliquots of two million cells . Each aliquot was incubated with a mix of PE- , APC- , and BV421-labeled tetrameric-complexes for two different BKPyV VP1 epitopes and one BKPyV LTag epitope ( Sanquin , Amsterdam , Netherlands ) , followed by incubation with a combination of the following antibodies: CD27 APC-eFluor780 ( eBioscience Inc , San Diego , CA , USA ) , CD8 BrilliantViolet ( BV ) 785 , IL-7Rα BV711 , CXCR6 PE-Cy7 ( BioLegend , San Diego , CA , USA ) , CD3 V500 , CD45RA BV650 , CCR7 Brilliant UltraViolet ( BUV ) 395 , PD-1 BrilliantBlue515 , CD14 PE-CF594 , CD19 PE-CF594 , CD21 PE-CF594 , CD95 BV711 ( BD Biosciences , San Jose , CA , USA ) , CD28 FITC ( Sanquin ) . Dead cells and duplets were excluded from analysis by using Live/Dead fixable staining ( Life Technologies Europe BV , Bleiswijk , Netherlands ) and height- and width event characteristics , respectively ( S1C Fig ) . The FOX-P3 staining kit ( eBioscience ) was used for intracellular stainings with the following antibodies: Eomesodermin PerCP-eFluor710 , granzyme K PerCP-eFluor710 , T-Bet PE-Cy7 ( eBioscience ) , Ki-67 BUV395 and granzyme B AlexaFluor700 ( BD Biosciences ) . Cells were washed twice , all aliquots of a sample were pooled and up to ten million PBMC per sample were measured on an LSRFortessa flow cytometer and analyzed with FlowJo Version 9 . 3 . 3 software . Only live CD19−CD4−CD20−CD8+CD3+ lymphocytes positive for both differently labelled but otherwise identical tetramers were considered specific for the BKPyV epitope presented in the HLA-A2 tetramer ( S1B Fig ) . CD8+ T cell differentiation was determined by surface expression patterns of CD45RA , CCR7 , CD28 and CD27 . We used a classification that defines the seven largest functionally distinct subsets , involving naïve and stem-cell memory cells ( sharing a similar phenotype ) , central-memory cells ( TCM ) , four different effector-memory ( TEM ) subsets and the TEMRA subset as described previously [21 , 53–55] . Please note that due to limited numbers of available PBMCs per patient we were not always able to do stainings with all the different antibody panels . This affects the data presented on granzyme K , granzyme B , Ki-67 , and CD95 expression ( which were stained in a separate panel ) , where we did not have sufficient samples to determine the expression of these markers by BKPyV VP1-specific CD8+ T cells in one NR patient , one Rlow patient and one Rhigh patient at t = pre-peak; three Rlow patients and three Rhigh patients at t = <6 months post-peak; and two Rlow patients at t = 2years post-peak . Expression of these markers could also not be measured in BKPyV LTAG-specific CD8+ T cells for one Rhigh patient and one BKVN patient at t = peak; nor in one Rlow patient at t = 2 years post-peak . Cytokine release after phorbol 12-myristate 13-acetate ( PMA ) /ionomycin stimulation was performed as described by Lamoreaux et al . [56] . In short , PBMC were thawed and rested overnight in suspension flasks ( Greiner ) in RPMI supplemented with 10% FCS , penicillin , and streptomycin ( culture medium ) . Samples were split into aliquots of two million cells . Each aliquot was stimulated with PMA ( 10 ng/ml ) and ionomycin ( 1 μg/ml ) in culture medium in the presence of CD107a FITC ( eBioscience ) ; αCD28 ( 15E8; 2 μg/mL ) , αCD29 ( TS 2/16; 1 μg/mL ) , brefeldin A ( Invitrogen; 10 μg/mL ) ; and GolgiStop ( BD Biosciences ) in a final volume of 200 μL for 4 hours ( PMA at 10 ng/mL/ionomycin at 1 μg/mL ) at 37°C and 5% CO2 in untreated , round-bottom , 96-well plates ( Corning ) . Subsequently , cells were incubated with a mix of PE- , APC- , and BV421-labeled tetrameric-complexes for two BKPyV VP1 epitopes and one BKPyV LTag epitope ( S1A Fig ) , followed by incubation with CD14 PE-CF594 , CD19 PE-CF594 , CD21 PE-CF594 , CD3 V500 , CD8 BV785 , and Live/Dead fixable red cell stain . Cells were then washed twice , fixed , and permeabilized ( Cytofix/Cytoperm reagent; BD Biosciences ) and subsequently incubated with the following intracellular mAbs: anti-IFNγ BUV 395 , anti-TNFα BV650 ( BD Biosciences ) , and anti—IL-2 PerCP-eFluor 710 ( eBioscience ) . Cells were washed twice; all aliquots of a sample were pooled and up to ten million PBMC per sample were measured on an LSRFortessa flow cytometer and analysed with FlowJo Version 9 . 3 . 3 software . Because of the relatively small study group size , non-parametric distribution was assumed . The two-tailed Mann-Whitney test was used to analyse differences between different patient groups . The Kruskal-Wallis test was used to simultaneuously compare all four study groups . To analyse HLA mismatches between different patient groups , we used chi-square testing and to compare all four study groups we used Fisher-Freeman-Halton Exact Testing . Analyses were done with IBM SPSS v24 . 0 . A p-value less than 0 . 05 was considered statistically significant .
|
In immunosuppressed renal transplant recipients ( RTRs ) , BKPyV frequently reactivates from latency and may cause severe interstitial nephritis in the allograft ( BKVN ) . Not only is there no effective treatment , it also not understood why BKVN arises in some RTRs but not in all . In the current study we investigated populations of CD8+ T cells targeting epitopes from structural and non-structural BKPyV proteins in RTRs over the course of transplantation . In contrast to RTRs who suffered from self-limiting reactivation of BKPyV , patients who developed severe viral reactivation and BKVN were found to have BKPyV-specific CD8+ T cells which did not , or less often differentiate into CD28− effector-memory cells during viral reactivation . Moreover , virus-specific CD8+ T cell activation and differentiation was not only impaired in the circulation , but possibly also in BKVN-affected renal allografts . In contrast to the CD8+ T cells in kidneys from three patients who did not develop BKVN , T cells in two BKVN-affected kidneys did not display typical cytotoxic effector traits . These findings suggest that impaired BKPyV-specific CD8+ T cell maturation in response to viral reactivation , possibly owing to inter-individual differences in sensitivity to immunosuppressive medication or to certain viral quasispecies , underlies the emergence of severe viral reactivation and BKVN .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"urinary",
"system",
"procedures",
"medicine",
"and",
"health",
"sciences",
"innate",
"immune",
"system",
"organ",
"transplantation",
"immune",
"cells",
"body",
"fluids",
"immune",
"physiology",
"cytokines",
"viral",
"transmission",
"and",
"infection",
"immunology",
"microbiology",
"cell",
"differentiation",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"developmental",
"biology",
"renal",
"transplantation",
"molecular",
"development",
"cytotoxic",
"t",
"cells",
"viral",
"load",
"transplantation",
"white",
"blood",
"cells",
"animal",
"cells",
"t",
"cells",
"viral",
"replication",
"hematology",
"immune",
"system",
"blood",
"cell",
"biology",
"anatomy",
"virology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] |
2016
|
Clinically Relevant Reactivation of Polyomavirus BK (BKPyV) in HLA-A02-Positive Renal Transplant Recipients Is Associated with Impaired Effector-Memory Differentiation of BKPyV-Specific CD8+ T Cells
|
Intracellular pathogens have complex metabolic interactions with their host cells to ensure a steady supply of energy and anabolic building blocks for rapid growth . Here we use the obligate intracellular parasite Toxoplasma gondii to probe this interaction for isoprenoids , abundant lipidic compounds essential to many cellular processes including signaling , trafficking , energy metabolism , and protein translation . Synthesis of precursors for isoprenoids in Apicomplexa occurs in the apicoplast and is essential . To synthesize longer isoprenoids from these precursors , T . gondii expresses a bifunctional farnesyl diphosphate/geranylgeranyl diphosphate synthase ( TgFPPS ) . In this work we construct and characterize T . gondii null mutants for this enzyme . Surprisingly , these mutants have only a mild growth phenotype and an isoprenoid composition similar to wild type parasites . However , when extracellular , the loss of the enzyme becomes phenotypically apparent . This strongly suggests that intracellular parasite salvage FPP and/or geranylgeranyl diphosphate ( GGPP ) from the host . We test this hypothesis using inhibitors of host cell isoprenoid synthesis . Mammals use the mevalonate pathway , which is susceptible to statins . We document strong synergy between statin treatment and pharmacological or genetic interference with the parasite isoprenoid pathway . Mice can be cured with atorvastatin ( Lipitor ) from a lethal infection with the TgFPPs mutant . We propose a double-hit strategy combining inhibitors of host and parasite pathways as a novel therapeutic approach against Apicomplexan parasites .
Toxoplasma gondii is an important intracellular pathogen causing disease in humans and animals . Most human infections are uncomplicated but the parasite persists and the chronic infection can be reactivated upon immunosuppression in patients undergoing organ transplants , cancer chemotherapy [1] , or AIDS due to HIV infection [2] . During pregnancy , infection causes congenital toxoplasmosis with serious consequences to the fetus [3] . There is also growing concern about outbreaks of severe ocular disease due to T . gondii in immunocompetent patients [4] . The parasite masterfully manipulates its host cell to insure favorable conditions for its survival and replication . T . gondii infection results in differential regulation of a variety of host signaling and metabolic pathways [5] . Many of these host changes are still not completely understood but it is quite likely that such modification of host pathways is essential for parasite growth and survival . Isoprenoids are lipid compounds with many important functions . The enzymes that synthesize and use isoprenoids are among the most important drug targets for the treatment of cardiovascular disease , osteoporosis and bone metastases and have shown promise as antimicrobials in a number of systems [6] . T . gondii lacks the mevalonate pathway for the synthesis of isoprenoid precursors that is used by mammals but harbors a prokaryotic-type 1-deoxy-D-xylulose-5-phosphate ( DOXP ) pathway in the apicoplast . This pathway generates isopentenyl diphosphate ( IPP ) and dimethyallyl diphosphate ( DMAPP ) . We recently demonstrated that the DOXP pathway is essential in T . gondii [7] . Knockout of 1-hydroxy-2-methyl-2- ( E ) -butenyl 4-diphosphate reductase ( LytB ) , which catalyzes the generation of IPP and DMAPP in the final step of the DOXP pathway , or of DOXP reductoisomerase ( DOXPRI ) , which catalyzes the second step of the DOXP pathway , were both lethal [7] . We also characterized the key enzyme of downstream isoprenoid synthesis in T . gondii , farnesyl diphosphate synthase ( TgFPPS ) [8] . Interestingly , we found it to be a bifunctional enzyme that can catalyze the condensation of IPP with three allylic substrates: DMAPP , geranyl diphosphate ( GPP ) , and farnesyl disphosphate ( FPP ) . The enzyme thus generates not only 15-carbon FPP but also 20-carbon GGPP [8] . A bifunctional FPPS has also been described in Plasmodium falciparum [9] . TgFPPS is inhibited by long alkyl chain ( lipophilic ) bisphosphonates , which are among the most active inhibitors of human GGPPS [10] , as well as by short chain bisphosphonates like risedronate ( aminobisphosphonates ) , which preferentially inhibit human FPPS . T . gondii engineered to overexpress TgFPPS requires considerably higher levels of bisphosphonates to achieve growth inhibition supporting the idea that the T . gondii enzyme is a target of bisphosphonates [11] . In this work we report that drugs acting on the mevalonate pathway , like statins , are active in vitro and in vivo against T . gondii . This is surprising as the parasite lacks this pathway . With the use of null mutants for the TgFPPS ( Δfpps ) we demonstrate why the parasite is sensitive to these inhibitors . We also show that the parasite is able to salvage some isoprenoid intermediates from the host while depending on its own synthetic machinery for others . Our results reveal a metabolic exchange between host and parasite that quite likely also occurs in other intracellular pathogens like Plasmodium or Cryptosporidium . To take advantage of these findings we propose a double-hit strategy combining inhibitors of both host ( statins ) and parasite ( bisphosphonates ) pathways . This strategy will allow leveraging the extensive clinical experience gained with statins towards the treatment of infections and potentially adapt it to other intracellular parasites .
FPPS is an essential component of the isoprenoid biosynthesis pathway in all cells studied so far . This enzyme synthesizes both FPP and GGPP in T . gondii and localizes to the mitochondria [8] . Previous work from our laboratory has shown that the T . gondii FPPS is inhibited by bisphosphonates , which also inhibit parasite growth . Considering the central role of this enzyme in the isoprenoid pathway we wanted to validate the entire pathway as potential target for chemotherapy . We approached this by creating a null mutant for the TgFPPS gene . We used the T . gondii Δku80 strain , which favors homologous recombination [12] , [13] . Our targeting construct was a large genomic cosmid recombineered to replace the gene with a drug resistance marker ( Fig . 1A ) [14] . After initial unsuccessful attempts , we were able to obtain null mutants when supplementing the medium with geranylgeraniol during the selection process . This requirement for geranylgeraniol for growth of mutant parasites is possibly because of their specific metabolite need during the stress of the transfection . We analyzed these mutant clones ( Δfpps ) by Southern ( Fig . 1B ) and western ( Fig . 1C ) blot and demonstrated the lack of both TgFPPS gene and protein . We isolated complemented clones by re-introducing the TgFPPS gene into the T . gondii genome ( Δfpps-cm1 and Δfpps-cm2; Figs . 1B , and 1C ) . We next introduced tandem tomato red fluorescent protein constructs into all strains ( Δku80-rfp , Δfpps-rfp , Δfpps-cm-rfp ) to measure parasite growth following the intensity of red fluorescence [15] . To our surprise , Δfpps mutant parasites were able to grow at a similar rate in fibroblasts ( the cells we routinely use for parasite culture , Fig . 1D ) , and formed plaques of similar number and size when compared to the parental and complemented strains ( Figure S1 , top row ) . Previous work has shown that T . gondii can enter macrophages by active invasion . However weakened or stressed parasites can be actively phagocytized by macrophages , resulting in parasite death making macrophages a more challenging host cell than fibroblasts [16] . We tested our mutants for their ability to grow in macrophages . Interestingly , Δfpps parasites showed a significant growth defect in these cells ( Fig . 1E , red line ) . We compared growth of our mutants in fibroblast vs . macrophages using a competition growth assay . For this , we mixed unlabeled Δfpps mutants with fluorescent parental ( Δku80-rfp ) or complemented strains ( Δfpps-cm-rfp ) at a 20∶1 starting ratio . Fig . 1F and 1G show that parental and complemented clones are able to rapidly outgrow mutant parasite in macrophages ( Mac , blue lines ) while they grow at a similar rate in fibroblasts ( Fib , black lines ) . Our interpretation of these results is that whether the enzyme is required or dispensable for growth of the parasite depends on the specific host cell and host-parasite interaction . In this context we note that growth of Δfpps mutants was well supported in low passage primary fibroblasts , as measured by plaque assay ( Fig . S1 , top row ) , but limited in aging fibroblast cultures ( fibroblast with ∼40 passages; Fig . S1 , bottom row ) . This suggests that the parasite isoprenoid metabolism may not only be sensitive to the cell type infected but also to its physiological and/or metabolic state . We next addressed whether Δfpps mutants would be less virulent in vivo . The RH strain is hypervirulent , which can make it difficult to appreciate modest attenuation . We observed a difference in virulence when infecting mice with low parasite numbers ( 5–10 ) while higher doses ( 15–100 ) lead to death at 9–10 days ( data not shown ) . We wondered whether the use of a less virulent strain would better highlight the difference in virulence between mutant and parental cell lines . We created conditional mutants using the described TATi cell lines [17] . There are two advantages for using these cells . First , the reduced virulence of the parental cell line allows the use of 104–105 parasites to infect mice . Second , these mutants are maintained in culture expressing an extra copy of TgFPPS , prior to suppression with anhydrotetracycline ( ATc ) thus avoiding preadaptation . We first expressed a regulatable copy of the TgFPPS in the TATi parental cell line ( Fig . S2A ) and created the cell line FPPS/FPPSi . In a clonal line derived from these cells we disrupted the endogenous TgFPPS gene as detailed before ( see legend for Fig . S2 ) and generated ΔFPPS/FPPSi mutants . We established ATc regulation and gene deletion by western and Southern blot analyses , respectively ( Fig . S2B and S2C ) . Plaque assays in fibroblasts in the presence of ATc showed no difference in the number and size of plaques ( Fig . S2D ) . In contrast , a highly significant difference in growth was observed when parasites were allowed to infect macrophages ( Fig . S2E ) , as seen before with the Δfpps mutants ( Fig . 1E ) indicating a fitness defect only evident under stressful conditions . With the purpose to define a dependable inoculum to use for virulence studies , we first established a protocol in which we passed our Tati-derived strains through mice and performed in vivo titration experiments ( see Materials and Methods and Fig . S3 ) . This treatment increased consistency dramatically and we found that using an inoculum defined in this way resulted in reproducible virulence outcomes . To establish whether FPPS knockdown affects the ability of these parasites to cause disease , mice were infected with 10 , 000 FPPS/FPPSi or ΔFPPS/FPPSi tachyzoites ( Fig . S3 ) ( a reproducible number found after our in vivo titration experiments ) . Mice challenged with the FPPS/FPPSi strain succumbed to the infection even if they were given ATc in the water ( Fig . S3 , black lines ) . In contrast , mice infected with the ΔFPPS/FPPSi and receiving ATc survived the infection while mice infected with the same parasites but given a placebo were susceptible to infection ( Fig . S3 , compare red lines ) . With the aim of understanding how the Δfpps mutant parasites manage to survive without the production of essential isoprenoids we measured growth of mutant Δfpps parasites and their parental strain after being deprived of host cells for a determined length of time . We exposed mutant , parental and complemented parasites for 30 min to an extracellular buffer and for a more accurate readout we switched to a plaquing efficiency protocol as described [18] in which there is only a 30-min contact interval between parasite and host ( Figure 2A ) . Plaques were counted after 7 days of incubation . We observed that the number of plaques was significantly lower for the Δfpps parasites after being exposed to these stress conditions . We also measured ATP levels of parasites incubated in extracellular media for one hour . No difference in the ATP levels was observed in recently egressed parasites but there was a significant decrease in the Δfpps mutants after incubating them for one hour in extracellular buffer with glucose ( Figure 2B ) . These results indicate that the Δfpps parasites do have a defect in energy generation , which is not evident under the protected intracellular environment . However , this defect becomes relevant when the parasite is outside the host and we were able to increase it by incubating them for an extended length of time before letting them continue with its lytic cycle ( Figure 2A ) . A possible cause of this defect could be the synthesis of ubiquinone , isoprenylated cofactor of the mitochondrial respiratory chain , which may be more important when the parasites are extracellular . We measured mitochondrial membrane potential of Δfpps and the parental Δku80 parasites using JC1 , a lipophilic , cationic dye that accumulates in mitochondria in a membrane potential dependent fashion and that changes color from green to red as it accumulates [14] ( Figure 2C , upper panels ) . The lower panels show that for the mutant parasites JC1 stays mostly green indicative of a partially depolarized mitochondrial membrane ( Figure 2C , lower panels ) . We also used flow cytometry to quantify the effect of knocking out the TgFPPS gene ( Δfpps ) and compare it with the parental strain Δku80 ( Figure S4 ) . We observed a dramatic drop of mitochondrial fluorescence in Δfpps parasites ( 56 . 2% , compared to 85 . 2% ) . This indicates a mitochondrial defect that is not important for intracellular life but becomes accentuated when the parasites are deprived of host cells . Our surprising findings could imply that intracellular parasites salvage isoprenoids from their host and that we impinge on this ability by cultivation in different cells . To test this hypothesis , we performed two labeling experiments testing different conditions . We first labeled infected fibroblasts with 14C-glucose ( Figure 3A ) . Under these conditions , radioactive glucose is available to both host and parasite to label isoprenoids generated by host and parasite specific de novo synthesis pathways . In the second experiment the strategy was to first label the fibroblasts with 14C-glucose , remove unincorporated label by washing the cells with fresh medium and only then infect with parasites ( Figure 3B ) . In both settings we compared parental ( Δku80 ) , mutant ( Δfpps ) , or complemented ( Δfpps-cm ) T . gondii . Parasites were purified through several filtration steps before isoprenoid extraction . Parasite isoprenoids were isolated by solvent extraction of purified tachyzoites and analyzed by thin layer chromatography and autoradiography ( TLC ) . When infected cells were labeled ( Figure 3A ) the most abundant isoprenoids were FPP , GGPP , an intermediate co-migrating with a 25C standard ( 25 C ) , and a longer unidentified derivative co-migrating with a 35C standard ( long prenyl diphosphate; LPP , 35C ) . The results indicate that mutant parasites ( Δfpps ) have levels of intermediates similar to the parental strain despite the lack of FPPS ( differences between labeled compounds were not statistically significant , n = 3 , data not shown ) . Figure 3B shows the isoprenoids obtained from the parasite after labeling only the host cells followed by infection with unlabeled parasites . FPP and GGPP were still present and labeled in the extracts obtained from mutant parasites despite the fact that they lack the enzyme required for their synthesis ( Figure 3B , Δfpps ) . However , labeling of the longer chain product was stronger in extracts from Δfpps mutants . This likely indicates that the parasite synthesizes these longer chain products using both host and its own precursors , and that labeling via the host pathway becomes more evident in the absence of parasite synthesis ( Figure 3B , Δfpps ) . The parental and complemented cells did not show this labeling arguing that it is generated in the parasite using unlabeled precursors . The results were quantified by calculating the ratio of labeling for this long chain product to that of GGOH and is displayed with bar graphs in Figs . 3C and 3D . This analysis shows no difference between mutant and wild type when parasite and host cells are simultaneously labeled with 14C-glucose ( Figure 3C ) . However , the ratio was significantly higher for the mutant parasite when only the host cells were prelabeled ( Fig . 3D ) . Taken together these results suggest that mutant parasites lacking their own production of FPP and GGPP import these intermediates from the host ( pre-labeled with 14C in our experimental set-up ) and convert them into the long chain isoprenoid . Under similar experimental conditions , when analyzing the isoprenoid products made by the parental strain , the labeling of this long chain isoprenoid product becomes diluted as a consequence of the endogenous production of unlabeled FPP and GGPP by the TgFPPS . If the parasite is taking up FPP and/or GGPP from the host , then inhibiting the synthesis of these host compounds may affect parasite growth . We directly tested this idea using an inhibitor of hydroxymethyl glutaryl-CoA reductase ( HMG-CoA reductase ) , the rate-limiting enzyme of the host mevalonate pathway ( this pathway is absent in T . gondii ) . We tested atorvastatin ( Lipitor ) in tissue cultures ( Fig . 4A and 4B , black lines ) and found that atorvastatin is able to inhibit growth of the parental strains with an IC50 of ∼40 µM . We thought that this modest level of efficacy points to other sources of FPP and GGPP for the parasite , in particular its own synthesis . We hypothesized that the Δfpps mutants , unable to produce FPP and GGPP should be more sensitive to the inhibition of the host by atorvastatin . This is indeed what we observed when testing the drug against the mutant parasites ( Fig . 4A and B , red lines ) and we calculated an IC50 of 2 µM ( 20 times lower than the efficacy against the parental cell lines ) . This effect of atorvastatin is specific to its inhibition on the production of isoprenoid metabolites because it was possible to rescue parasite growth by adding geranylgeraniol to the medium ( Fig . 4B , red lines: Δfpps and Δfpps+GGOH ) . To investigate whether atorvastatin inhibits parasite growth mainly as a result of impaired cholesterol synthesis we tested WC-9 , a known inhibitor of squalene synthase ( SQS ) [19] . We found WC-9 to inhibit parasite growth with an IC50 of 4–5 µM ( Fig . 4C ) . T . gondii does not encode SQS and acquires cholesterol from its host [20] , [21] , [22] . We therefore attributed the effect of WC-9 to its action against the host pathway . Importantly , we found no difference in the WC-9 susceptibility of Δku80 , Δfpps and Δfpps-cm parasites ( Fig . 4C ) . This suggests that WC-9 acts downstream of the formation of FPP and GGPP , and that inhibition of cholesterol synthesis is not the most important anti-parasitic effect of statin treatment . This is in agreement with previous findings that suggested that the parasite relies on LDL trafficking rather than de novo synthesis by the host cell to satisfy its cholesterol requirement [20] , [21] , [22] . We next tested another statin ( mevastatin ) on FPPS/FPPSi or ΔFPPS/FPPSi parasites grown in the presence or absence of ATc ( Figs . 4D ) . FPPS/FPPSi tachyzoites express an extra copy of TgFPPS ( Fig . S2B ) and possesses higher FPPS activity ( not shown ) . There was a reverse correlation between mevastatin inhibition and the expression level of TgFPPS ( Fig . 4D ) . ΔFPPS/FPPSi parasites were the most susceptible in the presence of ATc ( IC50∼4 µM mevastatin ) while FPPS/FPPSi cells with an extra copy of the TgFPPS gene were resistant to concentrations up to 18 µM ( Fig . 4D ) . The effect of mevastatin was rescued by supplementation of the medium with 1 µM geranylgeranyol ( Fig . 4E , compare red and blue lines ) , again supporting its direct effect on the production of FPP and GGPP . We also tested the efficacy of atorvastatin treatment against T . gondii infection of mice using wild type RH strain . Fig . 5A shows a summary of 3 experiments using groups of 5 mice treated with different doses of atorvastatin . While 100% of control mice died between 9–13 days post-infection , 80% of mice treated with the higher 40 mg/kg/day dose , survived more than 30 days . Note that this is not an excessive drug dose but the standard concentration of atorvastatin commonly used and well tolerated in mouse experiments [23] , [24] . An atorvastatin ED50 of 32 . 3 mg/kg per day was calculated ( Fig . 5A ) . We also were interested in comparing the efficacy of atorvastatin against the infection of mice with Δku80 , and Δfpps cells . We infected mice with a lethal dose of parasites ( parental and mutants ) to highlight the effect of atorvastatin against infection with the Δfpps clone . Fig . 5B shows that atorvastatin is highly effective at treating mice infected with Δfpps parasites: 9 of 10 mice survived the infection when treated with atorvastatin , while 8 of 10 mice died in the absence of atorvastatin . To establish further that knockdown of TgFPPS make T . gondii infection more amenable to treatment with atorvastatin we infected mice with a lethal dose of 100 , 000 ΔFPPS/FPPSi or FPPS/FPPSi tachyzoites and treated with ATc in their drinking water ( Fig . 5C ) . This high parasite dose was lethal even when infecting with ΔFPPS/FPPSi ( compare with Fig . S3 for which we used 10 , 000 parasites , ten fold difference in dose ) [25] . Most mice infected with FPPS/FPPSi and treated with atorvastatin succumbed to this high infection ( Fig . 5C , black lines ) . In contrasts most mice infected with ΔFPPS/FPPSi were cured by atorvastatin when the mutation was induced by ATc treatment ( Fig . 5C , red lines ) . T . gondii appears to be able to rely on both synthesis and salvage of isoprenoids . Δfpps mutants are more dependent on salvage . Could this be exploited pharmacologically by combining inhibitors of TgFPPS with atorvastatin ? Bisphosphonates are known inhibitors of FPPS and have shown antiparasitic activity [11] . We chose to test zoledronic acid [26] because our previous work had identified this compound as the bisphosphonate with the highest specificity against TgFPPS , and its activity decreased significantly when we overexpressed the parasite enzyme [11] . To evaluate interaction between atorvastatin and zoledronate , we mixed both drugs at different concentrations following a protocol designed for testing synergy [27] . This protocol measures and calculates the IC50 of one drug in the presence of subtherapeutic concentrations of the second drug [27] . The results were plotted in an inhibition isobologram using IC50s of individual drugs and of five different drug combinations ( Fig . 6A ) . The resulting curve is concave for atorvastatin and zoledronic acid and thus indicative of synergistic drug interaction ( Fig . 6A ) . FPP and GGPP production in the parasite requires the isoprenoid precursors IPP and DMAPP . We therefore next wanted to test whether atorvastatin would interact with fosmidomycin , a specific inhibitor of the DOXP pathway . T . gondii is insensitive to fosmidomycin because the drug is not able to cross the parasite membrane [7] . However a T . gondii transgenic parasite that expresses the bacterial transporter glycerol-3-phosphate transporter ( GlpT ) capable of importing fosmidomycin , is sensitive to fosmidomycin [7] ( Fig . 6B ) . We assessed the growth of these parasites in the presence of 50 µM atorvastatin and 0 . 78 µM of fosmidomycin . This represents the IC10 for fosmidomycin and this low concentration was deliberately chosen to be able to detect drug interaction . Individually these drugs affected parasite growth as expected , approximately 50% inhibition with 50 µM atorvastatin and very little inhibition with 0 . 78 µM fosmidomycin . Interestingly , combining both drugs abolished parasite growth , indicating strong interaction also between atorvastatin and fosmidomycin . We tested the interaction between atorvastatin and fosmidomycin in these transgenic parasites by a simplified checkerboard technique [27] and calculated the fractional inhibitory concentration ( FIC ) index to be 0 . 36 , confirming synergistic interaction ( FIC<0 . 5 ) [28] , [29] . This assay provided additional strong evidence that the parasite , although capable of generating its own isoprenoids , also depends on the host isoprenoids for continuous growth and successful infection . Our results show that therapeutic strategies aimed at interfering with both parasite and host isoprenoid synthesis could provide a higher rate of success in curing T . gondii infections .
Our work reveals a crucial metabolic interaction between the intracellular pathogen T . gondii and its host cell to secure the parasite's access to isoprenoids . Isoprenoids are essential for all cells and in most Apicomplexans their five carbon precursors are produced by the apicoplast [30] , [31] . The synthesis of these precursors is now viewed as the most important function of the apicoplast and the reason the organelle was maintained long after the loss of photosynthesis [32] . Genetic analysis in T . gondii demonstrates that loss of the apicoplast isoprenoid pathway is lethal and mimics complete loss of apicoplast metabolism [7] , [14] . Inhibiting this pathway with the antibiotic fosmidomycin is effective against Plasmodium , Babesia , and against Toxoplasma ( once parasites are engineered to take up the drug ) [7] , [33] , [34] . Most intriguingly , in Plasmodium falciparum cells cured of their apicoplasts by antibiotic treatment targeting plastid translation can nonetheless be continuously maintained in culture when the media are supplemented with high concentrations of IPP [35] . Overall these studies suggest that the synthesis of IPP and DMAPP by the parasite is essential and cannot be circumvented by salvage from the host under physiological conditions . This makes our observation that the parasite enzyme catalyzing the next step in the isoprenoid pathway – the synthesis of FPP and GGPP from IPP and DMAPP is dispensable for T . gondii in fibroblasts all the more surprising . FPPS-catalyzed reactions are essential in most organisms studied so far and are important drug targets . T . gondii is not only able to make its own isoprenoids but can also import from the host cell ( Fig . 7 ) . We note that this ability to salvage appears not to be universal but restricted to certain compounds ( Fig . 7 ) . At the moment it is not fully understood whether this difference is due to difference in transport capability of the parasite or availability and abundance of the metabolites in the host cell . However , in our experiments we measured a strong impact of the host cell environment for FPP and GGPP . Extracellular parasites or parasites infecting macrophages rather than fibroblasts show more pronounced defects upon loss of the synthesis capacity . The intracellular survival of T . gondii depends on its unique ability to invade cells actively . Active invasion is fundamentally different from phagocytosis and requires parasite motility [16] . When extracellular parasites are incubated in PBS , their ability to invade cells actively rapidly declines , and they are mostly internalized by phagocytosis with parasites engulfed in phagosomes , which fuse with endosome/lysosomes and are further digested [16] . Parasite fitness is essential for its ability to actively invade host cells and/or escape from the phagosome . Lack of endogenous production of FPP and GGPP by T . gondii renders them less able to grow in macrophages . This could be the result of a fitness defect or because of a shortage of metabolites in macrophages or a different mechanism of transport of isoprenoids in these cells . How dependent is the parasite on isoprenoid salvage under normal conditions with its synthesis capability intact ? Our labeling studies show robust import of host cell-synthesized isoprenoids even in wild type parasites . Import is also supported indirectly by microarray studies of T . gondii-infected fibroblasts that revealed a significant induction of genes encoding enzymes of the mevalonate pathway following infection [5] , [36] including the rate-limiting enzyme HMG-CoA reductase , and FPPS [5] . Previous work has shown that T . gondii does not synthesize cholesterol and imports it from the host low–density lipoprotein ( LDL ) [20] , [21] , [22] . It is possible that the inhibition of cholesterol synthesis by statins results in reduced parasite invasion or reduced parasite growth . Interestingly , a recent study has shown that atorvastatin treatment of endothelial cells reduced cytoadherence of Plasmodium falciparum [37] . We consider that inhibition of host cholesterol synthesis is unlikely as the reason for the effect of statins because of three reasons . First , the isoprenoid intermediate geranylgeraniol was able to rescue almost completely the growth inhibition by two statins , atorvastatin and mevastatin . Second , growth inhibition by an SQS inhibitor that blocks the pathway downstream to the production of FPP and GGPP was not enhanced in the Δfpps mutants , as observed with atorvastatin and mevastatin . And third , statins do not reduce overall plasma cholesterol levels in mice as they do in humans ( due to low levels of low density lipoproteins in rodents ) [38] , [39] . In addition , it has been demonstrated that host cell cholesterol production has no significant effect on parasite replication and that the bulk of parasite cholesterol requirement can be satisfied by exogenous cholesterol from low-density-lipoprotein delivered to the parasitophorous vacuole [20] . Our results with the squalene synthase inhibitor strongly support that conclusion . It is interesting to note that the growth in macrophages of Salmonella enterica serovar Typhimurium , which also lacks a mevalonate pathway , is inhibited by statins and this inhibition is not due to a deficient production of sterols but of intermediates of the pathway between mevalonate and squalene 2 , 3-oxide [38] . It would be very interesting to investigate whether these intermediates are also FPP and GGPP as in the case of T . gondii . Our labeling results indicate that T . gondii may use its own enzymes to make specific long chain isoprenoids . We previously reported that TgFPPS localizes to the mitochondria [8] . Our results showed that loss of TgFPPS resulted in alteration of the mitochondrial membrane potential , and rapid decrease in the ATP levels of extracellular parasites . These results suggest that TgFPPS functions to make FPP and GGPP in the mitochondrion as precursors for long chain isoprenoids and ubiquinone synthesis . Therefore we could deduce that TgFPPS plays an important role in maintaining mitochondrial function . This appears to be crucial for the parasite during its extracellular stage . Our results suggest a requirement for oxidative phosphorylation for generation of ATP in extracellular parasites . This is consistent with our previous results showing active oxidative phosphorylation in extracellular parasites [40] and a recent report showing that oxidative phosphorylation is responsible for >90% ATP synthesis in extracellular tachyzoites [41] . The deficient synthesis of ubiquinone precursors would not affect tachyzoites when they are intracellular while seriously impeding extracellular parasites as a consequence of rapid depletion of ATP , which is needed for gliding motility and invasion . Taken together , we show that T . gondii has a versatile system for its isoprenoid needs . During its replicative stage with needs for large quantities of isoprenoids , the parasite is able to manipulate the host and salvage isoprenoids . However , under stressful situations the parasite is able to provide by itself and this was emphasized when exposing Δfpps or conditional knockout mutants to challenging conditions like infection in vivo , growth in macrophages , growth in metabolic inactive host cells , or during extended extracellular life . These findings show that the endogenous activity of TgFPPS , while low compared to other FPPs ( Table S1 ) is needed under stress or for other functions . This ability of the parasite to use not only its own metabolites but also to manipulate the host cell metabolism and salvage its products makes it a challenge for drug therapy . However , in the case of isoprenoid metabolism this split reliance may also prove to be an opportunity as it can build on the massive investment made into controlling this pathway pharmacologically in the host . Our work demonstrates that inhibition of the host mevalonate pathway enhances the impact of blocking the parasite isoprenoid pathway and we propose a double hit strategy that combines inhibitors of the parasite enzyme with host isoprenoid pathway inhibitors . We tested combinations of two approved and widely used drugs , zoledronic acid ( Zometa ) and atorvastatin ( Lipitor ) and showed synergism in the inhibition of T . gondii growth . We demonstrated that impinging on host or parasite isoprenoid synthesis reduces parasite virulence but that blocking both produces stronger effects and affords considerable protection . This strategy could prove even more promising when tested in other parasites . For example our experiments combining fosmidomycin with atorvastatin suggest that atorvastatin may boost the efficacy of fosmidomycin as an antimalarial . This combination may also make it more difficult for the parasite to develop resistance extending the useful life of the drug . Our strategy will benefit from the extensive clinical knowledge on both statins and bisphosphonates and this knowledge will facilitate their use for the treatment of other infections .
Mice experiments in this work followed a reviewed and approved protocol by the Institutional Animal Care and Use Committee ( IACUC ) . Animal protocols followed the US Government principles for the Utilization and Care of Vertebrate animals . The protocol was reviewed and approved by the University of Georgia IACUC ( Protocol number A2012-3-010 ) . Oligonucleotide primers were obtained from Integrated DNA Technologies ( Coralville , IA ) . Taq DNA polymerase , and restriction enzymes were from Invitrogen or New England Biolabs . Plasmid miniprep and maxiprep and gel kits were from Qiagen Inc . ( Chatsworth , CA ) . IPP , DMAPP , GPP , FPP , GGPP were from Isoprenoids , LC ( FL , USA ) . [4-14C] Isopentenyl diphosphate triammonium salt ( 55 . 0 mCi/mmol ) and [14C ( U ) ]-glucose ( 319 mCi/mmol ) were from PerkinElmer Life Sciences . Atorvastatin ( Lipitor ) was from Pfizer . Zoledronic acid and WC-9 were a gift from Dr . Juan B . Rodriguez , University of Buenos Aires . Fosmidomycin was a gift from Dr . Yongcheng Song ( Baylor College of Medicine ) . All other reagents were analytical grade . Tachyzoites of T . gondii RH strain were cultured in human fibroblasts or hTERT cells in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 1% fetal bovine serum , 2 mM glutamine , and 1 mM pyruvate , and purified as described before [8] . A basic electroporation protocol was used for transfection . Briefly , 107 recently released parasites and 20 µg of sterilized cosmid ( see below ) or plasmid DNA were mixed in a 2-mm gap electroporation cuvette . Electroporation was performed using a Genepulser Xcell from BioRad and after 15 min of recovery the parasites were allowed to infect fibroblasts . Stable transfectants were selected with 20 µM chloramphenicol and cloned by limited dilution . For the TgFPPS cDNA complemented cell line , the Δfpps parasites were transfected with a TgFPPS cDNA construct [8] , then cultured in 15 µM atorvastatin for 4 passages . These parasites were sub-cloned by limited dilution medium containing 5 µM atorvastatin . Subcloned parasites were analyzed by PCR , Southern and western blot . All the clones that survived to atorvastatin selection have the TgFPPS cDNA stably integrated . The complemented clone used for the experiments was named Δfpps-cm . The protocol for creating TgFPPS conditional deletion mutants is described in the supporting information ( Fig . S2 legend ) . The cosmid PSBLI36TV was obtained from L . David Sibley ( Washington University ) . The knockout cassette from pH3CG was amplified by using the primers ( 5′-GCGGCCACCGTCCATAATTGCAAAAATGGAGCGGCTGTGTTTCCGTCTCCTCGACTACGGCTTCCATTGGCAAC-3′ and 5′-CTATTTCTGCCGTTT GTGGAGCCTCCCGAGGACGAGGCCGAAGAAGGCCTATACGACTCACTATAGGGCGAATTGG-3′ ) . The TgFPPS gene targeting cosmid construct was obtained by recombineering in E . coli EL250 as described previously [14] ( Fig . 1A ) . Plaque assays and growth assay of tagged parasites were performed as described before [7] . Plaquing efficiency was measured infecting hTERT monolayers with 1 , 000 parasites per well and allowing contact with host cells for 30 min . At this point , wells are washed with PBS , fresh media added and parasites allowed to grow for 4–7 days , fixed and stained with crystal violet [18] . Parental and mutant strains of T . gondii were transfected with a plasmid containing a tandem tomato RFP gene and red fluorescent parasites were sorted and subcloned by FACS analysis . Δku80-rfp , Δfpps-rfp and Δfpps-cm-rfp cell clones were obtained . Growth competition assays were performed by mixing strains: Δfpps parasites with Δku80-rfp and Δfpps-cm-rfp cell lines at 20∶1 ratio ( 5% of red cells in the mixture ) . These parasite mixtures were used to infect fibroblasts or macrophages . 1×106 parasites were inoculated in each passage . Percentage of red cells at each passage was calculated using a standard curve generated by measuring the fluorescence intensity for a fixed number of cells . T . gondii genomic DNA was digested with SalI , separated in a 0 . 8% agarose gel , and transferred to a nylon membrane . The DNA probe was generated by PCR with primers ( 5′-TGACGCGCTGAGCAGTGGTGAGCA-3′ and 5′-AGCCATTTCAACTTCAAACCGCA-3′ ) . The purified PCR product was 32P labeled by random priming . Western blots were done using an affinity purified rabbit polyclonal antibody raised against TgFPPS at 1∶1500 in PBS-T . The secondary antibody was horseradish peroxidase-conjugated goat anti-rabbit and immunoblots were visualized on blue-sensitive x-ray film by using an ECL detection kit . Purified parasites were washed in Ringer ( 155 mM NaCl , 3 mM KCl , 1 mM MgCl2 , 3 mM NaH2PO4-H2O , 10 mM Hepes , pH 7 . 3 , 10 mM glucose ) and resuspended in Buffer A with glucose ( BAG , 116 mM NaCl , 5 . 4 mM KCl , 0 . 8 mM MgSO4 , 5 . 5 mM D-glucose and 50 mM Hepes , pH 7 . 2 ) at 1–5×108 cells/ml . The parasite suspension was incubated at 37°C for 1 hour . The suspension was extracted with perchloric acid as described previously [42] . Briefly , the parasite suspension was centrifuged and resuspended in 100 µl of BAG and mixed with 300 µl of 0 . 5 M HClO4 and incubated in ice for 30 min . The supernatant was neutralized with 0 . 72 M KOH/0 . 6 M KHCO3 and used immediately or stored at −20C for ATP measurement . The kit A22066 from Invitrogen was used to measure ATP . Two labeling strategies were used . The first one consisted on labeling both host cells and parasites . Briefly , hTERT-fibroblasts were first infected with fresh tachyzoites and grown in DMEM medium containing 14C-glucose until the natural release of parasites . The second strategy consisted of labeling only host cells and infecting afterwards . Host cells were grown in medium containing 14C-glucose and before infection the monolayer was thoroughly washed and fresh medium containing glucose was added . For both experiments , released tachyzoites were purified by several filtration steps ( 8 , 5 , and 3 µM membranes ) , to ensure the absence of host cells , and lipids extracted in chloroform/methanol at 4°C overnight . After filtration followed a saponification step ( with KOH and ethanol ) and the radioactive prenyl products in the mixture were hydrolyzed to the corresponding alcohols with alkaline phosphatase at room temperature , overnight . The resulting alcohols were extracted with hexane and separated on a HP-TLC-RP18 plate using acetone∶H2O ( 10∶1; v/v ) as the moving phase . Standard prenyl alcohols were run in parallel and were visualized by iodine vapor . Radioactivity of the products was also measured by autoradiography or phosphorImaginer analysis . All parasite clones were grown in fibroblasts using similar conditions to those used for the RH strain . For in vitro drug testing , confluent hTERT monolayers in 96-well plates were first prepared with phenol-free medium containing the drugs serially diluted and infected with 4 , 000 parasites per well . The plates were incubated at 37°C and the fluorescence measured every day . Regression analysis and IC50 calculations were performed using SigmaPlot 10 . 0 . Isobolograms were constructed by plotting the IC50 of one drug against the IC50 of the other for each of five drug ratios , with a concave curve indicating synergy , a straight line indicating addition , and a convex curve indicating antagonism . For simplified checkboard studies , drugs were mixed in fixed ratios of their respective IC50s and dose-response curves generated from serial dilutions carried out in triplicate . Results were expressed as the sums of the fractional inhibitory concentration ( sum FIC = IC50 of drug A in mixture/IC50 of drug A alone ) + ( IC50 of drug B in mixture/IC50 of drug B alone ) , as described by Berenbaum [43] . Sum FIC values indicate the kinds of interactions as follows: <0 . 5 , synergy; 1 , addition; >2 , antagonism . For in vivo infection with T . gondii , fresh tachyzoites were harvested , washed with PBS twice , and resuspended in PBS before inoculation . Female Swiss Webster or BALBc mice were injected with 5–20 tachyzoites of the RH strain i . p . in a 200 µl PBS final volume or 10 , 000–100 , 000 tachyzoites of the TATi strain in a similar volume . When using low parasite numbers , plaque assays were performed with the parasite suspensions used to inoculate mice to ensure that the number represented the number of viable and infectious parasites . Because initial in vivo experiments with cultured TATi-derived strains gave inconsistent virulence results , we developed a protocol by which we first infected mice with a high dose ( 106 parasites ) of parasites ( parental , knock-in and knock outs ) and collected the peritoneal fluid ( containing tachyzoites ) five days p . i . This suspension was used to infect a confluent flask of fibroblasts and allowed to grow and lyse . The supernatant from these flasks containing tachyzoites was collected , centrifuged and parasites resuspended in the appropriate media to prepare aliquots for freezing in liquid nitrogen . For each experiment , one vial was thawed and passed once through tissue culture before used for infection . Tati-derived strains showed a remarkable recovery in virulence with this treatment . We performed titration experiments and determined that 10 , 000 parasites of Tati-derived strain ( no ATc ) are lethal to mice 9–10 days p . i . . Results shown in Figs . S3 and 5C were obtained with parasites previously treated as described . Drugs were dissolved in phosphate-buffered saline ( PBS ) containing approximately 2% DMSO , at pH 6 . 8 , and were also inoculated i . p . Treatment was initiated 6 hours after infection and administered daily i . p . for 10 days .
|
Toxoplasma gondii is an obligate intracellular parasite and is not able to replicate outside the host cell . The parasite lives in a specialized parasitophorous vacuole in contact with the host cytoplasm through the parasitophorous vacuole membrane . It is highly likely that a very active exchange of metabolites occurs between parasite and host cell . We present evidence for this exchange for isoprenoids , abundant lipidic compounds essential to many cellular processes including signaling , trafficking , energy metabolism , and protein translation . Our work shows that intracellular T . gondii tachyzoites are able to salvage farnesyl diphosphate ( FPP ) and/or geranylgeranyl diphosphate ( GGPP ) from the host , and the parasite is able to grow even when its endogenous production is shut down . However , when extracellular , the parasite depends entirely on its own production of isoprenoids . We propose to use a combination of inhibitors that would hit both the host and the parasite pathways as a novel therapeutic approach against Toxoplasma gondii that could also work against other Apicomplexan parasites .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Toxoplasma gondii Relies on Both Host and Parasite Isoprenoids and Can Be Rendered Sensitive to Atorvastatin
|
The biology of adult tsetse ( Glossina spp ) , vectors of trypanosomiasis in Africa , has been extensively studied – but little is known about larviposition in the field . In September-November 1998 , in the hot-dry season in Zimbabwe’s Zambezi Valley , we used artificial warthog burrows to capture adult females as they deposited larvae . Females were subjected to ovarian dissection and were defined as perinatal flies , assumed to have entered burrows to larviposit , if oocyte sizes indicated >95% pregnancy completion . Perinatal flies were defined as full-term pregnant if there was a late third instar larva in utero , or postpartum if the uterus was empty . All other females were defined as pre-full-term pregnant ( pre-FT ) . Of 845 G . m . morsitans captured , 91% ( 765 ) were female and 295/724 ( 41% ) of females dissected were perinatal flies . By contrast , of 2805 G . pallidipes captured only 71% ( 2003 ) were female and only 33% ( 596/1825 ) of females were perinatal . Among all perinatal females 67% ( 596/891 ) were G . pallidipes . Conversely , in burrows not fitted with traps – such that flies were free to come and go – 1834 ( 59% ) of pupae deposited were G . m . morsitans and only 1297 ( 41% ) were G . pallidipes . Thus , while more full-term pregnant G . pallidipes enter burrows , greater proportions of G . m . morsitans larviposit in them , reflecting a greater discrimination among G . pallidipes in choosing larviposition sites . Catches of males and pre-FT females increased strongly with temperatures above 32°C , indicating that these flies used burrows as refuges from high ambient temperatures . Conversely , catches of perinatal females changed little with maximum temperature but declined from late September through November: females may anticipate that burrows will be inundated during the forthcoming wet season . Ovarian age distributions of perinatal and pre-FT females were similar , consistent with all ages of females larvipositing in burrows with similar probability . Artificial warthog burrows provide a novel method for collecting tsetse pupae , studying tsetse behaviour at larviposition , assessing the physiological status of female tsetse and their larvae , and of improving understanding of the physiological dynamics of terminal pregnancy , and population dynamics generally , with a view to improving methods of trypanosomiasis control .
Adult tsetse flies ( Glossina spp , Diptera: Glossinidae ) are the vectors of human and animal trypanosomiasis in Africa and , as such , have been the object of intense study since the early 20th century and the source of an extensive literature on their field biology . Our knowledge of the ( briefly free-living ) larval and pupal stages is , by contrast , limited mainly to laboratory studies [1 , 2 , 3 , 4 , 5 , 6 , 7] . Studies on the physiology of reproduction in tsetse[8] , showed that they have a very unusual reproductive system , termed adenotrophic viviparity . Unlike most Diptera , which lay large numbers of small eggs that hatch to produce free-living larvae , tsetse have just two ovarioles in each of the left and right ovaries and only produce one large egg at a time . The egg is retained in the uterus , fertilized there by sperm stored in the female’s spermathecae , and hatches after 4–6 days to produce a first instar larva . This larval stage , and two subsequent instars , are fed via a highly modified uterine or milk gland producing ultimately a late third instar larva that often constitutes more than 50% of the female’s total body mass . In the field a full-term third instar larva is typically deposited in sand , or soft soil , covered with leaf litter . The larva burrows a few centimetres into the substrate and rapidly forms around itself a hard , waterproof , puparial case of chitin . Inside this puparial case the fly goes through the larval , pre-pupal , pupal and pharate adult stages before emerging as the teneral adult fly some weeks after larviposition , the period depending on temperature . Whereas this process has been studied in the laboratory , it is extremely unusual to observe female tsetse larvipositing in the field: indeed it is uncommon to find female tsetse with a full-term larva in utero and this difficulty has led to some uncertainty regarding the physiological status of female flies at terminal stages of pregnancy [9] . Field studies have accordingly been limited to finding puparia that have already been deposited . This involves methodical searches of suspected larviposition sites—under fallen logs and rocks , in rot-holes in trees , in shaded places on the edges of dry river-beds , or in loose soil , or sand , covered with leaf litter [10 , 11] . In Zimbabwe , from August to November , tsetse puparia can frequently be found in burrows dug by the aardvark Orycteropus afer Pallas and often used thereafter by warthog Phacochoerus aethiopicus Pallas [11 , 12] . Muzari & Hargrove showed that female tsetse larviposit in artificial versions of these ‘warthog burrows’[13] . They described the construction of such devices and recorded annual variations in their use as larviposition sites by the tsetse flies Glossina morsitans morsitans Westwood and G . pallidipes Austen at Rekomitjie Research Station in the Zambezi Valley of Zimbabwe . The original study was concerned only with the collection of puparia , but a device was also described whereby flies could be captured in the burrows . The primary purpose of the present study was , for the first time in tsetse fieldwork , to sample full-term pregnant flies just before they deposited a larva . More generally we studied the sample characteristics of this new capture system . Our null hypotheses were that: the exact location of larviposition was independent of burrow site and orientation; that the larviposition site within each burrow was randomly chosen; that these choices were the same for both species and for females of different ages; that the timing of larviposition was random throughout the day and that a full-term pregnant fly entering a burrow deposited its larva in that burrow . Since warthog burrows are always cooler than ambient during daylight hours [13] , it seemed likely that tsetse would also use the burrows as “refuge” sites—the cooler dark places where flies are found in increasingly large numbers as temperatures increase above 32°C [12 , 14] . Accordingly we also studied the effects of calendar time and temperature on the sex and species distribution of captured flies and , particularly for female flies , the distribution between full-term pregnant , and other , females . The study suggested that female tsetse do not larviposit randomly with respect to time , nor with position within burrows , and that full-term flies often enter a burrow but then leave it without larvipositing . This behaviour differed between species . The research opens the door for an entirely new area of tsetse studies—including comparisons between the physiological status of individual females and the pupae they have just deposited .
During 1998 , 36 artificial warthog burrows were deployed over a distance of 800m along the banks of the Chiuyi and Rukomechi Rivers near Rekomitjie Research Station , Zambezi Valley , Zimbabwe . Burrows were deployed at nine sites in groups of four , with burrow openings at each site facing north , east , south and west , respectively . Burrow construction ( Fig . 1 ) is fully described by Muzari & Hargrove ( 2005 ) . For each of the 20 burrows at sites 1 to 5 a wire-framed , net-covered , trap was inserted into the mouth of a burrow , such that tsetse could enter the burrow but were trapped when they flew out towards the light ( Fig . 1 ) . Between 8 September and 25 November traps were cleared daily at c . 1115 , 1230 , 1400 and 1645 h . With this regularity of clearing , it was possible to collect “perinatal” female flies—i . e . , postpartum flies that had just deposited a larva , or full-term pregnancy flies which generally produced a larva soon after capture . Male flies , and females that were not perinatal , were also collected and counted . The 16 burrows at sites 6 to 9 were always used without a trap so that adult females could enter and leave the burrow unhindered . Flies entering these burrows could deposit larvae in one of six sand-filled plastic trays placed on the bottom of the burrow or could leave the burrow without depositing a pupa . The positions of any puparia deposited were noted as being found in the left or right column of trays and in the back , middle or front tray , as viewed from the burrow entrance . The idea of this experiment was to study larviposition site selection , in terms of burrow orientation and position within the burrow , with a view to maximising future puparial collection rates . The experiment also facilitated study of the variation of larviposition rates , for each of the two tsetse species , as a function of calendar date and temperature . Puparia were collected from the trays in these burrows , at 3–18 day intervals , between 28 August and 8 November . Adult tsetse from burrow traps were placed , with any puparium found in the trap that could be attributed unequivocally to the female fly in question , in individually labelled ( 75 x 25 ) -mm glass tubes which were kept under a black cloth in a polystyrene box . Larvae could be attributed unequivocally to their mothers if there was only a single larva/pupa , and only one postpartum female , in the trap cage—or if a full-term pregnant female was transferred from the capture cage to a glass tube and the female thereafter deposited a larva prior to dissection . Adult females were subjected to ovarian dissection , 98% on their day of capture and the remainder the following morning , and assigned to one of eight ovarian age categories , using the disposition and relative sizes of the oocytes within the ovarioles in the left and right ovaries [15] . This procedure can be used to determine unequivocally the number of times a fly has ovulated as long as this number is less than four . Thereafter , the ovarian category can only be defined modulo 4 [15]: thus , it is not possible to differentiate flies that have ovulated four times from those that have ovulated 8 , 12 , 16 etc . times , and similar problems exist for those that have ovulated 5 , 6 or 7 times . This did not constitute a problem for the present study since we were not concerned with actual ages of flies but only the ovarian age distributions of different groups . Pregnancy stage was assessed from the linear dimensions of the ovarian and uterine contents [16 , 17] . Females which were seen to have produced a larva after the time of capture—either in the trap or in the collection tube after removal from the trap , but before ovarian dissection , were classified as “postpartum” . This category was also used for flies which had an empty uterus , and where the sizes of the first and second largest oocytes indicated that the female had completed >95% of the ovulation cycle . It was assumed that these flies had very recently larviposited . Females with a third instar larva in utero , of a size indicating that >95% of the pregnancy was complete , were classified as “full-term” pregnant . Postpartum and full-term flies were jointly referred to as “perinatal” . All other female flies were termed “pre-full-term” ( abbreviated below to pre-FT ) . This group includes flies in ovarian category zero , i . e . , flies which had not yet ovulated for the first time , and all flies with either an egg or a first , second , or small third instar larva in utero . Finally it includes flies with an empty uterus where the size of the oocytes in the ovaries indicated that the female had completed ≤95% of the ovulation cycle . Note that the empty uterus could have occurred due to natural causes , or to capture-related trauma . For present purposes the cause is not relevant: if oocyte size indicates that ≤95% of pregnancy has been completed then the fly is properly classified as pre-FT—regardless of the content of the uterus or , where the uterus is empty , how this came about . A mercury thermometer in a Stevenson screen at Rekomitjie Research Station was used to record daily maximum and minimum temperatures . A rain gauge sited next to the screen produced daily records of precipitation . Hourly mean measurements of shade temperature and relative humidity were also made using an automatic weather station ( type WS01 , Delta-T devices , Newmarket , UK ) at a site c . 200 m from the Stevenson screen . Analyses were carried out using Stata ( StataCorp , 1999 ) statistical package , version 12 . We denote a chi-squared statistic with n degrees of freedom as χ2 ( n ) : we used Yates correction whenever n = 1 . We denote statistical significance at the 0 . 05 , 0 . 01 and 0 . 001 levels of probability by * , ** and *** , respectively: P>0 . 05 is denoted by “ns” . All error terms are 95% confidence intervals ( 95% ci ) .
September marks the onset of the hot-dry season in the Zambezi Valley and temperatures typically increase steadily through October and into November , remaining high until the onset of the rains , which do not generally start in earnest until mid-December . In 1998 maximum temperatures ( Tmax ) increased steadily into mid-October , fell for a few days , then increased again until early November ( Fig . 2A ) , peaking at 42 . 5°C—at that time the highest temperature ever recorded at Rekomitjie . Readings of Tmax from the mercury thermometer were on average about 1°C higher than those from the logger . The transient declines in October temperatures were associated with increased cloud cover and increases in the relative humidity ( Fig . 2B ) . Rain measured at 0 . 5 , 1 . 5 , and 17mm fell on 19 , 20 and 23 November: heavier and more sustained rain fell in December . Ambient temperatures at the weather station showed strong diurnal changes , with a minimum at about dawn and a maximum at 1500–1600 h . Despite the high ambient temperatures and the large diurnal variation , mean temperatures varied only between 28 and 29 . 5°C in the well-insulated burrows , which were up to 3°C cooler even than artificial refuges during daylight hours—but were warmer between 2200 h and shortly after dawn [13] . Of 3131 tsetse puparia collected from sand-filled trays in the 16 burrows at sites 6 to 9 , 1834 ( 59%; 95% ci 57%- 60% ) were G . m . morsitans and 1297 ( 41%; 95% ci 40%- 43% ) were G . pallidipes ( Table 1 , 2 ) . There was no significant difference between the proportions of puparia found in the trays on the left or right side of the burrow ( χ2 ( 1 ) = 0 . 2 , P > 0 . 05: Table 1 ) but the vast majority of both species deposited their larvae in the trays at the back of ( deepest into ) the burrow ( χ2 ( 2 ) >1300 , P< 0 . 001 for each species: Table 1 ) , and the smallest proportion in the trays closest to the mouth . There was also a significant difference between species in this regard , with G . pallidipes shifted significantly more towards the trays at the back of the burrow than G . m . morsitans ( χ2 ( 2 ) = 48 . 1 , P< 0 . 001 ) . Other analyses showed that there were significant differences between the proportions of puparia found in burrows facing north , south , east or west ( χ2 ( 3 ) = 18 . 0 and 9 . 8 for G . m . morsitans and G . pallidipes , respectively: P<0 . 001 in each case ) . There were , nonetheless , fairly small deviations from a uniform distribution of 25% in each burrow: for G . m . morsitans the range was 22 . 3%- 29 . 2% and for G . pallidipes 21 . 8%- 27 . 9% . The largest proportions were found in west and north-facing burrows for the two species , respectively . The data in Table 1 are for puparial numbers pooled on all sites , but the distribution of puparia between trays within a given burrow , and between burrows at a given site , were all similar . Accordingly , all further analyses of these data are carried out on data pooled on burrow for each site . In summary , burrow orientation was of minor importance for larviposition: but for both species , and particularly G . pallidipes , the vast majority of larvae were deposited at the very rear of the burrow , raising the possibility that deeper/longer burrows might have encouraged greater rates of larviposition . The total number of puparia collected varied between the four sites by a factor of 6 . 4:1 . There were always fewer G . pallidipes than G . m . morsitans but the proportion of G . pallidipes was much smaller at site 9 than at the other three sites ( Table 2 ) . For both species , the number of puparia collected per site per day also changed with calendar time , with a peak at the end of September ( Fig . 3 ) . Numbers then declined steadily until mid-November: thereafter almost no puparia are found in burrows [13] . Multiple linear regression analysis indicated , for both species , linear and quadratic effects of time after 28 August , and significant site effects ( Table 3 ) . What is not clear from the above results is the extent to which changes in larviposition rates reflect changes in tsetse population levels , as opposed to changes in the probability that females used burrows as larviposition sites . Moreover , it is not clear how day-to-day variation in larviposition rates in burrows is affected by meteorological factors . These issues were investigated in experiments where tsetse were captured in traps inserted into burrow mouths ( Fig . 1 ) . Some care is required in the analysis of these results since both males and females of both species are now captured . Since burrows provide dark spaces where daytime temperatures are always markedly below ambient [13] , male flies , as well as females that are not about to deposit a larva , are likely to use burrows not for larviposition but as artificial “refuges” from the heat [12] . For the sampling period 8 September—25 November the total catches of male and female tsetse from the burrows at sites 1 to 5 were 80 and 765 for G . m . morsitans , and 802 and 2003 for G . pallidipes . Of the females captured , 295/724 ( 41%; 95% ci 37%- 44% ) of G . m . morsitans were successfully dissected and identified as perinatal ( Table 4 ) , using the definitions described in the Methods . This was a significantly higher proportion than the 596/1825 ( 33%; 95% ci 31%- 35% ) of G . pallidipes identified as perinatal . When the data for each month were analysed separately , the proportions of G . m . morsitans and G . pallidipes among perinatal and pre-FT females did not differ significantly ( χ2 ( 1 ) <1 . 4 in each month , P> 0 . 05: Table 4 ) . When data were pooled over all months , however , there was a significantly greater proportion ( 33%: 95% ci 30%- 36% ) of G . m . morsitans among the perinatal females than among the pre-FT group ( 26%: 95% ci 24%- 28% ) ( Table 4 ) . In contrast to these figures , 1834/3131 ( 59%: 95% ci 57%- 60% ) of the puparia collected from burrow trays were G . m . morsitans ( Table 1 ) . For September and October combined , the numbers of perinatal females captured per site per day were 1 . 65 and 3 . 29 for G . m . morsitans and G . pallidipes , respectively , compared with 1 . 74 and 1 . 23 puparia of these species collected per site per day ( data from Tables 1 and 4 ) . In short , whereas many more full-term-pregnant G . pallidipes enter burrows , a greater proportion of G . m . morsitans actually larviposit in those burrows , suggesting that greater proportions of full-term G . pallidipes leave burrows they have entered without larvipositing . Catches of pre-FT and perinatal female flies showed entirely different trends with calendar time: in univariate analyses , numbers of pre-FT G . pallidipes increased markedly with time during the experiment ( Fig . 4A ) , whereas exactly the opposite trend was seen among perinatal females of this species ( Fig . 4B ) . Similarly , increasing Tmax was associated with rapidly increasing catches of pre-FT flies ( Fig . 4C ) , whereas there was no effect on the numbers of perinatal females captured ( Fig . 4D ) . Maximum temperature , relative humidity and calendar time are clearly correlated over the period of the experiment and , to tease out their effects on rates of larviposition , multivariate linear regression analysis was carried out on the data for both tsetse species . For pre-FT females of both species , once the very strong temperature effect was removed , a quadratic effect of time was also evident: i . e . , the numbers captured initially increased , and then decreased as the hot weather continued ( Table 5 ) . Since temperature is correlated both with relative humidity and saturation deficit , the latter two both had significant effects on catches when used in univariate analyses: the effects were , however , smaller than the temperature effects and , once temperature had been included in the model , there were no detectible additional effects of either relative humidity or saturation deficit on the catches . For perinatal females of both species , catches declined significantly with time and increased with temperature—though the latter effect was not significant for G . m . morsitans . The effects of time and temperature on catches of male G . pallidipes were qualitatively the same as for pre-FT females and the values of the coefficients for the time and temperature effects were not significantly different in the two groups ( Table 5 ) . For G . m . morsitans males only 80 flies were caught over the whole experiment and only a positive effect of temperature was found . When the relationships in Table 5 were used to predict the changes in catches of G . pallidipes , for given fixed Tmax , predicted catches of pre-FT females peaked at the end of the second week in October ( Fig . 5 ) . By contrast , the catches of perinatal females decreased from the time that sampling started in early September . There was also a positive effect of Tmax on the numbers of perinatal female G . pallidipes captured , though the coefficient was much smaller than for pre-FT females ( Table 5 ) . Accordingly , the proportion of perinatal females in the catch also declined continuously after the beginning of September ( Fig . 5 ) . The model for male G . pallidipes was closely similar to that for pre-FT females and the predicted catches at a fixed temperature also peaked at the end of the second week of October . Since the numbers of perinatal female flies decreased monotonically with time for given Tmax , whereas catches of both males and females reached a peak in mid-October and then declined , the predicted proportion of female flies in the entire catch also declined with time ( Fig . 5 ) . If , however , the perinatal females were excluded from the above calculation—so that we were only considering flies assumed to be using burrows as refuges , rather than as larviposition sites , then the predicted percentage of females in the catch was almost constant at about 59% ( Fig . 5 ) . The plots in Fig . 5 are for an arbitrary choice of constant Tmax , but choices in the range 35–40°C gave qualitatively similar results: increasing temperature resulted in modest increases in the numbers of perinatal flies , larger increases in the pre-FT group , and thus lower overall percentages of perinatal flies . The percentage of females among flies assumed to be using the burrows as refuges was approximately constant at each temperature , the level shifting from about 55% to 62% as the temperature increased . The models in Table 5 predict that , at constant temperature , 68–80% of catches of pre-FT females will be G . pallidipes . The distribution of catches of female flies across the day differed significantly between the two groups of female tsetse ( Fig . 6 ) . For the pre-FT group the situation changed with the month of the experiment: in September only about 10% of the G . m . morsitans ( Fig . 6A ) and 20% of the G . pallidipes ( Fig . 6D ) were taken in the 1230h sample . This proportion increased markedly in October , and by November the greatest proportion was caught at this time ( Fig . 6C , F ) . These results are consistent with earlier findings that tsetse enter refuges at increasingly early hours of the day as temperatures increase [12 , 14] . The perinatal group of flies entered early in the day in all three months: for both species the greatest proportion ( 40–65% ) of the flies were caught in the first sample taken at 1230h , and 80–90% of the days’ samples were caught by 1400h , regardless of the temperature ( Fig . 6 ) . An obvious difference between the age structure of the pre-FT and perinatal groups of female is that none of the latter can be flies in ovarian category zero , which have not yet ovulated for the first time . When this age category was excluded from the analysis there was little difference , for either species , between the age structures of the two groups ( Fig . 7 ) .
While artificial burrows , designed as larviposition sites , may double as artificial refuges , the reverse does not seem to be true . When artificial refuges were provided with a floor of river sand , fewer than one puparium per week were found in these sites [12] . This emphasises the rather precise conditions that female tsetse appear to require of larviposition sites . When tsetse had the choice of larvipositing in trays filled with either plain sand , or sand covered with leaf litter , 97% of all tsetse puparia were found in the latter site [13]: in that study , as in the present one , it was also found that the vast majority of larvae were deposited in the trays furthest to the back of the burrows . The absence of added leaf litter in Vale’s ( 1971 ) refuge flooring , the slightly warmer conditions in the artificial refuges , and the more exposed situation compared with our burrows , will all presumably have reduced the probability that his refuges would be used as larviposition sites [12] . The greater proportion of G . m . morsitans among pupae collected , as against the proportion among perinatal flies captured ( cf Tables 2 and 4 ) might simply reflect differential proportions of adults of the two species around sites 6 to 9 , where puparia were collected , and around sites 1 to 5 , where adult flies were trapped . Given , however , that the sites were all within a few hundred metres of each other , and were all in similar habitat , this explanation seems unlikely . Alternatively the anomaly could reflect a higher level of discrimination among G . pallidipes in the selection of larviposition sites . In this scenario , among full-term pregnant females entering a burrow , a greater proportion of the G . pallidipes leave the burrow without larvipositing , and a greater proportion of G . m . morsitans deposit a larva . When there is a trap in place , however , full-term pregnant flies that have entered the burrow are prevented from leaving . Those full-term pregnant female G . pallidipes that would have left the burrow are now trapped and this changes the balance of the number of perinatal females trapped in favour of this species relative to G . m . morsitans . Several pieces of evidence are consistent with the above scenario . Firstly , the selection of the trays deepest in the burrows was stronger in G . pallidipes than in G . m . morsitans ( 83% vs 74% , Table 1 ) . Perhaps even deeper/longer burrows will elicit improved larviposition by both species , but particularly by G . pallidipes . Secondly , for G . pallidipes , the yield per site per day of perinatal females ( 3 . 29 , Table 5 ) was 2 . 7 times higher than the yield of puparia ( 1 . 23 , Table 1 ) from burrows where there was no trap and where flies were free to come and go . For G . m . morsitans the yields were very similar , 1 . 74 and 1 . 65 , respectively . This is consistent with a larger proportion of full-term pregnancy G . pallidipes entering burrows but then leaving , without larvipositing , if they were allowed to do so . Indeed the above figures suggest that for every ten full-term pregnant female G . pallidipes entering a burrow , on average perhaps only four would deposit their larvae there . Thirdly , the fact that markedly more G . m . morsitans than G . pallidipes larvae were deposited in the burrows , in an area where the latter species is clearly more numerous [18] , suggests that the artificial burrows were less attractive as larviposition sites to G . pallidipes than to G . m . morsitans . Fourthly , as reported previously [13] , the proportion of G . pallidipes from natural warthog burrows was double the proportion in artificial burrows . More carefully controlled experiments will be required to test the tentative idea that many heavily pregnant females entered the burrows but left without larvipositing and that this effect is more accentuated in G . pallidipes . This suggests an important determinant of larviposition site selection , but the stimuli and responses underlying this selection are unclear . Similarly , it is not clear what underlies the difference in larviposition responses of G . m . morsitans and G . pallidipes . An impressive array of differences in behavioural responses between tsetse species has been attributed to differences in habitat geometry [19] . Within habitats , differences in behaviour between species , and between males and females of the same species , were attributed to differences in fly size and mobility . It is possible that the differences in larviposition behaviour observed here between the two species has similar causes . G . pallidipes is a larger and more mobile fly than G . m . morsitans and may , therefore , be able to visit a greater number of potential sites before larvipositing . The study system described here provides ways of addressing this question and other issues . Temperature-adjusted catches of males and of pre-FT females continue to increase until mid-October , consistent with the idea that the decline prior to that time , in pupal deposition rates in burrows , reflects a falling probability of tsetse larvipositing in burrows , rather than a decline in the population of full-term-pregnant flies in the area . What is less clear is whether females are reacting to any meteorological or other cues in moving away from using the burrows as sites for larviposition . Regression analysis suggests that there is no effect of either ambient relative humidity or saturation deficit: moreover , the small effect of ambient temperature has a positive coefficient which would tend to increase , rather than decrease , the numbers of larvae deposited in the burrows as the hot season progresses . We cannot exclude the possibility that flies were reacting to changes in humidity within burrows , which was not measured here , but should be monitored in future studies . Reduced use of warthog burrows as larviposition sites during October and November , reflected both in the reduction in the absolute numbers of perinatal flies captured in burrows , and of puparia found in these sites , may be adaptive . When heavy rains fall , the burrows become waterlogged and may even flood , and any pupae present would presumably perish . In 1998 the first rains only fell on 19 November , whereas the decline in captures of perinatal flies started a month before this time . Nonetheless , it is not uncommon for heavy rain to fall in the first half of November; in 5/10 years in the period 1989–1998 > 12 . 5 mm of rain were recorded at Rekomitjie by 15 November . Larvae deposited in the last half of October might not , therefore , have emerged by the time the first heavy rains fell . Over the same period 0/10 years had recorded this level of rain by the end of October . There is therefore little danger of flooding for any flies deposited in burrows during September , but this danger increases in October and particularly November . It has been argued that refuges provide the least biased sample of tsetse currently available—at least in terms of age structure and pregnancy stage [20 , 21] . The similarity between the age structures of perinatal females and those apparently using the burrows as a refuge is then consistent with the idea that all ages of fly use the burrows as larviposition sites with approximately the same probability . For both species , perinatal flies were most often caught before the hottest time of the day ( at about 1500h ) : 45–65% were captured by 1230h on any given day and most often larviposited in the collection tube shortly thereafter . We do not know , however , when the flies would have larviposited if they had not been caught , nor even whether they would have larviposited in the burrow where they were caught had they not been trapped . Further field studies are required to better estimate the timing of larviposition in the field . Interpretation of catches from burrow traps is also complicated by the unknown efficiency of the traps . On one hand the presence of the trap may have discouraged some flies from entering the burrow . Conversely , traps occasionally contained a pupa that could not have been produced by any of the flies in the trap—implying that a postpartum fly had escaped . A better idea of the number of flies entering the burrows could be obtained by deploying electric nets [22] around the entrance of the burrow and/or even inside it . The estimate that 60% of full-term G . pallidipes leave a burrow that they have entered , without larvipositing there , suggests that we should seek ways of improving the burrows such that females are more likely to larviposit in the burrow they first enter . This study , and an earlier one [13] , made no progress in improving on the original burrow design . The use of the burrows as larviposition sites also seems to be limited to about four months of the year and our knowledge of the behaviour of perinatal flies in the field is thus currently limited to just two species of tsetse during a single season . At other times of the year pupae are deposited along the edges of dried up water-courses , in rot holes in trees and under fallen logs . Artificial versions of the latter two sites might be used to sample larvipositing flies at other times of the year at Rekomitjie , and indeed to sample other species in other parts of Africa . Currently , the burrows capture system provides the only method for collecting perinatal tsetse in the field , providing thereby a unique opportunity to study the physiology of female tsetse and the larvae that they have just deposited . Future papers will address this interesting new area of tsetse field biology .
|
Adult tsetse , vectors of trypanosomiasis , have been extensively studied for more than 100 years , but little is known about larviposition behaviour in the field . Pupae are generally collected in the field via arduous searches of putative larviposition sites . Females have never been sampled in the field as they deposit a larva , leading to confusion about the physiological dynamics at the end of pregnancy . We overcome these problems through the use of artificial warthog burrows , where tsetse deposit pupae during the hot dry season in the Zambezi Valley of Zimbabwe . When burrows were fitted with a retaining trap it was also possible to sample perinatal ( full-term pregnant and postpartum ) female tsetse . Comparisons of the numbers of pupae deposited in burrows without the trap , with the numbers of perinatal flies trapped in burrows , showed that many full-term pregnant female tsetse enter burrows but then leave without depositing a larva . G . pallidipes are more discriminating in this regard than G . m . morsitans . Capture of perinatal females will make it possible , for the first time , to compare the physiological status of female tsetse and the pupa they have just deposited , with important implications for the understanding of tsetse population dynamics .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Artificial Warthog Burrows Used to Sample Adult and Immature Tsetse (Glossina spp) in the Zambezi Valley of Zimbabwe
|
Infection with filarial parasites is associated with T cell hyporesponsiveness , which is thought to be partly mediated by their ability to induce regulatory T cells ( Tregs ) during human infections . This study investigates the functional capacity of Tregs from different groups of filarial patients to suppress filaria-specific immune responses during human filariasis . Microfilaremic ( MF ) , chronic pathology ( CP ) and uninfected endemic normal ( EN ) individuals were selected in an area endemic for Brugia timori in Flores island , Indonesia . PBMC were isolated , CD4CD25hi cells were magnetically depleted and in vitro cytokine production and proliferation in response to B . malayi adult worm antigen ( BmA ) were determined in total and Treg-depleted PBMC . In MF subjects BmA-specific T and B lymphocyte proliferation as well as IFN-gamma , IL-13 and IL-17 responses were lower compared to EN and CP groups . Depletion of Tregs restored T cell as well as B cell proliferation in MF-positives , while proliferative responses in the other groups were not enhanced . BmA-induced IL-13 production was increased after Treg removal in MF-positives only . Thus , filaria-associated Tregs were demonstrated to be functional in suppressing proliferation and possibly Th2 cytokine responses to BmA . These suppressive effects were only observed in the MF group and not in EN or CP . These findings may be important when considering strategies for filarial treatment and the targeted prevention of filaria-induced lymphedema .
Lymphatic filariasis ( LF ) , caused by nematodes Wuchereria bancrofti , Brugia malayi and B . timori , affects around 120 million people worldwide and additionally 2 billion people are at risk in endemic areas [1] . Although not life-threatening , chronic manifestation of disease causes major disabilities and deformities , especially in areas with minimal access to health care facilities . Indonesia is one of the endemic countries in the South-East Asia region and accounts for the second highest burden of LF in the world . All three filarial parasites are prevalent in the archipelago and efforts are being made to control the disease in various areas ( Global Programme to Eliminate Lymphatic Filariasis ) [2] , [3] . Helminths such as filarial parasites have been shown to induce immune modulation , resulting in T cell hyporesponsiveness and failure to expel parasites [4] . Initially a phase of immune activation and proinflammatory cytokine responses is induced by the larval stages of filarial parasites [5] . However in patent infection , with circulating microfilariae ( MF ) and/or filarial antigens , decreased proliferative responses and increased anti-inflammatory cytokines , such as IL-10 and TGF-β , reflect a state of immune hyporesponsiveness [6] . At the transcriptional level , it has been shown that in infected subjects both Th1 and Th2 pathways are downmodulated by the enhanced expression of molecules such as FOXP3 , CTLA-4 and TGF-β involved in regulatory networks [7] . In patients with chronic pathology this seems to be reversed; in PBMC from these patients enhanced inflammatory Th1 and Th17 responses as well as decreased levels of mRNA for different Treg markers were observed as compared to asymptomatic infected individuals [8] . The suppressive capacities of Tregs have been implicated in many infectious diseases , including filariasis . Induction of Tregs by pathogens is regarded as one of the mechanisms to evade the human immune system [9] . A recent report demonstrated that in animal models , early recruitment of Tregs affects the course of the immune response that leads to the development of chronic filariasis , indicating that Tregs are important regulators of the overall immune response to filarial nematodes in mice [10] . In human filariasis , different Treg subsets have been the focus of recent studies in different age groups and different clinical categories . While in India , higher frequencies of regulatory T cell markers were found in asymptomatic microfilaremics compared with chronic pathology patients , a recent study in Mali reported higher frequencies of Tregs ( CD4+CD25+FOXP3+CD127− ) in MF or circulating filaria antigen-positive versus uninfected adolescents , but also suggested a more prominent regulatory role for IL-10 producing , so-called adaptive Tregs ( CD4+CD25− ) cells [8] , [11] . Altogether , in previous studies of regulatory networks in human filariasis phenotypes of participating lymphocyte subsets and key regulatory molecules have been investigated , whereas the functional capacity of Tregs remained largely unknown . In this study we aimed to explore the immune regulatory activity in different disease states of microfilaremia ( MF ) , chronic pathology ( CP ) presented as elephantiasis and uninfected endemic normals ( EN ) as controls in a population living in an area endemic for B . timori in Indonesia . By in vitro depletion assays we determined the effect of Tregs on filaria-specific T and B cell proliferation and cytokine production .
In Sikka district , Flores , east Indonesia , an area endemic for B . timori was identified . Study participants were recruited from surrounding villages , written informed consent was obtained and night blood samples were collected to determine microfilaremia . Morning venous blood samples were collected from 24 MF-negative asymptomatic endemic normals ( EN ) , 24 MF-positive asymptomatic individuals ( MF ) and 26 MF-negative chronic pathology ( uni- or bilateral elephantiasis ) patients ( CP ) . 1 ml of blood was used for filtration to quantify mf load and thick blood smears were screened for the presence of malaria parasites . The study was approved by the Committee of Medical Research Ethics of the University of Indonesia . Peripheral blood mononuclear cells ( PBMC ) were obtained by gradient centrifugation of heparinized venous blood over Ficoll . Based on sufficient numbers of PBMC , of 69 individuals ( 23 in each group ) CD4+CD25hi T cells were isolated by magnetic cell sorting ( MACS ) using the CD4+CD25+ Regulatory T Cell Isolation Kit ( Miltenyi Biotec GmBH , Bergisch Gladbach , Germany ) ; details have been described previously [12] . The CD4+CD25hi -depleted PBMC were compared with PBMC which were treated in an identical manner , however to which the eluted CD4+CD25hi cells were added back to ( this is referred to as “mock-depleted” ) . The green-fluorescent dye carboxyfluorescein succinimidyl ester ( CFSE; Sigma-Aldrich , CA , USA ) was used to monitor proliferation . CFSE is divided over daughter cells upon cell division and this can subsequently be tracked by decreasing fluorescence intensity . After labeling with 2 µM CFSE , mock- and CD4+CD25hi -depleted PBMC were cultured in RPMI 1640 ( Gibco , Invitrogen , Carlsbad , CA , U SA ) supplemented with 10% FCS ( Greiner Bio-One GmbH , Frickenhausen , Germany ) with or without B . malayi adult worm antigen ( BmA , 10 µg/ml ) . After 96 h cell supernatants were collected and cells were fixed in 2% formaldehyde ( Sigma-Aldrich ) , after which all samples were preserved at −20°C first , then at −80°C . After thawing , the CFSE-positive cells were labeled with fluorochrome-conjugated anti-CD3 , anti-CD4 , anti-CD25 ( BD Biosciences , Franklin Lakes , NJ , USA ) and anti-CD19 ( biotinylated antibody from eBioscience Inc . , San Diego , CA , USA; streptavidin-Qdot525 from Invitrogen ) antibodies , acquired on a FACSCanto II machine ( BD Biosciences ) and analyzed with FlowJo software ( Treestar Inc . , Ashland , OR , USA ) . Proliferation of effector T cells was determined in a FlowJo Proliferation application by calculation of the fraction of cells from the starting population that had divided , within the CD3+CD4+CD25+ T cell and CD3−CD19+ B cell subsets . Since background levels of cell proliferation were high , spontaneous proliferation was subtracted from BmA-stimulated values to compare proliferative responses in the three study groups . Cytokine production was assessed using the Multiplex Bead Immunoassay for interferon-gamma ( IFN-γ ) , interleukin ( IL ) -13 , IL-17 and IL-10 according to the protocol supplied by the manufacturer ( Biosource , Invitrogen , Carlsbad , CA , USA ) . Samples were acquired with Luminex 100™ xMAP technology ( Luminex Corp . , Austin , TX , USA ) . Half the detection limit supplied by the manufacturer was used for values below detection limit and the values above upper detection limit were given the upper limit value . The cytokine data were not normally distributed and therefore are presented as raw unmanipulated data . Thus , there was no subtraction of or division over unstimulated samples , but data are shown separately as medium-stimulated or antigen-stimulated cytokines . Statistical analysis was performed in SPSS 16 . 0 . Not-normally distributed values ( cytokine levels in supernatants ) were log-transformed . Both age and sex were incorporated into univariate analysis to compare different infection and clinical groups . Resulting adjusted means were anti-log-transformed when needed . For mock- versus Treg-depleted samples , paired analysis was done using paired t-test or Wilcoxon Signed Ranks Test . In the multiplex cytokine analysis Bonferroni correction was taken into account where applicable , by multiplying the p-values by the number of non-correlated measurements .
Individuals from an area endemic for lymphatic filariasis in the north of Flores , Indonesia , were recruited for a night blood survey . Based on the microfilaremic status and sufficient number of PBMC , 23 MF-negative asymptomatic endemic normals ( EN ) , 23 MF-positives ( MF ) and 23 chronic pathology ( CP ) patients were included for immunological studies . Microscopic Plasmodium spp . parasitemia was found in 2 CP patients , but had no effect on the analyses shown here . The characteristics of the study population are summarized in Table 1 . Age was significantly higher in the CP group ( medians 42 , 46 and 54 years for EN , MF and CP respectively; p = 0 . 038 ) , while male to female ratio was lower in the CP group ( percentage male 48% , 65% and 22% in EN , MF and CP respectively; p = 0 . 012 ) . Because of these differences , comparisons between groups were adjusted for age and sex . The lymphocyte count ( PBMC/ml blood ) was similar in the three groups ( medians 1 . 09 , 0 . 97 and 0 . 92 for EN , MF and CP respectively ) , as well as the frequencies of T and B cells ( data not shown ) . To analyze suppression of lymphocyte proliferation during filarial infection , cell proliferation to filarial antigen was determined by CFSE dilution in PBMC . Divided cell subsets were measured in activated T ( CD4+CD25+ ) and in B ( CD19+ ) cell populations . Net T cell proliferation was lower in the MF group , which was mainly caused by high background proliferation in unstimulated condition ( response to medium , Figure 1A; age- and sex-adjusted means 2 . 66% , −0 . 236% , 4 . 02% divided in EN , MF and CP respectively; p = 0 . 043 for EN vs . MF , p = 0 . 010 for MF vs . CP ) . Also B cell proliferation was lower in MF , shown in figure 1B ( adjusted means 1 . 13% , −1 . 11% , 2 . 07% divided for EN , MF and CP respectively; p = 0 . 002 for EN vs . MF and p = 0 . 0002 for MF vs . CP ) . To assess the modulation of differentiated T helper cell subsets by filarial infection , hallmark cytokines for Th1 ( IFN-γ ) , Th2 ( IL-13 ) , Th17 ( IL-17 ) and regulatory ( IL-10 ) responses were assessed in culture supernatants from cells stimulated with BmA ( Figure 2 ) . IFN-γ production was lower in the MF group than in EN or CP ( adjusted means 573 , 146 and 1318 pg/ml in EN , MF and CP respectively; p = 0 . 004 for EN vs . MF , p = 0 . 00007 for MF vs . CP ) . Both IL-17 and IL-13 levels were decreased in the MF group compared to EN , however were not different from CP ( IL-17 adjusted means 130 , 48 and 100 pg/ml; p = 0 . 037 for EN vs . MF; IL-13 adjusted means 1472 , 895 and 1276 pg/ml in EN , MF and CP respectively; p = 0 . 029 for EN vs . MF ) . IL-10 production was similar in all three groups ( adjusted means 333 , 427 and 355 pg/ml in EN , MF and CP respectively ) . Spontaneous production of these cytokines was not significantly different between the groups and it was noted that IFN-γ and IL-17 levels in BmA-stimulated PBMC supernatants were hardly above spontaneous ( unstimulated ) production , particularly in the MF group ( Figure S1 ) . After applying correction for multiple analyses , only IFN-γ levels were significantly lower in microfilaremic individuals . To assess the functional contribution of Tregs to in vitro immune responses , we performed magnetic depletion of CD4+CD25hi cells . By flowcytometry mock- and Treg-depleted PBMC were assessed for expression of CD25 and FOXP3 on CD4 T cells , of which a representative example is shown in Figure 3A . Treg frequencies decreased in most cases ( Figure 3B ) , which was highly significant and similar in all three clinical groups ( Figure 3C; p = 1 . 7•10−4 for EN , p = 2 . 7•10−5 for MF , p = 3 . 1•10−5 for CP ) . Geometric mean of CD25hiFOXP3+ cell percentages of CD4 cells decreased from 1 . 69% to 0 . 83% after depletion ( mean extent of depletion was 46 . 5% ) . For 5 donors , 4 in EN and 1 in CP group , Treg frequency either could not be assessed or did not decrease after depletion , therefore these patients were excluded for further analysis . To evaluate the influence of Tregs on BmA-specific T and B cell proliferation , we analyzed CFSE dilution in CD4+CD25+ and CD19+ cell subsets before and after Treg depletion . Here we present unadjusted proliferative responses to BmA and medium separately . For CD4+CD25+ effector T cells we observed an increase in proliferation to BmA in the MF group after removal of Tregs , whereas proliferative responses did not change significantly in EN or CP groups ( Figure 4A; p = 0 . 004 for MF ) . Treg depletion did not enhance spontaneous proliferation ( medium condition ) in MF-positives , however spontaneous responses were increased in uninfected individuals ( Figure 4B , p = 0 . 04 for EN ) . Interestingly , B cell proliferative responses in response to BmA were also enhanced in Treg-depleted conditions for MF patients , although this fell short of statistical significance ( Figure 4C; p = 0 . 07 for MF ) . In contrast , after Treg depletion B cells proliferated to a lesser extent in CP patients ( Figure 4C; p = 0 . 01 for CP ) . Unstimulated B cell proliferative responses were not influenced by Treg removal ( Figure 4D ) . To check whether Treg depletion completely restored lymphocyte proliferative responses in the MF group to levels seen in the other groups , we compared age- and sex-adjusted net proliferative responses of T and B cells to BmA in Treg-depleted conditions . Although responses in the CP group remained high for both T and B cells , T and B cell proliferation in the MF group was no longer different from EN individuals ( Figure S2 ) . Next , we investigated the capacity of Tregs to suppress the filaria-specific cytokines by measuring IFN-γ , IL-13 , IL-17 and IL-10 in response to BmA in culture supernatants of mock- and CD4+CD25hi - depleted PBMC . In Figure 5 , unmanipulated cytokine responses to BmA and medium are shown separately . Filaria-specific IFN-γ production was significantly upregulated after removal of Tregs in the MF group only ( Figure 5A; p = 0 . 064 , p = 0 . 004 for EN and MF respectively ) . However , the IFN-γ response to BmA was weak and similar in magnitude to responses seen in medium-stimulated PBMC , which also increased after depletion of Treg in MF as well as CP ( Figure 5B; p = 0 . 084 , p = 0 . 0002 , p = 0 . 001 for EN , MF and CP ) . With respect to IL-13 , the response to BmA increased after depletion of Tregs in MF-positive individuals only ( Figure 5C; p = 0 . 41 , p = 0 . 008 , p = 0 . 20 for EN , MF and CP respectively ) . Spontaneous IL-13 production was low compared to BmA-stimulated conditions and also increased significantly upon removal of Treg , but this was still negligible compared to levels induced by BmA ( Figure 5D; p = 0 . 007 for EN , p = 0 . 038 for MF and p = 0 . 002 for CP ) . IL-17 and IL-10 responses before and after Treg depletion were comparable and unchanged in all three groups ( data not shown ) .
To investigate the function of Tregs in different infection and clinical groups of human filariasis , we studied the effect of Treg depletion on in vitro responses to BmA using human PBMC from individuals in an area endemic for B . timori lymphatic filariasis in Flores , Indonesia . Our main findings were diminished T and B cell proliferation as well as lower IFN-γ , IL-17 and IL-13 production in MF-positives , but similar IL-10 secretion compared to CP and EN groups . Treg depletion resulted in antigen-specific increase of lymphocyte proliferation and IL-13 responses in the MF group only . Since our study population was not optimally age- and sex- matched , it was necessary to adjust for age and sex in the comparisons made between the infection groups . In studies on human filariasis it is often difficult to obtain comparable patient groups . One reason for this is the pathophysiology of this disease; microfilaremia can be present in all ages but particularly in young adults , while CP is an end stage disease that develops in older age . Importantly , a recent paper demonstrated a relevant effect of age on infection-induced regulatory immune responses; intensity of infection with Schistosoma haematobium was positively correlated with Treg frequency in the age group 8–13 years , while the opposite was observed for the group older than 14 years [13] . Age and sex should thus be taken into account carefully when interpreting cellular immunological data . Lymphocyte proliferation in filariasis has been studied since the 1970s and is consistently shown to be diminished in microfilaremic patients , including previous population studies by our group in Sulawesi , Indonesia [14]–[19] . We have now established that the well-described T cell hyporesponsiveness can be measured by CFSE dilution assays in PBMC stimulated with BmA , and also show that in addition to T cells , B cell proliferation is considerably lower in MF-positives . Previously , it has been shown that the functional capacity of B cells , in terms of specific IgE and IgG production , was lower in MF versus CP patients [20] , [21] . Here , we extend this to B cell proliferation , showing for the first time to our knowledge B cell proliferative hyporesponsiveness in microfilaremics . Interestingly , despite lower IgG found in earlier studies , the number of positive individuals for filaria-specific IgG4 , an isotype shown to be associated with elevated plasma IL-10 [22] , was higher in microfilaremics ( data not shown ) . It is tempting to speculate that B cells in MF subjects that are hyporesponsive are also contributing to immune regulation by producing IL-10 and IgG4 , as is suggested for venom-specific B cells from beekeepers ( reviewed in [23] ) . The suppressed cytokine responses in MF-positive individuals here correspond with a recent study , showing higher IFN-γ and IL-17 responses to BmA in chronic pathology patients compared to MF-positive individuals [8] . In microfilaremics IFN-γ and IL-17 production were not induced above background levels; this is also in line with the findings by Babu et al . , who analyzed the production of IFN-γ and the expression of IL-17 mRNA in 24 h BmA-stimulated PBMC [8] . However , Treg removal did not affect the Th1 and Th17 cytokines which may suggest that these cytokines are not regulated by Tregs . IL-13 production in response to BmA was increased after Treg depletion , however this result must be considered with caution , since medium responses were also changed . Since IL-10 levels were high in MF before as well as after Treg depletion , IL-10 derived from CD4+CD25− T cells could be responsible for the observed decreased cytokine responses in microfilaremics , supported by two studies which showed the majority of IL-10 during filarial infection was produced by effector T cells , despite higher Tregs in the MF group [11] , [24] . Contrary to our expectations , Treg depletion had little or no effect on BmA responses in the other groups , although these individuals live in a filaria-endemic area and do have filaria-specific proliferative and cytokine responses . One explanation might be that active Tregs in MF are filaria- or BmA-specific , which are only actively induced and/or expanded during patent microfilaremia . Since there are very few studies on the function of Tregs in human helminth infections , it would be interesting for future studies to determine antigen specificity and functional characteristics of the Tregs in the different study groups . Furthermore , due to limited number of available cells we were unable to determine the mechanisms by which this CD4+CD25hi subset affects immune responses; an area that should be investigated in the future . A previous study concluded that in vitro blockade of CTLA-4 and PD-1 reverted suppression of M . tuberculosis-specific immune responses , suggesting cell-contact mediated mechanisms of suppression during microfilaremia [25] . Regarding the limitations of the current study , we were not able to evaluate whether the Treg depletion procedure has led to depletion of any other cell subsets , as a possible explanation for the reduced B cell proliferation in CP . In addition , our plan to confirm previous studies that show higher FOXP3 in ex vivo PBMC of MF patients failed due to a technical problem with FACS staining of FOXP3 . We only had 4-days cultured PBMC that we could stain for FOXP3 and thereby we were able to show the depletion of CD25hiFOXP3+ cells . However the level of CD25 and FOXP3 in medium-cultured cells may not be fully representative for the circulating levels of Tregs . Nevertheless , although important to gather data on Treg frequencies , the primary objective of our study was to assess their functional capacity in different infection and clinical groups . It should also be mentioned that our previous study of geohelminth infection in Indonesia indicated that it was not the number but the suppressive capacity of Tregs which was altered in infected children [12] . In conclusion , we report active contribution of Tregs to modulation of T and B cell proliferation and polarized cytokine production by effector T cells in MF-positive individuals in Flores , Indonesia . Since chronic lymphedema appears to be concurrent with lack of Treg-associated suppressive capacity , further research on targeted activation of specific Tregs would be important to be able to decrease the morbidity and disabilities induced by LF .
|
Lymphatic filariasis is a neglected disease still prominent in low-resource settings and is very disabling when it progresses to chronic pathology caused by lymphedema . Until now , studies on the contribution of Tregs to lymphocyte hyporesponsiveness in human filariasis have focused on frequency and phenotypic characteristics of these cells . We have looked at the functional consequence of the presence of Tregs in filaria-specific immune responses during different stages of human lymphatic filariasis . Proliferation of not only T cells , but also B cells , was decreased in patients with microfilaremia compared to uninfected individuals and chronic pathology ( lymphedema ) patients . The suppressed lymphocyte proliferative responses were increased after in vitro removal of Tregs in the microfilaria-positive group only , indicating the presence of filaria-specific functional Tregs in microfilaremic patients which are not as active in subjects with chronic pathology or without infection . Th2 cytokine responses were specifically enhanced in microfilaremics as well after Treg depletion , suggesting Treg-associated suppression of filaria-specific Th2 responses . Taken together , filaria-specific Treg contribute to immune modulation during microfilaremia and might need to be considered in therapeutic strategies to prevent chronic pathology induced by filarial infection .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"immune",
"cells",
"cytokines",
"immunity",
"to",
"infections",
"immune",
"suppression",
"immunology",
"adaptive",
"immunity",
"neglected",
"tropical",
"diseases",
"immunoregulation",
"immunomodulation",
"lymphatic",
"filariasis",
"infectious",
"diseases",
"inflammation",
"t",
"cells",
"immune",
"response",
"immune",
"system",
"clinical",
"immunology",
"immunity"
] |
2012
|
Regulatory T Cells in Human Lymphatic Filariasis: Stronger Functional Activity in Microfilaremics
|
ChIP-seq enables genome-scale identification of regulatory regions that govern gene expression . However , the biological insights generated from ChIP-seq analysis have been limited to predictions of binding sites and cooperative interactions . Furthermore , ChIP-seq data often poorly correlate with in vitro measurements or predicted motifs , highlighting that binding affinity alone is insufficient to explain transcription factor ( TF ) -binding in vivo . One possibility is that binding sites are not equally accessible across the genome . A more comprehensive biophysical representation of TF-binding is required to improve our ability to understand , predict , and alter gene expression . Here , we show that genome accessibility is a key parameter that impacts TF-binding in bacteria . We developed a thermodynamic model that parameterizes ChIP-seq coverage in terms of genome accessibility and binding affinity . The role of genome accessibility is validated using a large-scale ChIP-seq dataset of the M . tuberculosis regulatory network . We find that accounting for genome accessibility led to a model that explains 63% of the ChIP-seq profile variance , while a model based in motif score alone explains only 35% of the variance . Moreover , our framework enables de novo ChIP-seq peak prediction and is useful for inferring TF-binding peaks in new experimental conditions by reducing the need for additional experiments . We observe that the genome is more accessible in intergenic regions , and that increased accessibility is positively correlated with gene expression and anti-correlated with distance to the origin of replication . Our biophysically motivated model provides a more comprehensive description of TF-binding in vivo from first principles towards a better representation of gene regulation in silico , with promising applications in systems biology .
In order to adapt to different environmental challenges , microorganisms need to precisely control the expression of specific sets of genes at defined magnitudes at any given moment [1 , 2] . This control is mediated by regulatory proteins such as transcription factors ( TF ) that are able to recognize and bind specific DNA sequences to recruit or block the gene expression machinery . Recent advances in next-generation sequencing have now enabled us to measure TF-binding in vivo at the genome scale [3–5] . Chromatin Immunoprecipitation followed by sequencing ( ChIP-seq ) is a popular technology for in vivo measurements of TF binding [6–8] , which uses TF-specific antibody selection and high-throughput sequencing to identify the genomic regions that are bound by a query TF . In parallel , technologies for high-throughput characterization of TF-binding in vitro have also emerged [9–13] . Yet , only a fraction of the expected binding sites are bound under physiological conditions [8] and in vivo measurements are poorly correlated with in vitro ones [14 , 15] . TF-binding in vivo is often more complex than what can be measured in vitro due to multiple factors [16] . For instance , strength of TF-binding affinity [17 , 18] , presence of multiple binding sites [19] , cooperative interactions [18 , 20] , and genome accessibility [21 , 22] have all been shown to impact TF-binding in vivo . Incorporating these parameters in ChIP-seq analysis can lead to more accurate models of gene regulation across the whole genome [14 , 15] . As sequencing costs continue to decrease , challenges in ChIP-seq studies are transitioning from data generation to analysis and modeling [23] . Data analysis methods have moved from purely peak identification to physically-motivated models of ChIP-seq coverage [24] . Early computational methods focused on identifying statistically enriched peaks that correspond to TF-binding regions [5 , 25–28] . Recent methods are incorporating mechanistic principles to extract regulatory insights [24 , 29–31] . For example , the BRACIL method integrates ChIP-seq coverage , motif score , and thermodynamic modeling through a signal processing representation to predict binding site locations with high-resolution as well as cooperative interactions [24] . The growing abundance of ChIP-seq data creates a greater demand for more comprehensive models [15 , 23 , 32] and an opportunity to evaluate key parameters of TF-binding in vivo . Within the cell , transcription factors need to have physical access to the relevant regulatory regions in order to control gene expression . In eukaryotes , genome accessibility is mostly caused by different chromatin states due to epigenetic factors such as histone modification and nucleosome structures [33] . The chromatin state can lead to gene silencing throughout the genome and have been used to estimate genome accessibility . In contrast , bacteria do not organize their genome in nucleosomes , thus genome accessibility is a subtle feature that is hard to be measured . In general , accessibility is not uniform across the genome due to the presence of global factors such as nucleoid associated proteins ( NAPs ) that alter genomic architecture [15 , 21 , 22] or local factors such as presence of repressor elements that block recruitment of RNA polymerase [21 , 34] . Alteration of global genome structure can lead to changes in gene expression [35 , 36] . For example , NAPs are associated with highly expressed genes that are organized into transcription factories [21] . The challenges in measuring and estimating genome accessibility have impeded the incorporation of this feature into bacterial ChIP-seq analysis . Here , we present a novel biophysically motivated model that incorporate genome accessibility and highlights its importance in assessing TF-binding in bacteria . Extending our previous efforts to mechanistically characterize ChIP-seq coverage information [24] , our model treats ChIP-seq binding profiles as a Boltzmann distribution with two parameters: genome accessibility and binding affinity . We applied this model on a large-scale dataset used to map the regulatory network of M . tuberculosis and compared the results to a simplified model that only considers binding affinity . Our results show that genome accessibility can explain variability in ChIP-seq coverage and peaks , and is associated with specific groups of gene function .
Using ChIP-seq data , biophysically motivated models can provide a quantitative framework for determining key parameters of in vivo TF-binding . We represent the ChIPseq profile in region bins of 500 bp and look for the influence of genome accessibility in TF binding . From a thermodynamic perspective , the probability , pij , that a TF j binds to a genome region i depends on the affinity between the TF and the specific sequence it binds , wij , as well as on the degree that this region is accessible , ai . Formally , the probability of binding is defined by the following equation ( see Methods for detailed derivation ) : log ( pij ) =ai+wij ( 1 ) TF-binding is represented in terms of binding affinity alone by constraining ai = 0 for all i . The accessibility parameter is inferred indirectly by performing linear regression on a large-scale dataset of ChIP-seq experiments [15 , 32] . The affinity parameter is obtained from the motif score . The parameter ai describes a global trend in the probability of binding to region i by any TF . Here , we refer this as the genome accessibility for better biological interpretation of the results . Fig 1A and 1B illustrates schematically how genome accessibility influences TF-binding . Eq 1 is motivated by the poor correlation between ChIP-seq coverage and motif score ( S1 Fig ) . For example , genomic regions with weak motif scores are observed with strong binding signal and vice versa ( Fig 1 ) . We evaluated the extent to which genome accessibility can explain ChIP-seq data . We model our data according to Eq 1 and use a linear fixed effect model to estimate parameters and predict ChIP-seq profiles . The dataset comprises ChIPseq data for a total of 64 unique TFs obtained under same protocol and growth condition ( see Methods ) . The ChIP-seq profile for a specific TF is defined as the normalized abundance of sequence reads that align to each region . The result suggests that the accessibility parameter is a global trend that provides preferential binding on specific genomic regions . We observe that genome accessibility improves prediction of ChIP-seq profiles when compared to a model that considers only binding site affinity . Quantitatively , the accessibility model explains 63% of the observed variance , while motif score alone explains only 35% ( p-value <10−16 , Fig 2 ) . We also explored a more complex representation for binding affinity that considers best motif match , number of binding motifs and a combined score for all motif matches . The combined score is defined as the sum of -log ( pvalue ) for all motif matches . The accessibility values estimated by the more complex model is almost the same as the one estimated by the model that considers only best motif match ( correlation above 99 . 9%; S8 Fig ) . Our model can predict functional features that are useful in ChIP-seq analysis . The most common task in ChIP-seq analysis is the identification of TF-binding peaks , i . e . genomic regions that are bound by the TF under query , which shows a peak in ChIP-seq coverage [5 , 28] . We classify regions into two groups: peaks or not peaks , according to peak-caller method described in previous work [15] . Each region is ranked with a score that indicates how likely they are to contain a peak . Given a threshold , false positives represent regions classified as peak by peak-calling but labeled as not peaks by the ranking score for de novo peak prediction . Similarly , false negatives represent regions that are classified as not peaks by peak-calling but labeled as peaks by the ranking score for de novo peak prediction . The rank for peak classification is defined according to motif and accessibility score and used to construct the ROC curve . Motif score is defined as the maximum log ( p-value ) of motif match per region bin and accessibility score is the estimated value for parameter ai from Eq 4 . We consider three models for peak classification: motif only , motif plus accessibility , and normalized motif plus normalized accessibility . The first model predicts peaks using only motif score obtained by motif scan; the second model uses the sum of motif score and accessibility value; the last model rescale the values of motif score as well as accessibility in the interval from 0 to 1 and use their sum for peak prediction ( see Methods ) . The results show that DNA accessibility improves de novo ChIP-seq peak predictions when compared to predictions that consider motif only . As measured by the area under a receiver operating characteristic ( ROC ) curve , de novo ChIP-seq peak prediction occurs with values 0 . 69 , 0 . 75 , and 0 . 82 for method that uses motif only , motif score plus accessibility , and normalized motif score plus normalized accessibility , respectively ( Fig 3A ) . The affinity values are sequence specific and by definition do not dependent on experimental conditions while the accessibility parameters may vary depending on experimental condition ( S9 Fig ) . Therefore , given that TF-binding affinity score is previously known , one would only need to measure genome accessibility to predict TF-binding under novel growth conditions or for TFs with known binding motifs . This rationale can significantly reduce the need for additional ChIP-seq experiments . The ability to predict ChIP-seq peaks de novo depends on the robustness of the genome accessibility metric and the ease to estimate its parameters under novel experimental conditions . The robustness of DNA accessibility values is illustrated by plotting the accuracy of accessibility values as a function of dataset size used for their estimation , i . e . the expected Pearson correlation between the accessibility estimated in a subset of given size versus the accessibility estimated in the entire dataset ( S2 Fig ) . The expected accuracy for the accessibility values is estimated from 100 distinct samples for each subset size . We observe that as low as 10 ChIP-seq experiments is sufficient to estimate the accessibility values with ~90% accuracy ( Fig 3B and S2 Fig ) . The global trend in genome accessibility is robust to overexpression of a single TF . The ChIP-seq experiments used in this analysis were obtained under the same experimental set , with the exception that the TF under query was overexpressed [15] . We observe that removing any single TF from our dataset does not affect the estimated accessibility value ( correlation between estimates are >99% ) . This indicates that the estimation of genome accessibility is robust to single TF overexpression . Moreover , we observe that just a few ChIP-seq experiments are sufficient to estimate genome accessibility with high correlation to its reference value . Only two ChIP-seq experiments are sufficient to estimate accessibility values with expected 0 . 7 correlation to the reference ( Fig 3B and S2 Fig ) . We also observed that binding profile of some TFs are better correlated with the estimated accessibility values ( S4 Fig ) . This result may indicate TFs that play a key role on genome structure or good candidates to infer genome accessibility . Our model can be used to measure the accessibility state of each region in the genome . We sought to determine if genome accessibility is associated with various genomic features . Consistent with previous studies [37] , intergenic regions are more accessible than protein coding regions ( Fig 4A ) . Genome accessibility also appears to vary between genes or their regulatory regions based on their Clusters of Orthologous Groups ( COG ) assignments . In particular , genes or their regulatory regions in COGs for metabolism and transport of amino acids ( COG category E ) as well as carbohydrates ( COG category G ) are less accessible , while COGs for translation ( COG category J ) and transcription ( COG category K ) are more accessible ( p<0 . 05 after Bonferroni correction; see Fig 4B ) . The observation of higher genome accessibility in transcription and translation genes is consistent with previous observations that DNA structure plays a critical role in expressing rRNA operons [21 , 38 , 39] . Finally , we observe that expression levels are positively correlated with genome accessibility ( R2 = 0 . 23 , Pearson correlation , Fig 4C ) . Interestingly , our results show that the expected expression level is the highest at intermediate values of genome accessibility ( Fig 4D ) , which suggest that there may be a non-linear relationship between accessibility and gene expression . Furthermore , our analysis shows that genome accessibility is biased by genomic position and GC content ( Fig 5 ) . Accessibility has a strong negative correlation with GC content ( Fig 5A ) . In addition , accessibility is negatively correlated with distance to the origin of replication , oriC ( Fig 5C ) , while no apparent correlation is observed in comparison to genome position alone ( Fig 5B ) . This suggests two possible mechanisms that may influence genome accessibility: ( i ) DNA replication makes genomic regions more accessible for TF-binding , or ( ii ) there is a higher copy number of genomic regions near the oriC , leading to an apparent increase in genome accessibility ( Fig 5D ) . These two mechanisms are not necessarily mutually exclusive and would be interesting to explore in future studies .
In this study , we developed a biophysically motivated formulation for bacterial ChIP-seq analysis that contributes to new biological insights of the role that genome accessibility plays in bacterial gene regulation . The model highlights the importance of binding affinity and genome accessibility for in vivo TF-binding . The model formulates the TF-binding process in thermodynamic terms and derives a linear relationship between accessibility , binding affinity , and probability of binding . This relationship enables us to estimate the model parameters from ChIP-seq data . We optimized our statistical framework with a fixed-effects representation to make parameter estimation more computationally efficient . Numerous studies have investigated the role of genome accessibility on TF-biding in eukaryotic organisms [30 , 40–44] . However , to the best of our knowledge , the work described here is the first attempt for a genome-scale quantitative measurement of DNA accessibility in bacteria . In eukaryotes , reads from DNAse I assays are well-correlated with binding regions [40 , 41] . Pique-Regi et al . reported that DNAse I assays can inform genome accessibility for predicting ChIP-seq peaks from ENCODE data using a Bayesian probabilistic model that integrates accessibility with motif information from position weight matrix ( PWM ) , TSS location and evolutionary conservation [29] . Other studies [43 , 44] used a threshold on the coverage of DNAse I signal was used to distinguish accessible from silent genome regions and infer TF-TF interaction as well as set of TFs that drive tissue , cell type , and developmentally specific gene expression patterns in Drosophila . Foat et al . developed a thermodynamics model of binding based on equilibrium dissociation constant between bound and unbound states and used a least square regression model to infer binding affinity from ChIP-chip data of Saccharomyces cerevisiae [42] . However , genome accessibility was not considered in the model . Peng et al . developed another thermodynamic model that includes accessibility and binding energy to predict expression dynamics in Drosophila [30] . Accessibility was inferred from DNAse I assays and model parameters were trained based on an objective function that rewards good fit on highly expressed bins . The method proposed in this paper has several novel features in comparison to those outlined above for eukaryotes . In contrast to eukaryotic genome accessibility models , which are inferred directly from DNAse I assays , our method infer accessibility from binding profiles of multiple ChIP-seq characterized TFs . Our thermodynamics model of TF-binding is derived in terms of binding affinity and genome accessibility by using Lagrange multipliers and free energy of Helmholtz ( see Methods ) . A mixed effects linear regression model is used to make fit efficient and computationally feasible . In addition , the quantitative assessment of DNA accessibility in bacteria provides the possibility of testing hypothesis , novel biological insights , and applications . The framework described here could be used to assess TF-binding using a reduced set of necessary ChIP-seq measurements . Instead of collecting ChIP-seq data for each TF in every new experimental condition , one would only need to perform a small set of experiments to estimate the state of genome accessibility . Then , in combination with established TF affinity data , one can accurately predict TF-binding genome-wide as demonstrated here . This approach could link both in vitro and in vivo experimental datasets under a unifying framework . Our model provides a step forward in our ability to infer TF-binding at different growth states in silico to capture the dynamic nature of gene regulation in bacteria . Biophysical processes in vivo as well as experimental protocols should be considered for proper interpretation of accessibility values . Variance in DNA structure , binding competition , or in vitro artifacts in immuno-precipitation affects the measured genome accessibility . NAPs can shape genome structure at a global scale , while specific genome modification factors can affect accessibility within a particular regulon . Multiple transcription factors that bind to the same genomic region may lead to binding competition , causing a decrease in the observed accessibility . Variations in immuno-precipitation protocols and inherent noise in the technique may lead to variation in the estimation of binding specificity and sensitivity . These and other factors may cause genome accessibility to contain bias from ChIPseq experiments and could be helpful in providing better background estimation . Ultimately , the importance of accessibility in bacteria genome remains to be further explored . In eukaryotic cells , genomic accessibility is critical in fine-tuned gene regulation [45] through controlled activation [46] , minimizing biological noise [47] , and providing epigenetic regulation [33] . These processes may be similarly important in bacteria physiology . For instance , genomic accessibility could cause stochastic gene expression and influence cell fate [48] . Engineering or altering genome accessibility may lead to new approaches in synthetic gene regulation and advance research in systems and synthetic biology [33] . Our work highlights that new biological insights can be obtained through biophysically-motivated mechanistic models of gene regulation . This approach should inspire more refined models of cellular physiology and adaptation . Here , we showed that thermodynamic principles can improve our understanding of TF-binding and genomic structural states . More refined models that integrate accessibility and binding affinity with other factors such as cooperative interaction and multiplicity will enhance our understanding of gene regulation , which will lead to a more comprehensive representation of whole cell physiology [49] .
The probability of TF-binding to a specific region is represented as a Boltzmann distribution that depends on two parameters: accessibility and affinity . The accessibility parameter , ai , is specific to the DNA region and represents how likely a region i is to be bound by any transcription factor . The affinity parameter , wij , represents the specific affinity between a transcription factor j and a region i . Formally , the probability that a TF j binds at region i , pij , is defined as: pij=eai+wij ( 2 ) This representation omits negative signs and the temperature parameter because they are not relevant to the approach in this study . In thermodynamic terms , Eq 2 represents a grand-canonical ensemble in which each region bin can exchange particles ( i . e . TFs ) and energy . The parameter ai represents the chemical potential in region i and the parameter wij represents the energy associated with TF binding ( see S1 Text for detailed mathematical derivation ) . The probability pij can be measured directly from the ChIP-seq data . In order to make this parameter robust and independent on the sequencing depth , we define pij as pij=Ci , j∑iCi , j ( 3 ) where the coverage parameter , Ci , j , represents the number of reads from experiment j that lies in region i . A formal definition for region bins is presented in the next section . Eq ( 2 ) can be transformed to a linear representation . This representation is shown in Eq 1 in the main text and repeated here for clarity: log ( pij ) =ai+wij ( 4 ) Eq 4 permits that we use simple linear regression to estimate the parameters that determine ChIP-seq profiles . This study is restricted to TFs whose binding sites can be summarized by a position weight matrix ( PWM ) . Motif PWM was obtaining as the output of BRACIL [24] . The PWM provides a first order approximation of the affinity between the TF and the region it binds [10 , 50] . We call si , j the affinity score of TF j to region i estimated according to the PWM . This approximation can be placed in Eq 4 , and simplify linear regression as following: log ( pij ) =ai+tj⋅sij ( 5 ) The parameter tj is a constant that represents underlying variables specific to each ChIP-seq experiment , such as TF concentration , ChIP-seq coverage as well as quality of immuno-precipitation . The affinity score , si , j , is defined as the -log10 ( p-value ) of motif match with highest score in region i . Motif scan is performed using FIMO [51] . A affinity score of 2 was given to regions without any motif match . We assume binding affinity to the sequence decreases monotonically with motif p-value . The p-value indicates the probability a score as good as the one observed in motif match occurs by chance according to the reference motif PWM . Thus , the binding affinity is monotonically correlated with–log10 ( p-value ) of a motif match . By expanding it in Taylor series , the term–log10 ( p-value ) becomes a first order approximation for binding affinity that suffices for the purpose of this research . The genome is binned in regions of 500 bp to create a standardize profile and enable comparison of multiple TF experiments simultaneously . Cases with very low coverage are removed from analysis . In numbers , we classified the M . tuberculosis genome in 8824 region bins and only considered data points in which log ( pij ) > -10 . Our rationale is to set up a threshold that considers data points that are informative for analysis and remove noisy ones that decrease the quality of genome accessibility estimation . 82 . 5% of the data points are used for analysis after applying the threshold of log ( pi ) >-10 . This choice is supported by a sensitivity analysis that considers a wide range of minimum coverage threshold ( S5–S7 Figs ) . The results are also robust for varying size of region bins ( S10 Fig ) . We optimized the statistical representation of Eq 5 to make the analysis practical and more efficient . The naïve approach would be to solve Eq 5 by a simple least square minimization . However , the number of data points and parameters needed would exceed 106 and 104 , respectively . The least square minimization by QR decomposition ( function lm in R ) is impractical and we used a linear mixed-effects model ( function lmer , R package lfe ) instead . The linear mixed-effects model optimizes regression because the parameter related to regional accessibility can be described as a random effect that shift the intercept of the probability of binding . As most parameters of Eq 5 correspond to the accessibility value of a region bin , the linear mixed-effects representation makes computation much more efficient . In lmer annotation , our model uses the following formula: `log ( p ) ~ s·t + ( 1|region_bin ) ` , where p , s , and t are general representation of the corresponding parameters in Eq 5 and ` ( 1|region_bin ) `represents the random effect caused by accessibility to each region bin . The model that considers binding affinity only is represented as: `log ( p ) ~ s·t` . We use three methods for de novo peak prediction: motif only , motif + accessibility and normalized motif + normalized accessibility . Motif only rank regions according to best motif match . Motif + accessibility sums the score of motif match ( in terms of -log10 ( pvalue ) ) with the accessibility values estimated from fitting Eq 1 in the data . Finally , we define the minimum score to be 0 and maximum score to be 1 and re-scale motif as well as affinity score accordingly . This sum of the re-scaled score is used to rank regions for the method normalized motif + normalized accessibility . The ChIP-seq data used for this analysis was obtained from a large-scale study that mapped the regulatory network of M . tuberculosis [15 , 32] . The TF under query was FLAG-tagged and over-expressed under control of a mycobacterial tetracycline-inducible promoter . The enriched regions were computed according to the log-normal background model described in [15] . The binding motif was obtained as the output of the algorithm BRACIL [24] , which uses MEME [51] to perform motif identification . FIMO [51] was then used to scan for binding sites at each region . Only TFs that recognize a binding motif with E-value < 10−5 were selected for this analysis . This resulted in a total of 99 ChIP-seq experiments that comprises 64 TFs . Gene expression was defined as the median expression from the set of TF overexpression data , as described previously [15 , 52] . COG categories were obtained from ftp://ftp . ncbi . nih . gov/pub/COG/COG2014/data and mapped to H37rv loci according to GENBANK annotation . The code and corresponding documentation are available at https://sourceforge . net/projects/brasolia .
|
A quantitative description of transcription factor ( TF ) binding in vivo is critical for our understanding of gene regulation . Chromatin Immunoprecipitation followed by sequencing ( ChIP-seq ) provides a genome-scale map of TF-binding . However , a quantitative characterization of the impact of genome accessibility on TF-binding in bacteria remains elusive . In order to help recruit or block gene expression , TFs must have physical access to regulatory regions . This paper presents a thermodynamics model that describes TF-binding in terms of genome accessibility and binding site affinity . We apply this model in a ChIP-seq dataset for Mycobacterium tuberculosis and observed that genome accessibility is critical to our understanding of TF-binding in vivo . This new model provides practical applications , such as de novo prediction of TF-binding peaks and a framework to measure DNA accessibility from ChIP-seq data . Our model enables us to quantify the relationship of genome accessibility with genomic features and suggest mechanisms that influence genome accessibility in vivo ( e . g . distance to oriC ) . The model proposed in this study gives new perspective for ChIP-seq analysis in bacteria towards an improved description of gene regulation in silico .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"bacteriology",
"sequencing",
"techniques",
"engineering",
"and",
"technology",
"gene",
"regulation",
"synthetic",
"biology",
"regulatory",
"proteins",
"microbiology",
"dna-binding",
"proteins",
"invertebrate",
"genomics",
"transcription",
"factors",
"bacterial",
"genetics",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"microbial",
"genetics",
"thermodynamics",
"microbial",
"genomics",
"synthetic",
"genomics",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"bacterial",
"genomics",
"proteins",
"gene",
"expression",
"molecular",
"biology",
"animal",
"genomics",
"physics",
"biochemistry",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics"
] |
2016
|
The Role of Genome Accessibility in Transcription Factor Binding in Bacteria
|
The motility and invasion of Plasmodium parasites is believed to require a cytoplasmic actin-myosin motor associated with a cell surface ligand belonging to the TRAP ( thrombospondin-related anonymous protein ) family . Current models of invasion usually invoke the existence of specific receptors for the TRAP-family ligands on the surface of the host cell; however , the identities of these receptors remain largely unknown . Here , we identify the GPI-linked protein Semaphorin-7A ( CD108 ) as an erythrocyte receptor for the P . falciparum merozoite-specific TRAP homolog ( MTRAP ) by using a systematic screening approach designed to detect extracellular protein interactions . The specificity of the interaction was demonstrated by showing that binding was saturable and by quantifying the equilibrium and kinetic biophysical binding parameters using surface plasmon resonance . We found that two MTRAP monomers interact via their tandem TSR domains with the Sema domains of a Semaphorin-7A homodimer . Known naturally-occurring polymorphisms in Semaphorin-7A did not quantitatively affect MTRAP binding nor did the presence of glycans on the receptor . Attempts to block the interaction during in vitro erythrocyte invasion assays using recombinant proteins and antibodies showed no significant inhibitory effect , suggesting the inaccessibility of the complex to proteinaceous blocking agents . These findings now provide important experimental evidence to support the model that parasite TRAP-family ligands interact with specific host receptors during cellular invasion .
Plasmodium falciparum is the etiological agent of the most severe form of malaria causing over one million deaths annually , primarily in African children [1] . The parasite lifecycle is complex and involves distinct stages that can recognise and invade differentiated cell types of both the human host and the mosquito vector . These stages are characterised by different invasive properties: ookinetes must cross the epithelial cells of the mosquito gut; sporozoites target both the secretory cells of the mosquito salivary glands and the hepatocytes of the human host , which they can either traverse or invade; and merozoites invade human erythrocytes . The ability of each stage to invade their target cells is an obligatory step in the lifecycle of the parasites and therefore these events have been considered attractive points for therapeutic intervention . Invasion and motility requires a single-headed class XIV myosin anchored to the inner membrane complex that unidirectionally propels short actin filaments to impart motive force [2] . The actin filaments are coupled via the glycolytic enzyme aldolase [3] , [4] to parasite cell surface proteins or “invasins” belonging to the TRAP ( thrombospondin-related anonymous protein ) family , which in turn are thought to bind via their extracellular region to host cell surface receptors thereby coupling the actin-myosin power-stroke into forwards movement of the parasite ( Figure 1 A ) . Each different motile form of the parasite is distinguished by its own stage-specific cell surface TRAP-family member [5] . In Plasmodium species , the TRAP-family proteins include TRAP , S6 ( also known as TREP ) , CTRP , MTRAP and TLP . TRAP and S6 are expressed on sporozoites [6] , [7] , [8] , CTRP on ookinetes [9] , MTRAP on merozoites [10] and TLP on both sporozoites and merozoites [11] , [12] . Attempts to target the genes encoding these proteins have shown that most of them are essential for motility and invasion . TRAP is critical for sporozoite invasion of the salivary glands and for infection of mammalian liver as well as sporozoite gliding motility [13]; CTRP is essential for invasion of the mosquito midgut [9]; and S6 is important for sporozoite gliding motility and invasion of mosquito salivary glands [6] , [8] . TLP deletion initially showed no effect indicating a redundant role for this protein [11]; however , recent studies indicate a role in sporozoite cell traversal [12] , [14] . The TRAP-family can be extended to include other cell surface and secreted proteins that contain similar domains and include CSP , SPATR , TRSP , WARP and PTRAMP [5]; PTRAMP , like MTRAP , is expressed in merozoites [15] . To date , it has not been possible to genetically delete MTRAP , indicating it may be essential for parasite growth in blood stage culture [10] . Structurally , TRAP-family proteins are predicted type I cell surface proteins characterised by having one or more extracellular thrombospondin type-I repeats ( TSR ) domains , and/or von Willebrand factor ( vWF ) -like A-domain ( s ) and an acidic cytoplasmic tail with a sub-terminal tryptophan residue [5] . Studies of the individual domains have implicated distinct functions in motility and invasion . The cytoplasmic tail of TRAP , CTRP , TLP and MTRAP have all been shown to interact with aldolase [4] , [10] , [11] , and the cytoplasmic tail of TRAP was shown to be essential for gliding motility and invasion of both salivary glands and hepatocytes [16] . The extracellular TSR and vWF A-domains have been implicated in host cell interaction and invasion; indeed , both the TSR and A-domain of TRAP have been shown to be essential for invasion into both mosquito salivary glands and mammalian hepatocytes [17] , whereas for CTRP , only the A-domains are essential for infectivity [18] . In contrast to the cytoplasmic regions of these proteins , much less is known about the host binding partners for the extracellular regions of TRAP-like proteins and how they play a role in motility , invasion and host cell tropism . The human host extracellular molecules that have so far been identified as binding TRAP-family proteins are not restricted to particular cell types: TRAP ectodomains are known to bind sulphated glycoconjugates [7] , [19] and heparin [20] , [21] , [22] which are both widely distributed molecules . Indeed , both biochemical and functional studies have suggested the presence of additional TRAP receptors on hepatocytes [17] , [20] , [23] but their identities remain unknown; very likely , this is due to the technical challenges of biochemically manipulating membrane proteins and the fact that their extracellular interactions are typified by highly transient interaction strengths [24] . The identification of the host cell surface receptor proteins for TRAP-like parasite ligands remains an important unanswered question towards a better understanding of their role in host cell recognition and invasion ( Figure 1 A ) . Here we report how we have used an assay that is specifically designed to circumvent the technical difficulties in identifying low affinity extracellular interactions called AVEXIS ( for AVidity-based EXtracellular Interaction Screen ) [25] , [26] to identify Semaphorin-7A as an erythrocyte receptor for the P . falciparum merozoite-specific TRAP homolog protein , MTRAP . We report the biochemical characterisation of the interaction and examine its role in erythrocyte invasion .
To identify an erythrocyte receptor for P . falciparum MTRAP , we expressed the entire predicted extracellular region as a secreted recombinant protein in human embryonic kidney ( HEK ) 293E cells . Given the known difficulties in expressing functional Plasmodium proteins [27] , we codon-optimised the MTRAP gene for expression in mammalian cells , replaced the signal peptide with a high-scoring exogenous sequence from a mouse antibody , and mutated the predicted N-linked glycosylation sequons to prevent inappropriate glycan addition that might mask potential protein interaction interfaces . The ectodomain was expressed as both a monomeric and a pentameric his-tagged protein . Pentamerisation was achieved by using a peptide sequence derived from the cartilage oligomeric matrix protein ( COMP ) and was used to increase binding avidity so as to increase the likelihood of detecting transient binding events that are a common feature of extracellular receptor interactions . Both the monomeric and pentameric forms of MTRAP bound human erythrocytes ( Figure 1 B ) relative to controls , which confirmed that the proteins were biochemically active and that MTRAP binds an erythrocyte cell surface receptor . As expected , the binding of the more avid pentameric protein was more resistant to washing steps ( Figure 1 B ) . To determine the molecular identity of the human erythrocyte MTRAP receptor , we took a systematic approach by using the AVEXIS assay and a protein library that represents the cell surface receptor repertoire of the human erythrocyte . This approach has been successfully used to identify basigin as the erythrocyte receptor for P . falciparum RH5 [26] . The pentameric β-lactamase-tagged MTRAP ectodomain was screened against the library of 40 erythrocyte receptor baits used previously . A single interaction was observed ( Figure 1 C , upper panel ) corresponding to Semaphorin-7A ( also known as CD108 ) . The same single interaction was identified in the reciprocal bait-prey orientation ( Figure 1 C , lower panel ) . Semaphorin-7A is a GPI-linked surface protein that is broadly expressed in several tissues [28] , [29] , and particularly on activated lymphocytes where it has been shown to be involved in regulating immune responses [30] , [31] and neurons of both the central and peripheral nervous systems where it has documented roles in axon guidance [32] . Semaphorin-7A is the antigen for the John-Milton-Hagen blood group , although its function on erythrocytes isn't known . To show that MTRAP and Semaphorin-7A interact directly and to quantify the biophysical parameters of the interaction , we used surface plasmon resonance ( SPR ) . The entire ectodomain of Semaphorin-7A was expressed as a soluble recombinant protein and purified before serial dilutions were injected over MTRAP immobilised on a sensor chip . Clear saturable binding was observed ( Figure 1 D ) from which an equilibrium binding constant ( KD ) of 1 . 18±0 . 40 µM was derived . Independent kinetic parameters were in agreement with the equilibrium data ( Table S1 ) and were within the expected range for a typical membrane-tethered receptor-ligand pair that have been shown to have physiological relevance [24] , [26] , [33] . Taken together , these data show that Semaphorin-7A is an erythrocyte receptor for the P . falciparum merozoite TRAP-family ligand , MTRAP . The interaction interface on erythrocyte receptors bound by merozoite surface ligands have been shown , in several cases , to be dependent on the glycosylation state of their erythrocyte receptors [34] , [35] , [36] . Human Semaphorin-7A contains five predicted N-linked glycosylation sites and so to determine whether MTRAP binding was influenced by glycans , we treated recombinant Semaphorin-7A with PNGase F to remove N-linked glycans ( Figure 2 A ) . PNGase F-treated Semaphorin-7A was indistinguishable from untreated Semaphorin-7A in its ability to bind MTRAP using either the AVEXIS assay ( Figure 2 B ) or more quantitative SPR ( Figure 2 C ) . These data suggest that the interaction of MTRAP with Semaphorin-7A is not influenced by the presence of glycans on the receptor . It is known that TRAP can bind sulphated glycoconjugates on hepatocytes [19] . To investigate whether MTRAP was able to bind sulphated glyconjugates , we tested a panel of natural and synthetic glycoconjugates and determined whether they could bind MTRAP using SPR . Chondroitin sulphate A , chondroitin sulphate C , dextran sulphate , heparin and heparan sulphate were each injected at high concentrations over MTRAP immobilised on a sensor chip . No detectable binding for any of the glycoconjugates was observed relative to the Semaphorin-7A positive control ( Figure 2 D ) . We estimate that interactions as weak as 100 µM would have been detected using this approach and conclude that glycoconjugates are unlikely to be major MTRAP ligands . Structural and biochemical studies have shown that semaphorins exist as homodimers [37] , [38] , [39] . Size exclusion chromatography ( SEC ) confirmed that our recombinant soluble monomeric Semaphorin-7A ectodomain eluted in a fraction consistent with it forming a homodimer ( Figure 3 A; top panel ) , as has been shown before [39] . Surprisingly , the ectodomain of MTRAP also eluted at an increased size , perhaps suggesting it also forms a homodimer in solution ( Figure 3 A; middle panel ) . To further investigate these findings , both proteins were subjected to SEC immediately followed by multiangle light scattering ( MALS ) . This analysis demonstrated that while soluble Semaphorin-7A formed quasi-stable homodimers , the soluble MTRAP ectodomain was monomeric ( Figure 3 B ) . The early elution behavior of MTRAP in SEC may therefore be caused by a large hydrodynamic shape possibly due to the protein being highly flexible or adopting a long rod-like shape , as has recently been suggested from atomic force microscopy studies [40] . To determine the stoichiometry of the Semaphorin-7A:MTRAP complex , both purified proteins were mixed at equimolar concentrations and allowed to form a complex prior to separation by SEC . As expected , the complex eluted at a higher mass than each protein individually ( Figure 3 A; bottom panel ) and both proteins were present in these fractions ( Figure 3 C ) . The unusual behaviour of these proteins by SEC made determining the stoichiometry of the complex by this method difficult and so the fraction corresponding to the peak was subjected to amino acid analysis ( Figure 3 D ) . The amino acid compositions determined experimentally were compared to expected theoretical stoichiometries of 1∶1 , 2∶1 and 1∶2 ( Semaphorin-7A∶MTRAP ) . The amino acids that were most characteristic of either Semaphorin-7A or MTRAP indicated that a 1∶1 ratio best fitted the data ( Figure 3 D ) . Calculating the sum of the squared residuals for all amino acids gave values of 1×10−4 , 14×10−4 and 17×10−4 for the 1∶1 , 2∶1 and 1∶2 models , respectively; again , indicating that the 1∶1 ratio best fitted the data . These results therefore suggest that two MTRAP monomers bind one Semaphorin-7A homodimer . The paired TSR domains are the most conserved region of the MTRAP ectodomain across different Plasmodium species , and the TSR domain of TRAP has previously been shown to contribute to receptor binding [19] . To investigate whether the TSR domains of MTRAP contain the Semaphorin-7A binding site , a 74 amino acid fragment that contained both predicted TSR domains ( TSR1+2 ) , and two additional fragments containing each TSR domain individually , were expressed as biotinylated bait proteins ( Figure 4 A ) . The TSR1+2 MTRAP fragment bound to Semaphorin-7A indistinguishably from the entire MTRAP ectodomain using the AVEXIS assay demonstrating that the Semaphorin-7A binding site was localised to the two TSR domains ( Figure 4 B ) . A quantitative analysis using SPR demonstrated a slightly weaker interaction affinity for TSR1+2 ( KD = 1 . 96±0 . 03 µM; Figure 4 C; Table S1 ) compared to the entire MTRAP ectodomain , suggesting that residues outside of the TSR domains make only minor contributions to the binding affinity . Supporting this , purified pentameric TSR1+2 was able to bind erythrocytes ( Figure S1 ) , and it has recently been shown that recombinant MTRAP lacking the TSR domains could not [40] . Neither of the two individual TSR domains bound Semaphorin-7A by AVEXIS ( Figure 4 B ) or SPR ( Figure 4 C lower inset ) demonstrating that the Semaphorin-7A binding site requires both TSR domains . Similarly , to localise the MTRAP binding site on Semaphorin-7A , we expressed constructs containing each of the three recognisable domains: Sema , PSI and Ig-like ( Figure 4 A ) . Only the entire ectodomain of Semaphorin-7A bound MTRAP using the AVEXIS assay , irrespective of the bait-prey orientation ( Figure 4 B , Figure S2 ) . By SPR , however , detectable binding to MTRAP was observed using the Sema domain alone ( Figure 4 D; top graph ) with similar binding parameters to the full-length ectodomain ( KD = 0 . 83±0 . 43 µM ) . No binding was observed with the individual PSI or Ig-like domains ( Figure 4 D; bottom graph ) . Taken together , these data demonstrate Semaphorin-7A and MTRAP directly interact via their Sema and tandem TSR domains respectively . Malaria is thought to have been a powerful selective force in human evolutionary history and given the essential role of MTRAP in parasite blood stage culture we asked whether any naturally-occurring polymorphisms in human Semaphorin-7A would influence the binding of MTRAP . Eight variants in the extracellular region of human Semaphorin-7A are known ( seven within the Sema domain and one within the PSI domain ) and all were individually introduced by site-directed mutagenesis ( Figure 5 A; Table S2 ) . All variants were expressed ( Figure 5 B ) and the dissociation rate constants ( kd ) for MTRAP binding were determined using SPR ( Figure 5 C , Table S3 ) . No significant differences were observed in the interaction strengths for any of the variants suggesting that at least the known common variants in Semaphorin-7A have not been selected due to differences in the ability to bind P . falciparum MTRAP . To examine the role of the MTRAP-Semaphorin-7A interaction in erythrocyte invasion , we attempted to block invasion in vitro using purified recombinant proteins and antibodies raised against either the parasite ligand or erythrocyte receptor . Addition of purified highly avid pentamerised Semaphorin-7A or MTRAP in increasing concentrations had no inhibitory effect on erythrocyte invasion ( Figure 6 A ) , even at concentrations 10-fold higher than the measured interaction strength between monomeric proteins ( Figure 1 D ) . Previous studies of the PfRH5-basigin interaction suggest that antibodies more potently block receptor-ligand interactions during erythrocyte invasion , presumably because their interaction affinities are much higher . We therefore tested an anti-Semaphorin-7a monoclonal antibody in P . falciparum invasion assays at increasing concentrations . No inhibitory activity was seen even at the highest concentrations , unlike a monoclonal directed against the PfRH5 receptor , basigin , which has a >80% invasion inhibitory effect at 10 µg/ml ( Figure 6 B ) . We also raised rabbit polyclonal antibodies against purified , recombinant , monomeric MTRAP and Semaphorin-7A . Both antibodies were able to detect proteins of the expected size in parasite supernatants ( MTRAP ) and erythrocyte ghost preparations ( Semaphorin-7A ) by Western blot ( Figure 6 C ) ; we also showed that the anti-MTRAP antibodies were able to block binding of MTRAP to Semaphorin-7A by AVEXIS ( Figure S3 ) . When added to invasion assays , however , neither had an inhibitory effect on P . falciparum erythrocyte invasion , even at the highest concentration , in contrast to antibodies against AMA-1 ( Figure 6 D ) . Other attempts to block invasion through antibodies targeting MTRAP have yielded similar results [10] , [40] , suggesting that the MTRAP-Semaphorin-7A interaction is either not accessible to blocking agents in in vitro assays , perhaps because it takes place at a late time point during the invasion process , or it is not essential for erythrocyte invasion .
In this study , we have successfully expressed a functional recombinant P . falciparum MTRAP protein and shown that it binds erythrocytes . This protein and a library of human erythrocyte receptor ectodomains were used to identify Semaphorin-7A as its erythrocyte receptor using a systematic screening assay ( AVEXIS ) that is specifically designed to detect low affinity extracellular protein interactions . Importantly , this represents the first example of a host cell surface receptor protein for the TRAP-like family of parasite ligands that provide the crucial link between the target host cell and the parasite's cytoplasmic actin-myosin motor that powers the invasion process in any Plasmodium species . Saturation binding behaviours showed that the MTRAP-Semaphorin-7A interaction was specific and , as expected , was of moderately low affinity as is typical of other measured extracellular receptor-ligand interactions [24] and is consistent with low recovery of bound recombinant MTRAP to erythrocytes performed by others [40] . The biochemical characterisation of the interaction suggests that two MTRAP monomers interact via their tandem TSR domains with the Sema domains of a Semaphorin-7A homodimer . This result is supported by the recent finding that a recombinant MTRAP protein lacking the TSR domains was unable to bind erythrocytes [40] , whereas a protein containing just the two TSR domains could ( Figure S1 ) . This was not unexpected as the TSR domains contribute to binding of other TRAP family proteins to their host cells [5] , [19] and this region is conserved across MTRAP orthologues in other Plasmodium species [10] . In contrast to the sporozoite TRAP protein , we could find no evidence that MTRAP bound sulphated glycoconjugates despite using a highly sensitive assay . This might be explained by the presence of the sequence “WSPCSVTC” in the TSR domains of TRAP , TLP and the related protein CSP which is believed to be a sulfatide binding motif ( Muller et al . , 1993 ) and is absent from MTRAP . Recent work by Uchime et al . confirmed that the TSR domains of MTRAP structurally differ from previously studied TSR domains based on its disulphide bonds suggesting a more compact structure , perhaps indicating that both TSR domains function together [40] and explaining the requirement for tandem TSR domains in Semaphorin-7A binding . Our experiments to define the MTRAP binding site on Semaphorin-7A were complicated by the fact that Semaphorin-7A , like other semaphorins , is known to form a dimer with a large ( 2860 Å2 ) and primarily hydrophobic contact interface that involves the whole molecule , including the Ig domain [39] . Individually expressing each of the constituent domains to map the MTRAP binding site therefore disrupted this homodimeric structure . MTRAP , however , did interact with a construct containing the Sema domain alone immobilised at sufficient density on a Biacore chip ( Figure 4 D ) suggesting that the Sema domain contained the MTRAP interaction interface . We also established that the stoichiometry of binding is likely to be two MTRAP monomers binding to a single Semaphorin-7A homodimer . This binding model is also used by the endogenous semaphorin ligands , the plexins , as shown by crystallisation of the complex [39] , [41] . It is possible , as for the plexins , that binding of MTRAP to a dimeric receptor triggers local MTRAP clustering which is then necessary for function by bringing the cytoplasmic regions into close proximity . Attempts to genetically disrupt MTRAP in multiple P . falciparum strains were unsuccessful suggesting that it is essential for blood stage growth [10] . We therefore attempted to disrupt the MTRAP-Semaphorin-7A interaction; however , neither purified highly-avid pentameric proteins of both MTRAP or Semaphorin-7A nor polyclonal antibodies raised against either MTRAP or Semaphorin-7A showed any discernible effect on erythrocyte invasion in vitro . The inability of polyclonal antibodies raised against MTRAP to affect invasion is consistent with findings from other groups and suggests that MTRAP is unlikely to be a component of an effective subunit blood-stage vaccine [10] , [40] . MTRAP may therefore have an important receptor-independent role similar to TRAP which is required not only for cellular invasion but also gliding motility [13] . One other possible explanation is that the interaction may occur in close physical proximity to the moving junction - an electron dense thickening formed at the nexus of the erythrocyte and merozoite plasma membranes , which would almost certainly be inaccessible to large soluble proteinaceous blocking reagents . This hypothesis can be supported by the lack of non-synonymous polymorphisms found in the MTRAP ectodomain , suggesting that unlike other merozoite ligands it is not exposed to strong selection pressure by the immune system [10] . In agreement with this , MTRAP does not appear to be a primary target of the adaptive immune system , with low anti-MTRAP reactivity in human sera from malaria endemic regions [40] . This is in clear contrast to the highly polymorphic sporozoite TRAP protein [19] . Similarly , Semaphorin-7A had very few known polymorphisms , all of which did not quantitatively affect its interaction with MTRAP , and to our knowledge , there is no evidence of selective pressure on this receptor in malaria endemic regions . Interestingly , loss of Semaphorin-7A expression on erythrocytes can be acquired , typically with increasing age , and levels of erythrocyte Semaphorin-7A expression have been observed to fluctuate significantly during pregnancy [42] . Our finding that Semaphorin-7A is a receptor for P . falciparum MTRAP makes correlating the levels of cell surface Semaphorin-7A with clinical parameters of P . falciparum infection an important area for future study . Our finding that Semaphorin-7A is a receptor for MTRAP provides the first example of a host receptor protein for a member of the TRAP-like family . Semaphorin-7A , similar to MTRAP , is a member of a larger family of cell surface proteins , the semaphorins , that can be subdivided into eight different structural classes [43] . Whether the other members of the TRAP-like family will have identifiable receptors within the broader semaphorin family is a key area for future research . Whether the essentiality of MTRAP simply lies in its function as a membrane-tethered link to the parasite cytoplasmic myosin-based motor or has additional roles in determining the cellular tropism of invasion is still not clear . However , the identification of an erythrocyte receptor for the extracellular region of MTRAP supports a mechanism whereby TRAP-family ligands directly interact with a protein displayed on the surface of the target cell . This interaction may therefore provide the traction required to couple the activity of the parasite myosin-based motor into a relative cellular movement that is necessary for invasion . We believe that this finding together with the successful demonstration of an experimental approach to identify host receptors for parasite TRAP-like ligands will stimulate further research into the challenging task of identifying receptors for this important class of parasite ligands .
Use of erythrocytes and serum from human donors for P . falciparum culture was approved by the NHS Cambridgeshire 4 Research Ethics Committee . All subjects provided written informed consent . The use of animals to raise antisera was performed according to UK Home Office governmental regulations and approved by the local Sanger Institute ethical review board . A list of the erythrocyte receptor proteins used in this study and the numbering used in Figure 1 C can be found in Supplementary Table 1 in Crosnier et al . , 2011 . Proteins within the human erythrocyte protein library were produced as bait and prey constructs as previously described [26] . Briefly , for the proteins containing a signal peptide , each expression construct contained the entire extracellular region ( including the native signal peptide ) flanked by unique NotI and AscI sites to facilitate cloning into a vector containing a C-terminal rat Cd4d3+4-tag and either an enzymatically biotinylatable peptide ( baits ) or a peptide from the rat cartilage oligomeric matrix protein ( COMP ) which spontaneously forms pentamers followed by the enzyme beta-lactamase ( preys ) . The ectodomain fragments of the four type II proteins ( which lack a signal peptide ) were expressed only as monomeric baits and not preys . Baits for the type II proteins were produced by flanking the predicted extracellular regions with NotI and AscI restriction enzymes and cloning them into a vector containing a mouse immunoglobulin kappa light chain signal peptide followed by the biotinylatable tag and Cd4d3+4 at the N-terminus of the insert . Bait proteins were enzymatically biotinylated during expression by cotransfection of a secreted form of the E . coli BirA protein biotin ligase [25] . The MTRAP ectodomain bait and prey constructs differed from the erythrocyte receptors in that the low-scoring endogenous signal peptide was replaced by a high-scoring signal peptide from a mouse immunoglobulin kappa light chain and the serines and threonines in the context of potential N-linked glycan sites were systematically mutated to alanine to prevent inappropriate glycosylation . All ectodomains were codon optimised for mammalian expression and chemically synthesized ( Geneart AG , Regensburg , Germany ) . The constituent Sema , PSI and Ig-like domains of human Semaphorin-7A were produced by identifying the domain boundaries using the crystal structure of the Semaphorin-7A extracellular region as a guide [39] . The MTRAP TSR1+2 , TSR 1 and TSR 2 domain boundaries were estimated based on the location of conserved cysteine residues identified by protein alignments of TSR repeats . The sequences corresponding to these domains were amplified using primers with flanking NotI and AscI cloning sites for cloning into the appropriate expression vectors . The PSI , Ig and all TSR domains were cloned into the same vectors as MTRAP to add an exogenous signal peptide required for protein secretion . Naturally-occurring variants of Semaphorin-7A were found in dbSNP ( www . ncbi . nlm . nih . gov/projects/SNP/ ) . Constructs containing these variants were produced by site directed mutagenesis ( GeneArt AG ) . Variants were mapped onto the structure of Semaphorin-7A using PyMOL ( www . pymol . org ) . Monomeric His-tagged proteins were prepared by subcloning the NotI/AscI flanked extracellular regions into a vector containing a C-terminal Cd4d3+4 tag followed by a hexa-His tag [25] . An additional monomeric His-tagged MTRAP lacking the Cd4d3+4-tag was produced by amplifying the MTRAP coding region with primers containing NotI and EcoRI sites and inserting into a NotI/EcoRI-digested His-vector using standard cloning procedures . Pentameric His-tagged proteins were similarly made by cloning the inserts into a NotI/EcoRI-digested prey vector where the COMP-beta-lactamase region had been replaced by a COMP-hexa-His tag . All proteins were expressed as secreted proteins by transient transfection of the human HEK293E cell line grown in suspension as described [25] , [44] . His-tagged proteins were purified from supernatants from transient transfections on HisTrap HP columns ( GE Healthcare ) using an ÄKTAxpress ( GE Healthcare ) according to manufacturer's instructions . Size exclusion chromatography of nickel purified samples was carried out on a Superdex 200 Tricorn 10/600 column ( GE Healthcare ) in HBS-EP ( GE Healthcare ) . Amino acid analysis was performed by the PNAC Facility , University of Cambridge , Cambridge , UK . For PNGase F treatment , biotinylated Semaphorin-7A was incubated with 50 U/µl of PNGase F ( NEB ) for 10 , 30 and 60 min at 37°C for Western blot analysis , and 60 min at 37°C for AVEXIS and Biacore analysis . The P . falciparum AMA-1 ectodomain was produced in a similar way to MTRAP , cloned into the vector containing a C-terminal Cd4d3+4 tag followed by a hexa-His tag , then expressed and purified as described above . Interaction screening was carried out as previously described [25] , [26] . Briefly , both bait and prey protein preparations were normalised to activities that have been previously shown to detect transient interactions ( monomeric half-lives less than 0 . 1 second ) with a low false positive rate [25] . Biotinylated baits that had been dialysed against HBS were immobilised in the wells of a streptavidin-coated 96-well microtitre plate ( NUNC ) . Normalised preys were added , incubated for 1 hour at room temperature , washed three times in HBS plus 0 . 1% Tween- 20 , and once in HBS , after which 125 µg/ml of nitrocefin was added and absorbance values measured at 485 nm on a Pherastar plus ( BMG laboratories ) . For the screen , a positive control interaction using rat Cd200 as a bait and rat Cd200R as a prey , and a negative control interaction using rat Cd4d3+4 as a bait and rat Cd200R as a prey , was used ( + and − in Figure 1 C ) . Where AVEXIS was used for interaction site mapping and PNGase F experiments ( Figure 2 B and 4 B ) , the Cd4d3+4 tag alone was used as a negative control bait , and a biotinylated anti-Cd4 antibody as positive control to capture the Cd4d3+4-tagged prey . Erythrocyte binding assays were carried out as described previously [45] but with slight modifications . Briefly , 60 µg of purified proteins were mixed with 50 µl of packed fresh erythrocytes for 2 hours at 4°C . The erythrocytes were separated from supernatant by spinning through 400 µl of ice cold dibutyl phthalate ( Sigma ) at 12000 g for 30 seconds , after which the erythrocyte pellet was washed in ice cold PBS . Proteins bound to the erythrocytes were eluted by incubating with 20 µl of 1 . 5 M NaCl at room temperature for 45 minutes , and collected after 30 seconds of 12 , 000 g centrifugation . The unbound , wash and eluted material were analysed by Western blotting as described below . Erythrocytes were pelleted then washed twice in 5 volumes of ice cold 2 mM HEPES , 154 mM NaCl , pH 7 . 1 and centrifuged for 15 mins at 500 g at 4°C . The pellet was transferred into 15 volumes of ice cold 10 mM Tris-HCl 1 mM EDTA pH 7 . 1 and left on ice for 30 mins . After centrifugation at 20 , 000 g for 15 mins at 4°C , the supernatant was discarded and the pellet gently resuspended whilst leaving behind the denser dark pellet of unlysed cells . The pellet was centrifuged at 20 , 000 g for 15 mins at 4°C then washed in 2 mM HEPES , 154 mM NaCl , pH 7 . 1 four times . The washed ghosts were resuspended in 10 mM Tris-HCl pH 7 . 1 and centrifuged at 20 000 g for 15 mins at 4°C , after which , the pellet was resuspended in 1 volume of 10 mM Tris-HCl pH 7 . 1 , then stored at −20°C . To make culture supernatants , synchronised schizonts were purified by centrifugation onto an 80% Percoll cushion , collected at the cushion interface , placed back into in vitro culture at 2 . 5×107 parasites/ml in the absence of additional erythrocytes and allowed to rupture overnight . Cells were removed by centrifugation and supernatants stored at −80°C . To raise polyclonal antisera against Semaphorin-7A , MTRAP , and AMA-1 , purified proteins were injected into rabbits ( Cambridge Research Biochemicals , Billingham , UK ) . The sera were purified on Hi-Trap Protein G HP columns ( GE Healthcare ) , and the mouse anti-human Semaphorin-7A , MEM-150 monoclonal antibody [46] was purified from mouse ascites on a HiTrap IgM Purification HP column ( GE Healthcare ) , using an ÄKTA Xpress ( GE Healthcare ) according to the manufacturer's instructions . Proteins were resolved by SDS-PAGE using NuPAGE 4–12% Bis Tris precast gels ( Invitrogen ) . Where reducing conditions were required NuPAGE reducing agent and anti-oxidant ( Invitrogen ) were added to the sample and the running buffer , respectively . Proteins were blotted onto PVDF membranes ( Amersham ) and blocked in 2% BSA . Membranes were incubated with either peroxidase-conjugated streptavidin ( Jackson Immuno Research ) or anti-C-term His-HRP antibody ( Invitrogen ) as appropriate , and proteins detected using SuperSignal West Pico Chemiluminescent substrate ( Thermo Scientific ) . When using rabbit-anti-Semaphorin-7A or anti-MTRAP , an anti-rabbit-IgG-HRP ( Invitrogen ) secondary antibody was used . Surface plasmon resonance studies were performed using a Biacore T100 instrument ( GE Healthcare ) . Biotinylated bait proteins were captured on a streptavidin-coated sensor chip ( GE Healthcare ) . Approximately 150 RU of the negative control bait ( biotinylated rat Cd4d3+4 ) were immobilised in the flow cell used as a reference and approximate molar equivalents of the query protein immobilized in other flow cells . Purified analyte proteins were separated by size exclusion chromatography on a Superdex 200 Tricorn 10/600 column ( GE Healthcare ) in HBS-EP ( GE Healthcare ) just prior to use in SPR experiments to remove any protein aggregates that might influence kinetic measurements . Increasing concentrations of purified proteins were injected at 100 µl/min to determine kinetic parameters , or at 20 µl/min for equilibrium measurements . The surface was regenerated with a pulse of 2 M NaCl at the end of each cycle . Duplicate injections of the same concentration in each experiment were super imposable demonstrating no loss of activity after regenerating the surface . Both kinetic and equilibrium binding data were analysed in the manufacturer's Biacore T100 evaluation software version 1 . 1 . 1 ( GE Healthcare ) . Equilibrium binding measurements were taken once equilibrium had been reached using reference-subtracted sensorgrams . Both the kinetic and equilibrium binding were replicated using independent protein preparations of both ligand and analyte proteins . All experiments were performed at 37°C in HBS-EP . Sulfated-glycoconjugates were obtained from Sigma and resuspended in 1× HBS-EP ( Biacore , GE Healthcare ) and used for surface plasmon resonance studies at 1 mg/ml . Size exclusion chromatography was performed on a Superdex200 10/30 column ( GE Healthcare ) equilibrated in 50 mM Tris . HCl , pH 7 . 5 , 150 mM NaCl at 0 . 4 ml/min . The column was followed in-line by a Dawn Heleos-II light scattering detector ( Wyatt Technologies ) and an Optilab-Rex refractive index monitor ( Wyatt Technologies ) . Molecular mass calculations were performed using ASTRA 5 . 3 . 4 . 14 ( Wyatt Technologies ) assuming a dn/dc value of 0 . 186 ml/g . The 3D7 P . falciparum parasite strain was cultured in human O+ erythrocytes at 5% hematocrit in complete medium ( RPMI-1640 containing 10% human serum ) , under an atmosphere of 1% O2 , 3% CO2 , and 96% N2 . Invasion assays were carried out in round-bottom 96-well plates , with a culture volume of 100 µL per well at a hematocrit of 2% . Parasites were synchronized at early stages with 5% ( w/v ) D-sorbitol ( Sigma ) , trophozoite stage parasites were mixed with the specified protein blocking reagent , and then incubated in the plates for 24 hours at 37°C inside a static incubator culture chamber ( VWR ) , gassed with 1% O2 , 3% CO2 , and 96% N2 . At the end of the incubation period , erythrocytes were harvested and parasitized erythrocytes were stained with 2 µM Hoechst 33342 ( Invitrogen ) , as described previously [47] . Purified MEM-150 , rabbit polyclonal antibodies and pentamerised MTRAP and Semaphorin-7A ectodomains were dialysed into RPMI ( GIBCO ) prior to use . A monoclonal antibody targeting basigin ( MEM-M6/6 , Abcam , Cambridge , UK ) was purchased and dialysed into RPMI before addition into invasion assays . Hoechst 33342 ( Invitrogen ) stained samples were excited with a 355 nm UV laser ( 20 mW ) on a BD LSRII flow cytometer ( BD Biosciences ) and detected with a 450/50 filter . BD FACS Diva ( BD Biosciences ) was used to collect 100 , 000 events for each sample . FSC and SSC voltages of 423 and 198 , respectively , and a threshold of 2 , 000 on FSC were applied to gate the erythrocyte population . The data collected were further analyzed with FlowJo ( Tree Star ) . All experiments were carried out in triplicate . GraphPad Prism ( GraphPad Software ) was used to plot the generated parasitemia data .
|
Apicomplexan parasites are one of the most significant groups of pathogens infecting humans and include Plasmodium falciparum , the parasite responsible for malaria . These parasites critically depend on their human host and must invade our cells to multiply; therefore , understanding this invasion process - with the eventual aim of therapeutically preventing it - has been a focus for scientific investigation . A key component of the invasion machinery is a family of proteins ( the “TRAP” family ) which traverse the membrane surrounding the parasite: the part remaining within the parasite connects to a molecular motor that powers invasion , whilst the surface-exposed region is thought to interact with proteins on the surface of the target host cell . One major question that remains unanswered is the identity of the host receptors for the TRAPs . In our paper , we use a method specifically designed to detect interactions that occur in the extracellular space between host and pathogen proteins to reveal a host receptor called Semaphorin-7A for the TRAP-family member used by the blood stage of the malarial parasite – a protein called MTRAP . The characterization of this host-parasite interaction may therefore lead to novel therapies based upon preventing parasite invasion .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biomacromolecule-ligand",
"interactions",
"protein",
"interactions",
"microbiology",
"host-pathogen",
"interaction",
"parasitology",
"glycoproteins",
"proteins",
"cell",
"adhesion",
"extracellular",
"matrix",
"biology",
"recombinant",
"proteins",
"proteomics",
"biochemistry",
"transmembrane",
"proteins",
"cell",
"biology",
"molecular",
"cell",
"biology",
"glycobiology"
] |
2012
|
Semaphorin-7A Is an Erythrocyte Receptor for P. falciparum Merozoite-Specific TRAP Homolog, MTRAP
|
Imaging studies have revealed a putative neural account of emotional bias in decision making . However , it has been difficult in previous studies to identify the causal role of the different sub-regions involved in decision making . The Ultimatum Game ( UG ) is a game to study the punishment of norm-violating behavior . In a previous influential paper on UG it was suggested that frontal insular cortex has a pivotal role in the rejection response . This view has not been reconciled with a vast literature that attributes a crucial role in emotional decision making to a subcortical structure ( i . e . , amygdala ) . In this study we propose an anatomy-informed model that may join these views . We also present a design that detects the functional anatomical response to unfair proposals in a subcortical network that mediates rapid reactive responses . We used a functional MRI paradigm to study the early components of decision making and challenged our paradigm with the introduction of a pharmacological intervention to perturb the elicited behavioral and neural response . Benzodiazepine treatment decreased the rejection rate ( from 37 . 6% to 19 . 0% ) concomitantly with a diminished amygdala response to unfair proposals , and this in spite of an unchanged feeling of unfairness and unchanged insular response . In the control group , rejection was directly linked to an increase in amygdala activity . These results allow a functional anatomical detection of the early neural components of rejection associated with the initial reactive emotional response . Thus , the act of immediate rejection seems to be mediated by the limbic system and is not solely driven by cortical processes , as previously suggested . Our results also prompt an ethical discussion as we demonstrated that a commonly used drug influences core functions in the human brain that underlie individual autonomy and economic decision making .
Research within behavioral economics and psychology has demonstrated that human decisions are based on more dimensions than simply maximization of monetary reward [1]–[3] . One important factor with prominent impact on decision making is the influence of emotional processes [4] . Emotional processes include both emotional responses [5] and feeling states [6] . Emotional responses are rapid and automatic in order to meet the demands for fast contextual adaptation . On the other hand , the representation of feeling states and the regulatory control of emotions reflect a slower adjustment to long-term considerations and goals [6] . Recently , brain imaging studies have shown that emotional systems are active in decision making [2] , [7]–[9] . However , it is difficult to identify the causal role of the different sub-regions involved in decision making based on these correlational studies . A human universal in social cooperation is the tendency to respond with immediate aggression upon perceived threat or unfairness [10] . Evolution seems to have favored the act of punishing those who violate perceived norms of the group [11] . A suitable paradigm to study the punishment of norm-violating behavior is the Ultimatum Game ( UG ) [12] . In the UG , a proposer suggests a way to divide a fixed sum of money . The responder has to accept or reject the proposal . If the responder accepts the proposal , the suggested split is realized . If the responder rejects the offer , neither of the two gets anything . Proposers often offer an even split , and unfair offers are frequently rejected; offers of ≤20% are rejected roughly half of the time [3] . These findings are robust with respect to learning effects , stake size , and other manipulations [3] . Although both individual genetic traits and cultural variation influence the response pattern , the general propensity to punish norm violators seems to be universal [13] , [14] . While there have been previous attempts within the field to include anatomical information in the behaviorally validated theoretical models of decision making , a mechanistic explanation of how emotions actually influence choice is still missing [15] . Existing interdisciplinary economic models are incomplete [16] in that the biological framework is not fully defined . Several studies have suggested that emotional theory may add important information for choice behavior [1] , [2] . The evolution of the frontal lobes in humans has extended the ability for long-term reasoning [17] . In a pioneering paper Sanfey et al . [7] suggested that the forebrain network ( anterior insula , dorsolateral prefrontal cortex [dlPFC] , and anterior cingulate cortex [ACC] ) has a pivotal role in the rejection response in the UG [7] . This result is at variance with a vast literature that attributes a crucial role in emotional decision making to a subcortical structure ( i . e . , the amygdala ) [2] , [4] , [18] . Here we propose an anatomy-informed model that may link these seemingly opposing views . Gläscher et al . [19] proposed that intuitive decisions could be considered as cognitively model-free reinforcement learning ( RL ) , whereas deliberate choices could be viewed as model-based RL . By combining computational learning models with functional imaging data they found that model-free RL was associated with subcortical processing and model-based RL was linked to cortical processing . They suggested that decision making involves at least two neural networks that seem to have distinct neural correlates . Thus , both cortical and subcortical levels may influence the final behavior in either direction in the UG , and the major difference is that the cortical level has a richer representation of future outcomes of a decision [17] , [19] . In the UG , the payoff-maximizing strategy for the responder ( individual level ) is to accept all offers , and reciprocally for the proposer it is to make the smallest possible offer [12] . The UG demands a simple and rapid yes/no answer , but the underlying reason for a response is complex and extends beyond payoff maximization and includes influences from social interactions . For example , the responder needs to consider: social hierarchy , tit-for-tat , preparing for the next encounter , maintenance of social norms , reputation building , and avoidance of social rejection . Short-term unaware responses are instantiated in the subcortical emotion system ( e . g . , the amygdala ) [4] , [20] , [21] , whereas the long-term considerations pertain to frontal cortex and insula [22] , [23] . In order to reconcile the paradox in the literature on the UG [7] , [24] and decision making [2] , [4] , we hypothesized that the response in the UG rests on a balance between phylogenetically older structures involved in the automatic reactive emotional response and neocortical areas associated with the neural processing of feeling states , future representation , and regulation of emotions [5] , [6] , [17] . Previous imaging studies [7] , [24] on decision making in the UG did not specifically aim to separate instant automatic responses from slower affective processes associated with awareness [6] . The immediate responses are likely to be transient and mitigated when emotional regulation sets in [25] . Thus , a prerequisite to detect these responses is to have a strictly defined onset time for when unfairness may be detected . Fast automatic responses in the UG have not previously been captured since the proposals were presented for 6 s [7] , [24] , thereby not providing a clear definition of the onset time for possible detection of unfairness . Thus , in the previous studies , there is a possibility that the experimental design precluded a detection of early and transient responses . In our experiment , the proposals were given orally in movie clips , and we formulated the UG proposal so as to maintain ambiguity of fairness until the final word of the proposal ( i . e . , the amount that would be taken by the proposer was spoken last ) , yielding a well-defined onset time ( Figure 1 ) . As the amygdala is crucial for both the mediation of aggressive responses [26] and of biasing decision making [2] , [4] , we suggest a parallel between reactive aggression and the behavior associated with rejection . Thus , we hypothesized that the amygdala drives immediate rejection in the UG . GABA receptors are abundant in the amygdala , and benzodiazepines can potentiate GABA activity , reduce behavioral signs of aggression [26] , and decrease amygdala activity in emotional tasks [27] , [28] . In the present study , we posited that the benzodiazepine oxazepam ( 20 mg orally ) could inhibit amygdala activity and , thus , change behavior in the UG . We assumed that unfair proposals would generate an amygdala response and increase rejection rate in the non-medicated group , while oxazepam would inhibit this process and therefore reduce the rejection rate in response to unfair proposals , in parallel with a mitigated amygdala response to unfair proposals . Thus , to test whether the amygdala is involved in the rejection of unfair offers , we randomly allocated subjects to a treatment or a placebo group prior to scanning them while playing the UG . As will be seen , our design allowed for the detection of the functional anatomical response to unfair proposals in a subcortical network that mediates rapid reactive responses . This response was possible to manipulate selectively with a benzodiazepine both on the behavioral and the neural response level .
In line with previous studies , fair proposals were never rejected in either group [7] , [29] . The rejection rate ( not controlled for gender ) for unfair proposals was significantly lower in the oxazepam group ( 19 . 0%; n = 18 ) than in the placebo group ( 37 . 6%; n = 17; p = 0 . 0475; Figure 2A ) . The same comparison but controlling for gender in a probit regression showed borderline significant results ( p = 0 . 0675 ) ; the predicted rejection rate was 19 . 5% in the oxazepam group and 36 . 1% in the placebo group . However , the gender coefficient was insignificant ( p = 0 . 278 ) . The treatment-induced drop in the rejection rate was 20 percentage points for men and 15 percentage points for women . Neither the rejection rate nor the effect of treatment on the rejection rate varied significantly between men and women ( see Text S1 for details ) . The functional MRI ( fMRI ) contrast of unfair versus fair in the placebo group essentially confirmed the results from Sanfey et al . [7] ( Figure 3A and 3B ) . The corresponding contrast in the oxazepam group showed a subsignificant activation in the right insula ( see Figure S1 and Table S1 ) . Given that oxazepam inhibits rejection of unfair offers , the interesting contrast is the interaction that probes for changes in response to unfairness with or without oxazepam ( placebo unfair−fair proposals>oxazepam unfair−fair proposals ) . We confirmed our primary hypothesis that the amygdala was relatively more activated in the placebo group than in the oxazepam group for unfair offers ( left amygdala: Montreal Neurological Institute space coordinates ( x , y , z ) [−18 −6 −18] , Z = 3 . 25 , p = 0 . 05 , corrected; right amygdala: [18 0 −18] , Z = 3 . 03 , p<0 . 05 , corrected; Figure 3D ) . The amygdala response for the different conditions is also visualized in Figure 4 , where we show that both groups have similar amygdala activation patterns in response to the fair condition; however , unfairness up-regulates the amygdala response in the placebo condition while oxazepam reduces the amygdala response in the treatment group . Moreover , the extended fMRI analysis revealed interaction differences in medial prefrontal cortex ( mPFC ) ( [−6 66 18] , Z = 3 . 77 , p<0 . 05 , cluster level corrected ) and right ACC ( [9 48 24] , Z = 3 . 57 , p<0 . 05 , cluster level corrected ) ( Figure 3C ) . Thus , subjects who were treated with oxazepam had a diminished activation in a subset of regions in the neural network normally activated by unfair proposals [7] . The change in rejection rate induced by oxazepam did not result in any significant effects in the dlPFC or the insula , two areas that have previously been linked to the rejection of unfair offers in the UG [7] , [29] . Given the previous data in the literature , we extended the analyses and performed post hoc specific searches with a single-region-of-interest approach in the dlPFC and the insula . We noted effects in the interaction also in these regions ( see Text S1 ) . Importantly , the general cerebral response in the decision making task was unaltered by the drug ( see Text S1 and Figure S2 ) . We predicted that increased amygdala activity would correspond to an increased rejection rate also within the placebo group , as the amygdala is known to have a crucial role in decision making [2] , [4] and aggression [26] . To test this hypothesis we did a within-subject analysis in placebo subjects that both accepted and rejected unfair proposals ( n = 6 ) in both experimental sessions . The contrast unfair proposals rejected>unfair proposals accepted showed increased amygdala activity ( [21 −3 −12] , Z = 3 . 50 , p<0 . 05 , corrected; Figure 5A ) . As testosterone can increase aggressive behavior [26] , we tested whether males drive the reactive amygdala response and hence would show an increased amygdala activity in response to unfairness . To test this hypothesis , we compared amygdala activity between sexes for the contrast placebo unfair−fair proposals>oxazepam unfair−fair proposals . Strikingly , males ( n = 5 ) showed a greater right amygdala activity than females ( n = 12 ) in the placebo condition , while there was no difference between sexes in the oxazepam condition ( males , n = 8; females , n = 10; Figure 5B ) . Thus , the interaction sex × treatment was significant ( F[1] = 8 . 50 , p = 0 . 007 ) . However , it is important to emphasize that the effect seen in the interaction contrast cannot be explained as a gender effect as the result remains significant ( left amygdala: p = 0 . 035; right amygdala: p = 0 . 042 ) even after adjusting for gender ( see Text S1 ) . The subjects treated with oxazepam displayed a decreased rejection rate to unfair proposals ( Figure 2A ) . In order to probe the possibility that the change in behavior was due to drug-altered perception of unfairness or likeability of the proposer , we compared subjective ratings between groups . Subjects in the oxazepam group had similar perception of unfairness ( Mann-Whitney U test , two-tailed , p = 0 . 5103; Figure 2B ) and perceived likeability of the proposers ( Mann-Whitney U test , two-tailed , Z = 0 . 63 , p = 0 . 5467 ) as in the control group ( Figure 2C ) . This is in concordance with the finding that there was no difference in insula activity between the groups in unfair versus fair offers ( see Figure S1 ) ( the insula is involved in the coding of feeling states [6] ) . Thus , we found that the observed change in choice behavior between the treatment groups was not explained by an altered feeling of unfairness or insula activity .
We are the first , to our knowledge , to show the functional anatomical response to unfair proposals in a subcortical network for rapid reactive responses . Our results suggest that the act of immediate rejection of unfair proposals is driven by a phylogenetically old structure ( the amygdala ) and may be viewed as a reactive aggressive response . This finding adds additional information , as previous UG studies have shown involvement of only a cortical network [7] , [24] . We propose that the subcortical and the cortical networks can operate separately and that the major differences between the two are that cortical networks have a richer future representation and operate more slowly [17] , [19] . Moreover , we demonstrated that the amygdala-driven rejection response was inhibited with oxazepam treatment without affecting the perception of unfairness . This suggests that the GABA system can influence the decision making network via an alteration of the balance between phylogenetically young ( prefrontal cortex ) and old structures ( amygdala ) . As timing is crucial for detection of transient responses [30] , our design had the necessary elaboration to allow detection of fast automatic emotional responses to unfairness and not only slow components . We observed a clear amygdala activation that is in line with previous studies ( with proper onset timing ) on emotional bias in decision making [2] , [28] . Our study generates two arguments for a causal role of the amygdala in the generation of an instant rejection response . First , in the treatment group , the amygdala response was mitigated in conjunction with a decreased rejection rate ( as compared with the placebo group ) . Second , in the within-subjects comparison in the unmedicated group , rejections were associated with increased amygdala activity . In light of this , we question the suggested exclusive causality of the insula in the generation of a rejection response that was derived from the correlation between insula activity and acceptance rate of UG offers [7] . Instead , we propose that the amygdala is involved in instant rejection of unfair UG offers , whereas the insula might be more involved in a late rejection response . Importantly , neither result excludes that separate neural operations can give rise to the same behavior . Moreover , it is important for future research to segregate which responses are related to perception of unfairness and which drive rejection behavior , and how these neural processes interact . Our data are compatible with the two-level model for decision making [19] . We have pharmacologically manipulated both levels in our study . The effects we observed may be based either on direct pharmacological action or on interaction effects between regions in the decision making circuit . In another manipulation of this circuit Knoch et al . [29] showed that transcranial magnetic stimulation of the dlPFC leads to an increased acceptance rate for unfair proposals in the UG , without changing the perception of unfairness . The authors concluded that dlPFC exclusively drives rejection in response to unfair proposals . We suggest that the interpretation could be modified , as dlPFC does not seem to play a role in low-level model-free decision making [19] , [22] . In support of this modified interpretation we note that , in a decision making task where the subject had to remember explicit values supporting the decision , the dlPFC was shown to be crucial [31] . As the prefrontal regions mature over the first years of life , the concept of fairness and theory of mind develop across the same age . The above interpretation predicts that children will not act as adults . Indeed , Takagishi et al . [32] demonstrated that preschoolers do reject unfair offers in spite of having no explicit account of unfairness or theory of mind . Thus , these findings suggest that dlPFC might be sufficient but not necessary for rejection , and support a two-level model for decision making . Our data provide some additional insights into the role of the prefrontal regions in decision making . We observed a relative increase in rostral ACC/ventromedial PFC for unfair versus fair proposals in the unmedicated versus oxazepam treatment group , but no changes in the dlPFC . We suggest that this treatment-related change in rostral ACC/ventromedial PFC is secondary to reduced amygdala input mirroring a reduction of conflict [33] . However , it is not possible to exclude that the rostral ACC/vetromedial PFC are part of the attentional control system and therefore could be directly modulated by the treatment . We have shown that the basis for decision making in the UG has underpinnings in several brain regions of different phylogenetic origin , and this underlines the complexity of responses in the UG . Our data suggest that the automaticity driven rejection response has a phylogenetically older representation than the calculated acceptance based on a consciously determined self-optimizing strategy . The amygdala-driven reactive aggressive response generates a behavior that , for example , yields an acceptable splitting of a prey within the group , and such an inequity aversion is seen in children [32] . Thus , automatic individual reactions to detected unfairness seem , to a certain extent , to support the long-term group norms that allow sharing . More developed sharing schemes like formal trade and abstract rule obedience require that each individual can maintain concepts of future effects of present decisions [29] , [31] . Such social interactions rest on the development of the human frontal lobe function . We have demonstrated that an anxiolytic drug alters the balance between rapid emotional reactions and reflected-feeling-based decisions . The finding prompts an ethical discussion , as we showed that a commonly used drug influences core functions in the human brain that underlie individual autonomy and economic decision making .
Thirty-five right-handed volunteers with the mean age of 23 . 7±4 . 2 y ( 13 men , 22 women ) were included in the study . Subjects were randomly assigned to either group independent of gender ( five males and 12 females in the placebo group and eight males and ten females in the oxazepam group ) and had no prior or present history of psychiatric illness or neurological disease . All subjects were healthy and took no medications , with the exception of birth control pills and mild allergy medications . All participants gave their informed consent . The study was approved by the local governmental ethics committee in Stockholm , Sweden . Each subject was exposed to 45 different movie clips . In each movie there was a different human proposer who made a fair , unfair , or neutral suggestion ( see below ) on how to split a sum of money . The fair proposals implied an equal split of the money; the proposer said , for example , “You get 50 Swedish crowns , and I take 50 Swedish crowns” ( 7 Swedish crowns [SEK]≈US$1 ) . The unfair proposals implied that the responder should receive 20% and the proposer 80% of the money , for example , “You get 20 Swedish crowns , and I take 80 Swedish crowns . ” The total stakes ( e . g . , 100 SEK ) were deliberately never mentioned , to maintain ambiguity of fairness until the final proposition of the amount that would be awarded the responder was revealed . All proposals had the exact same wording , except for the monetary amounts , since the total stakes varied . Subjects were instructed to respond with either “yes” or “no” to the proposals by pressing a button . In the neutral control condition the subjects were shown films with proposers saying “this is not a proposal , ” and subjects were instructed to respond “no” to these . The three different stake levels yielded a total of seven different kinds of messages . Each subject encountered six 50/50 offers , seven 20/80 offers , five 125/125 offers , five 50/200 offers , four 250/250 offers , three 100/400 offers , and 15 neutral messages . The genders of the proposers were thoroughly balanced ( 22 males , 23 females ) . Before each movie clip the subject was presented with a resting frame containing a hair cross , for a duration that was randomized between 1 and 5 s . Thereafter , a film clip with an offer was presented . The onset times when the proposer finished the sentence were included as regressors of interest ( individual regressors for fair , unfair , and no proposals ) in the subsequent general linear model analysis of the fMRI analysis . Each movie lasted for 7 s . The clip was followed by a pause , which was again randomized between 1 and 5 s . Thereafter , a frame was shown saying “respond now , ” instructing the subject to make a choice . This frame lasted until a choice had been made or maximally 3 s . The onset times when the subject pressed the button for “yes” or “no” were included as a covariate of no interest in the subsequent general linear model analysis of the fMRI analysis . Finally , a frame confirmed the decision , with a sign stating “you got X Swedish crowns , your counterpart got Y Swedish crowns” ( 2 s ) . A total of 92 individuals were filmed while they read each of the messages as explained above . All individuals were filmed under the same conditions , with light from the front , against a white background , and with the eyes located in the middle of the screen while speaking into the camera . The films with 45 of these individuals were kept , and the others were discarded because of low sound quality or because the person did not look into the camera as desired . Subjects acting as responders were paid 300 SEK for showing up . In addition , three of the 45 movies presented to them were selected at random , and paid out with real money , both for themselves and for the proposer . If the subject had answered “yes” to such a selected proposal , both participants were subsequently paid the corresponding amounts of money . In contrast , if the subject had declined the proposal then neither of the two received any money from that film . This information was given to the subjects before the experiment . The persons acting as proposers on the film clips were given 100 SEK for making the films . They were also subsequently paid their part of the proposals that were drawn randomly , as described for the proposers . Average payment to proposers was 380 SEK , and average payment to responders was 625 SEK . Upon arrival , subjects were randomly assigned to either the placebo group ( five men , 12 women ) or the oxazepam group ( eight men , ten women ) . The subjects in the oxazepam group received 20 mg of the drug . Both treatment and placebo were administered orally in a single-blind fashion . The subjects had been asked in advance not to eat for 2 h before the experiment or drink alcohol for 24 h prior to the experiment . After drug administration , the subjects were asked to fill out two questionnaires: Swedish Universities Scales of Personality [34] and State Trait Anxiety Index–Trait [35] . Before a subject entered the MRI scanner , the rules of the UG were explained , and the subject's understanding of the game was checked with a questionnaire . All subjects passed this test . Approximately 1 h after treatment , the first experimental session was conducted . The order in which the film clips were presented was randomized in advance , creating 18 different sequences of clips , or protocols . Each protocol , except for one , was presented for one subject receiving treatment and for one subject in the control group . We used an fMRI-compatible glove answering device in the scanner to register the subjects' responses . Subjects responded “yes” by pressing a button with their thumb and “no” by pressing a button with their index finger . All subjects underwent two scanning sessions , with a pause of approximately 1 min in between . The first session contained 23 movies , and the second session contained 22 movies . After the scanning was completed , subjects were given a questionnaire and asked to rate the fairness of all the kinds of offers they had received , on a scale 1–7 [7] . They also viewed pictures of the proposers and rated the likeability of the faces they had seen , on a visual analog scale ( 0–100 ) . Only corrected results are reported in this article , with the one exception of the following contrast: placebo unfair proposals>fair proposals . All results are reported as voxel level corrected , unless otherwise stated ( i . e . , cluster level corrected ) .
|
It is well-established that emotions influence decision making . One way of studying this relationship is the Ultimatum Game , which has revealed that subjects punish unfair behavior in others in spite of receiving a concomitant economic loss . Previous brain imaging studies have suggested that this decision to punish involves complex cortical processing . However , punishment also involves an instant aggressive emotional response , a behavior often linked to subcortical structures such as the amygdala . In this study , we present a model that joins these views . We designed a paradigm that allows us to measure the activity of subcortical brain regions during decision making in the Ultimatum Game , while at the same time using a pharmacological approach that can suppress emotional responses and amygdala activity . The pharmacological treatment made subjects punish unfair behavior less , and decreased brain activity in the amygdala in response to unfair proposals , without changing the subjects' feeling of unfairness . In the control group , punishment was directly linked to an increase in amygdala activity . Thus , immediate punishment of unfair behavior involves the amygdala and is not solely driven by cortical processes , as previously suggested . Our results show that a commonly used drug influences autonomy and decision making , which may have ethical implications for its use .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cognitive",
"neuroscience",
"behavioral",
"neuroscience",
"social",
"and",
"behavioral",
"sciences",
"psychology",
"emotions",
"cognition",
"decision",
"making",
"behavior",
"biology",
"neuroscience"
] |
2011
|
Limbic Justice—Amygdala Involvement in Immediate Rejection in the Ultimatum Game
|
Burkholderia pseudomallei is a soil-dwelling bacterium and the causative agent of melioidosis . The global burden and distribution of melioidosis is poorly understood , including in the Caribbean . B . pseudomallei was previously isolated from humans and soil in eastern Puerto Rico but the abundance and distribution of B . pseudomallei in Puerto Rico as a whole has not been thoroughly investigated . We collected 600 environmental samples ( 500 soil and 100 water ) from 60 sites around Puerto Rico . We identified B . pseudomallei by isolating it via culturing and/or using PCR to detect its DNA within complex DNA extracts . Only three adjacent soil samples from one site were positive for B . pseudomallei with PCR; we obtained 55 isolates from two of these samples . The 55 B . pseudomallei isolates exhibited fine-scale variation in the core genome and contained four novel genomic islands . Phylogenetic analyses grouped Puerto Rico B . pseudomallei isolates into a monophyletic clade containing other Caribbean isolates , which was nested inside a larger clade containing all isolates from Central/South America . Other Burkholderia species were commonly observed in Puerto Rico; we cultured 129 isolates from multiple soil and water samples collected at numerous sites around Puerto Rico , including representatives of B . anthina , B . cenocepacia , B . cepacia , B . contaminans , B . glumae , B . seminalis , B . stagnalis , B . ubonensis , and several unidentified novel Burkholderia spp . B . pseudomallei was only detected in three soil samples collected at one site in north central Puerto Rico with only two of those samples yielding isolates . All previous human and environmental B . pseudomallei isolates were obtained from eastern Puerto Rico . These findings suggest B . pseudomallei is ecologically established and widely dispersed in the environment in Puerto Rico but rare . Phylogeographic patterns suggest the source of B . pseudomallei populations in Puerto Rico and elsewhere in the Caribbean may have been Central or South America .
The Burkholderia genus is a group of diverse , primarily soil-dwelling , Gram-negative bacteria that have many strategies to survive and persist in soil , including acid tolerance [1] and intrinsic antibiotic resistance [2] . These species employ a wide variety of ecological strategies , including degradation of common pollutants , mutualistic relationships with plants , and also pathogenic relationships with plants , humans , and/or animals ( 3–10 ) . The taxonomy of this genus remains incomplete and new species are regularly described [3–5] . The genus is commonly separated into two major phylogenetic groups: the B . pseudomallei complex ( BPC ) , consisting of B . pseudomallei and its most closely related phylogenetic relatives , and the B . cepacia complex ( BCC ) [6] . The BCC includes a number of species that can be opportunistic pathogens of immunocompromised individuals , especially cystic fibrosis ( CF ) patients [7 , 8] . Some other Burkholderia species are not assigned to either of these complexes , including the important plant pathogens B . glumae and B . gladioli; B . glumae causes bacterial panicle blight , a devastating disease in rice plants [9] . B . pseudomallei is the causative agent of the disease melioidosis and considered a Tier 1 Select Agent by the US Centers for Disease Control and Prevention ( CDC ) [2 , 10] . Melioidosis can be contracted via cutaneous inoculation , inhalation , or ingestion , and can present with extremely varied symptoms [11]; these vague symptoms and diverse clinical presentations , along with culture-based diagnostic anomalies , make it difficult to properly diagnose in clinical settings [2 , 12] . No vaccines against B . pseudomallei are currently available , making rapid detection and specific antibiotic treatment crucial for favorable outcomes in infected humans . Successful antibiotic treatment typically includes a strict and long regimen of intravenous antibiotics , such as ceftazidime or meropenem , for at least two weeks , followed by oral antibiotics , such as co-trimoxazole , for up to six months [2 , 10] . However , treatment can be complicated by the fact that B . pseudomallei is intrinsically resistant to several clinically relevant antibiotics [13] . Importantly , other Burkholderia species that co-exist with B . pseudomallei in the environment are known to have an intrinsic resistance to other clinically relevant antibiotics , such as meropenem resistance in B . ubonensis [14] , and thereby represent a possible source of similar resistance in B . pseudomallei . B . pseudomallei has an “open genome” that can readily incorporate new genomic content via lateral gene transfer [15] . As a result , it has a relatively large accessory genome ( i . e . the genomic features variable present among different B . pseudomallei strains ) and a relatively small core genome ( i . e . the genomic features present in all B . pseudomallei strains ) . The core genome is currently estimated at ~1 , 600 genes but will likely continue to decrease due to a process known as core genome decay , just as the accessory genome will continue to increase [6] . This is because , as additional B . pseudomallei genomes are generated from novel isolates , components previously identified as part of the core genome will be missing in some of the new genomes and completely novel components will also be identified , both of which increase the size of the accessory genome [6] . Genomic islands , often associated with tRNA sequences [16] , contribute much of the genomic diversity observed in the B . pseudomallei accessory genome and some are hypothesized to contain virulence components [16 , 17] . The adaptive potential of the large accessory genome in B . pseudomallei may be substantial , but remains poorly understood . Determining where B . pseudomallei is present in the environment is crucial for understanding the potential risk to humans of acquiring melioidosis . This is because almost all infections with B . pseudomallei are independently acquired from the environment ( 27 ) ; human to human transmission of melioidosis is extremely rare [18] . B . pseudomallei has long been known to be endemic in tropical regions in northern Australia and Southeast Asia but the true global distribution appears to be much larger . Because melioidosis can be difficult to diagnose , it is possible that B . pseudomallei is also present in the environment in other regions of the world and causing human disease in these areas but going undetected [19] . The majority of Puerto Rico experience a tropical rainforest climate ( based on the Köppen climate classification ) , which is commonly associated with the presence of B . pseudomallei and human melioidosis cases have been previously reported from the island . Since 1982 , there have been a total of seven reported human cases from Puerto Rico , all from the more populated eastern portion of the island ( Fig 1 ) [20 , 21] . A recent human melioidosis case from Puerto Rico , in 2012 , occurred in the southeast municipality of Maunabo . Subsequent soil sampling in this region in 2013 resulted , for the first time , in the isolation of B . pseudomallei from the environment in Puerto Rico [20] . These previous human melioidosis cases ( all but one with no travel history ) and the isolation of B . pseudomallei from soil indicated that B . pseudomallei was present in the environment in Puerto Rico . The primary goal of this study was to gain a better understanding of the prevalence and geographic distribution of B . pseudomallei and other Burkholderia spp . in the environment in Puerto Rico . To achieve this goal , we conducted widespread soil and water sampling around the island and analyzed the samples using PCR and culture-based approaches to identify the presence of B . pseudomallei and other Burkholderia species .
Methods for environmental sampling were based upon international consensus guidelines for sampling for B . pseudomallei in the environment [22] , with additional modifications developed by the Menzies School of Health Research in Darwin , Australia [23] .
Our broad environmental survey in April 2017 resulted in the identification of B . pseudomallei from soil samples collected at only one new location , in the northern municipality of Arecibo ( Fig 1 ) . Just three of the DNA extracts obtained from the 600 enriched Ashdown’s broth samples contained B . pseudomallei DNA; B . pseudomallei was not detected in any of the 100 water samples . The positive samples originated from three adjacent soil samples collected from a single sampling site ( site 23; Fig 1 ) . Site 23 was located on a farm in the municipality of Arecibo where swine , goats , chickens , and cattle were present . DNA extractions from three Ashdown’s broth samples ( 23–07 , 23–08 , and 23–09 ) tested positive with the B . pseudomallei-specific TTS1 PCR assay ( run in triplicate ) . It is important to note that the presence of B . pseudomallei DNA in these complex DNA samples did not definitively indicate that live B . pseudomallei was present in the enriched broth samples or the original soil samples . However , this information allowed us to refocus our culturing efforts on these samples to attempt to isolate B . pseudomallei . Two of the three soil samples did yield B . pseudomallei cultures: B . pseudomallei isolate Bp9039 was obtained from soil sample 23–07 on the first culturing attempt , and a second round of culturing from another 20g of soil from sample 23–09 yielded B . pseudomallei isolate Bp9110 . Additional rounds of culturing yielded other isolates from 23–07 and 23–09 ( S3 Table ) but no B . pseudomallei isolates were ever obtained from soil sample 23–08 despite a positive B . pseudomallei DNA signal from the Ashdown’s broth extraction and multiple attempts at culturing; this is not an uncommon occurrence when surveying for B . pseudomallei in the environment [22] . B . pseudomallei isolates Bp9039 and Bp9110 were both susceptible to meropenem; other collected B . pseudomallei isolates were not tested . The pH of the soil samples collected around the island varied greatly from highly acidic to highly alkaline , whereas the water samples varied from a neutral pH to a highly alkaline pH ( S4 Table ) . All soil samples across all sites had an average pH of 7 . 3 with a range of 3 . 2–11 and all water samples from all sites had an average pH of 7 . 8 with a range of 6 . 8–10 . 2 . The three B . pseudomallei-positive soil samples from site 23 ( 07 , 08 , and 09 ) yielded pH values of 4 . 9 , 4 . 9 , and 5 . 1 , respectively . The pH of soil samples 01–06 from site 23 were 7 . 6 , 7 . 2 , 7 . 0 , 5 . 1 , 6 . 2 , and 6 . 8 , respectively; the pH of soil sample 10 from site 23 was 4 . 9 . No association between soil pH and the occurrence of B . pseudomallei was detected in this study , which was not unexpected given the very small number of B . pseudomallei-positive soil samples ( n = 3 ) . Within a core genome phylogeny of 414 globally diverse B . pseudomallei isolates ( Fig 2 , S2 Table ) , two B . pseudomallei isolates from the municipality of Arecibo ( Bp9039 and Bp9110 ) are highly similar to each other ( both isolates were assigned to MLST ST297 ) and are nested within a large monophyletic group that contains all included B . pseudomallei isolates from other locations in the Caribbean , Central and South America , Mexico , and Africa ( Fig 2 , panel B ) . Within this larger group , the new isolates from Puerto Rico and the previous B . pseudomallei isolates obtained from Puerto Rico form a distinct subgroup together with one isolate from Trinidad . Within that subgroup are two distinct lineages: one including the new environmental isolates from Arecibo with some previous clinical isolates from Puerto Rico , and a second lineage including the previous environmental isolates collected near the 2012 clinical isolate from Maunabo , as well as the one 2012 clinical isolate from Trinidad . The root-to-tip regression analysis identified weak clocklike behavior among the Puerto Rico and Martinique sample set with an R2 value of 0 . 1201 . However , the positive regression slope indicates molecular clock analysis is still reliable for mutation rate estimation [70] . The best-fitting nucleotide substitution model implemented based on MEGA7 model testing was GTR . The 10 , 000-date randomization permutation testing produced a p-value of 0 . 184 , suggesting that the R2 value produced in the root-to-tip regression analysis was not statistically different than random chance . Stepping-stone and path-sampling analyses did not show marked differences; a relaxed clock and extended Bayesian skyline plot was selected as the model combination for this analysis . The BEAST timing analysis had a mean estimate of the year 1950 ( 95% HPD , 1923 to 1975; S1 Fig ) for the TMRCA of chromosome 1 for the eight B . pseudomallei isolates from Puerto Rico and Trinidad . The evolutionary rate was estimated at 5 . 01E-6 ( 95% HPD , 2 . 81E-6 to 8 . 28E-6 ) for all eight samples and the Martinique outgroup . This is in contrast to another study that found an evolutionary rate of 1 . 80E-6 ( 95% HPD , 1 . 36E-6 to 2 . 66E-6 ) for chromosome 1 for multiple B . pseudomallei isolates from the Americas [71] . We observed fine-scale genomic diversity among multiple B . pseudomallei isolates obtained from a single sampling site and even from a single soil sample . A total of 55 B . pseudomallei isolates were isolated from two soil samples at site 23 , with 50 isolates from soil sample 23–07 and five isolates from soil sample 23–09 ( S3 Table ) . It is important to note that these isolates were obtained from enriched culture medium so it is possible that less than 55 individual B . pseudomallei cells were present in the original samples . All 55 isolates were similar in regards to being assigned to the same ST ( 297 ) and to ITS type G; they all also contained the YLF gene cassette . However , variation was still observed among these strains in the core genome phylogeny ( 48 unique SNP genotypes were identified among the 55 isolates ) . There are three distinct clades ( A-C ) observed in the core genome phylogeny for these isolates , with isolates from soil sample 07 ( n = 50 ) assigning to all three clades , while all isolates from soil sample 09 ( n = 5 ) assigned to just clade A along with two of the isolates from soil sample 07 ( S2 Fig ) . Interestingly , there were two B . pseudomallei isolates from different soil samples ( Bp9046-sample07 and Bp9110-sample09 ) that were very similar: these two strains exhibit no SNP differences in the core genome ( S2 Fig ) . We identified four distinct genomic islands ( GI1-GI4 ) among the 55 B . pseudomallei isolates obtained from site 23 ( Fig 3 ) , and these contain a subset of genes not found in any other B . pseudomallei genomes . The insertion of these genomic islands appears to be associated with tRNA gene loci ( S5 Table ) , which is similar to previous patterns described from B . pseudomallei [16] . GI1 ( comprised of 61 genes; S5 Table ) is conserved across all 55 of the B . pseudomallei isolates from site 23 , whereas GI2 ( comprised of 15 genes; S5 Table ) , GI3 ( comprised of 29 genes; S5 Table ) , and GI4 ( comprised of 5 genes; S5 Table ) are variably present among the 55 isolates from site 23 ( Fig 3B ) . The accessory genome of the 55 B . pseudomallei isolates from site 23 is comprised of 58 genes: 49 ( 1–49; S2 Fig , S5 Table ) are contained in GI2 , GI3 , and GI4 and nine others occur at different genomic locations ( 50–58; S2 Fig , S5 Table ) . GI2 is conserved among clade A isolates but not found in the other two clades , GI3 is conserved among clade B isolates but not found in the other two clades , and GI4 is conserved among clade B isolates and variably present in clade A and clade C isolates ( S2 Fig ) . None of the four genomic islands were found in a complete form in 412 other globally diverse B . pseudomallei genomes that were examined , including the genomes of B . pseudomallei isolates obtained from other locations in Puerto Rico ( Fig 3B , S5 Table ) , nor in the genomes of 781 other Burkholderia spp . isolates ( S5 Table ) . A majority ( n = 44 ) of the 61 genes within GI1 were found in at least one of the genomes of the 1 , 193 other B . pseudomallei and/or Burkholderia spp . isolates that were examined , but the other 17 genes in GI1 were only found in the site 23 isolates ( S5 Table ) ; none of the genes in GI2 were found in these other genomes ( S5 Table ) . A majority of the genes within GI3 ( 25/29 ) and GI4 ( 4/5 ) were not found in any of the other 1 , 193 genomes , but all of the nine accessory genes that occurred outside of the genomic islands were found in other B . pseudomallei genomes and some were also found in the genomes of other Burkholderia spp . ( S5 Table ) . Interestingly , it was more common for genes from GI1 and GI3 to be found in the genomes of other Burkholderia spp . than in the genomes of other , global B . pseudomallei isolates ( S5 Table ) . A number of other Burkholderia species are widespread and common in both soil and water throughout Puerto Rico ( S4 Table ) . A total of 686 sub-cultures were selected from Ashdown’s agar plates from 301 soil samples ( collected from all 50 soil collection sites ) and 77 water samples ( collected from all 10 water collection sites ) . Of these sub-cultures , 129 were identified as members of the Burkholderia genus according to the sequence of a recA gene fragment . Most of the 129 Burkholderia isolates ( n = 104 ) were isolated from 61 different soil samples ( originating from 20 of the 50 soil sampling sites ) , with only 25 isolated from 22 different water samples ( but originating from seven of the 10 water sampling sites ) . Burkholderia spp . were cultured from the environment in 21 of the 41 sampled municipalities within Puerto Rico . It is important to note that this does not indicate that there were not Burkholderia spp . present in the environmental samples collected at the other 20 municipalities , only that we did not successfully culture any Burkholderia spp . from environmental samples collected from those locations using our methods . The 129 Burkholderia spp . isolates were identified in a whole genome phylogeny ( Fig 4 ) as follows: B . anthina ( n = 2 ) , B . cenocepacia ( n = 29 ) , B . cepacia ( n = 15 ) , B . contaminans ( n = 5 ) , B . glumae ( n = 1 ) , B . seminalis ( n = 2 ) , B . stagnalis ( n = 36 ) , B . ubonensis ( n = 11 ) , and other unidentified novel Burkholderia spp . ( n = 28 ) ( S1 Table ) . A total of 332 novel MLST alleles were identified from the 129 isolates , resulting in 102 novel STs ( S1 Table ) . All Burkholderia isolates cultured from this study belong to the B . cepacia complex with the exception of B . glumae , which is genetically distinct from both the BPC and BCC . The single B . glumae isolate was identified from a water sample collected in Patillas , Puerto Rico . B . ubonensis appears widespread throughout Puerto Rico , with 11 isolates obtained from five municipalities spread across the island [Barceloneta ( n = 1 ) , Cabo Rojo ( n = 4 ) , Ceiba ( n = 4 ) , Juncos ( n = 1 ) , and Maunabo ( n = 1 ) ; S4 Table] . All 11 B . ubonensis isolates were resistant to meropenem ( >32μg/mL ) ( S1 Table ) .
Widespread environmental surveys for B . pseudomallei in the environment in Puerto Rico identified the pathogen from soil samples collected in a region of Puerto Rico from which it had never been previously detected in the environment or in humans . This study demonstrates that although B . pseudomallei is present in the environment in several widespread locations in Puerto Rico , it is also rare . Given how rare it is in the environment , B . pseudomallei does not appear to pose a large public health risk in Puerto Rico . There have been no known human melioidosis cases reported from the specific location in Arecibo where we detected it in the environment , or even that general region of the island . However , B . pseudomallei is clearly ecologically established in Puerto Rico and , as previously suggested [20] , both the public and clinicians in Puerto Rico should be made more aware of it . Of note , all but one of the previous human melioidosis cases in Puerto Rico occurred in immunocompromised individuals [20] . The International Diabetes Federation stated there were over 400 , 600 cases of diabetes in Puerto Rico in 2017 with a total prevalence of diabetes in adults of 15 . 4% [84] . As diabetes is an important risk factor for melioidosis , clinicians should particularly be aware of the possibility of melioidosis in these individuals .
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention .
|
The objective of this study was to examine the distribution and abundance of Burkholderia pseudomallei in the environment in Puerto Rico . B . pseudomallei is a microbe that lives in soil and causes the disease melioidosis . We conducted sampling around Puerto Rico to survey for the presence of B . pseudomallei in the environment . Of the 600 environmental samples collected , we isolated live B . pseudomallei from just two soil samples collected from the same site , which was in a region of the island where B . pseudomallei had never been previously reported . These results suggest B . pseudomallei is widely dispersed but rare in the environment in Puerto Rico . B . pseudomallei isolates from Puerto Rico are most closely related to other strains from the Caribbean . Caribbean strains are inside a larger group that contained all analyzed isolates from Central/South America , suggesting that B . pseudomallei populations in the Caribbean may have been introduced from Central or South America .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Disclaimer"
] |
[
"taxonomy",
"medicine",
"and",
"health",
"sciences",
"burkholderia",
"pseudomallei",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"melioidosis",
"geographical",
"locations",
"microbiology",
"north",
"america",
"bacterial",
"diseases",
"burkholderia",
"cepacia",
"complex",
"phylogenetics",
"data",
"management",
"caribbean",
"bacteria",
"extraction",
"techniques",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"genomics",
"medical",
"microbiology",
"microbial",
"pathogens",
"comparative",
"genomics",
"evolutionary",
"systematics",
"people",
"and",
"places",
"puerto",
"rico",
"dna",
"extraction",
"genetics",
"biology",
"and",
"life",
"sciences",
"burkholderia",
"computational",
"biology",
"evolutionary",
"biology",
"organisms"
] |
2019
|
Burkholderia pseudomallei, the causative agent of melioidosis, is rare but ecologically established and widely dispersed in the environment in Puerto Rico
|
Cells of the embryonic vertebrate limb in high-density culture undergo chondrogenic pattern formation , which results in the production of regularly spaced “islands” of cartilage similar to the cartilage primordia of the developing limb skeleton . The first step in this process , in vitro and in vivo , is the generation of “cell condensations , ” in which the precartilage cells become more tightly packed at the sites at which cartilage will form . In this paper we describe a discrete , stochastic model for the behavior of limb bud precartilage mesenchymal cells in vitro . The model uses a biologically motivated reaction–diffusion process and cell-matrix adhesion ( haptotaxis ) as the bases of chondrogenic pattern formation , whereby the biochemically distinct condensing cells , as well as the size , number , and arrangement of the multicellular condensations , are generated in a self-organizing fashion . Improving on an earlier lattice-gas representation of the same process , it is multiscale ( i . e . , cell and molecular dynamics occur on distinct scales ) , and the cells are represented as spatially extended objects that can change their shape . The authors calibrate the model using experimental data and study sensitivity to changes in key parameters . The simulations have disclosed two distinct dynamic regimes for pattern self-organization involving transient or stationary inductive patterns of morphogens . The authors discuss these modes of pattern formation in relation to available experimental evidence for the in vitro system , as well as their implications for understanding limb skeletal patterning during embryonic development .
Skeletal pattern formation in the developing vertebrate limb depends on interactions of precartilage mesenchymal cells with factors that control the spatiotemporal differentiation of cartilage . The most fundamental skeletogenic processes involve the spatial separation of precartilage mesenchyme into chondrogenic and nonchondrogenic domains [1] , and can be studied in vitro as well as in vivo ( Figure 1 ) . In high-density “micromass” cultures of chondrogenic ( i . e . , cartilage-forming ) embryonic limb mesenchymal cells [2 , 3] , as well as in the developing limb itself [4] , morphogens of the TGF-β family induce the local aggregation or condensation of these cells by a process that involves the upregulation of the adhesive extracellular glycoprotein fibronectin [3 , 5] . Cells first accumulate in regions of increased cell–fibronectin adhesive interactions [6–8] and then acquire epithelioid properties by upregulation of cell–cell adhesion molecules [9 , 10] . Cartilage differentiation follows at the sites of condensation both in vitro and in vivo ( see [11–13] for reviews ) . In certain developmental processes , such as angiogenesis ( sprouting of capillaries ) and invasion by cancer cells of surrounding tissues , pre-existing multicellular structures become more elaborate . Precartilage condensation , by contrast , is an example of a developmental process in which cells that start out as independent entities interact to form multicellular structures . Others in this second category include vasculogenesis ( the initial formation of blood vessels ) , the formation of feather germs , and the aggregation of social amoebae into streams and fruiting bodies . Both continuous [14–18] and discrete [18–29] models have been used previously to analyze a wide range of pattern formation behaviors in both categories using concepts such as chemotaxis , haptotaxis , and reaction–diffusion instability . Discrete models describe the behaviors and interactions of individual biological entities such as organisms , cells , proteins , etc . They are often applied to microscale events where a small number of elements can have a large ( and stochastic ) impact on a system . In a previous study [26] we presented a discrete “biological lattice gas” model for high-density cultures of precartilage mesenchymal cells derived from the embryonic vertebrate limb . This model , which was based on the physical notion of a lattice gas , in which individual particles are free to move from point to point on a lattice at discrete time-steps , accurately simulated the formation of patterns of mesenchymal condensations observed in high-density micromass cultures of such cells . In these simulations , the distribution and relative size of the condensations corresponded to in vitro values when appropriate quantities for cell behavioral parameters were chosen , and the simulated patterns were robust against small variations of these values . Moreover , the simulated patterns were altered similarly to the cultures when cell density and exposure to or expression of molecular factors represented in the model were altered in a fashion analogous to their counterparts in the living system . In the earlier model , each of the limb precartilage mesenchymal cells , and each molecule from a “core” subset of the molecules they secrete ( the diffusible activator morphogen TGF-β , a diffusible inhibitor of TGF-β's effects , the extracellular matrix [ECM] protein fibronectin ) , was represented as a single particle ( pixel ) on a common grid . Default motion of the cell particles was random , but cell movement was also biased by the presence of fibronectin particles produced and deposited by the cells according to a set of rules involving TGF-β and inhibitor particles . The latter in turn were produced in a cell-dependent fashion according to a reaction–diffusion scheme , the network structure of which was suggested by in vitro experiments [2 , 3 , 30] . The ability of the model of Kiskowski et al . [26] to simulate both qualitative and quantitative aspects of precartilage condensation formation and distribution suggested that the core genetic network–cell behavioral mechanism that underlies this biological lattice gas might be sufficient to account for pattern formation in the limb cell micromass system and corresponding features of in vivo limb development . However , the model deviated from biological reality in several important ways . ( 1 ) Mesenchymal cells in vitro are initially surrounded by a small layer of ECM that separates them by less than a cell diameter . Those that undergo condensation round up , reducing their surface area , but do not move away from adjacent noncondensing cells . Therefore , unlike the situation in the model of Kiskowski et al . [26] , mesenchymal condensation in micromass culture does not involve accumulation of cells at particular sites with concomitant depletion of cells in surrounding zones . ( 2 ) The representation of cells , morphogens , and ECM on a common grid is physically unrealistic . This is not simply a matter of pixel scale: molecular substances can indeed form deposits and gradients on the same linear scale as cells ( ∼10 μm ) , and a “molecular” pixel could be considered to correspond to thousands of molecules . Nonetheless , the dynamics of morphogen transport is continuous and is represented in an inauthentically saltatory fashion by pixel displacement on a grid of the same mesh size as that supporting cell translocation . ( 3 ) Whereas the model of Kiskowski et al . made the assumption that cells halt their motion when they encounter suprathreshold levels of extracellular fibronectin [26] , this does not agree with measurements [31 , 32] indicating that cells actually slightly increase their speed of motion as they enter condensation centers and have a finite probability of escaping from these foci . Despite the successes of the model of Kiskowski et al . [26] , it was unknown whether removing its artifactual aspects and replacing them with more realistic assumptions would lead to similarly authentic results . We have therefore designed a more sophisticated model that overcomes each of the listed deficiencies of the earlier one . The cells in the new model are extended , multipixel objects that can change shape in the plane and “round up” by moving pixels into a virtual third dimension . The model cells are separated by less than a cell diameter , condense without denuding the regions surrounding condensation centers , and are not irreversibly trapped upon entering a center . Finally , two grids of different mesh size are used for cell and molecular dynamics . We have found that not only does this improved model reproduce the experimental data accounted for by the model of Kiskowski et al . , but that additional morphogenetic features of the micromass culture system are simulated as well . Moreover , potential dynamic properties of the developmental process not seen in the earlier simulations , and not capable of being distinguished on the basis of existing experimental data , were disclosed in simulations using the new model , which has therefore provided motivation for further empirical tests .
We represented each model cell as an extended object on a 2-D spatial grid . The rate and probability at which cells move ( by random walk ) and change shape are parameterized separately from movement of molecules so that they can be calibrated to the scale of actual biological cells . Each model cell behaves according to a predefined set of experimentally motivated rules involving morphogen dynamics controlling the production and deposition of fibronectin ( see Materials and Methods ) . We chose the simplest multipixel representation of limb mesenchymal cells subject to the following biological constraints: ( 1 ) cells have essentially isotropic geometry ( i . e . , they do not elongate in the direction of migration , but rather probe their environment by extending short randomly oriented projections ) ; ( 2 ) the cell nucleus is also isotropic but is relatively unchanging in shape and comprises on the order of half the cell volume; and ( 3 ) cells in fibronectin-rich , condensing areas of the micromass round up such that their cross-section in the plane of the culture is significantly reduced [32] . Model cells ( initially comprising seven pixels; Figure 2A ) are therefore permitted to change shape consistent with maintaining four pixels in a two-by-two square ( kernel ) configuration that represents the portion of the cell that contains the nucleus ( Figure 2B ) ( although one or more pixels of the central block can exchange with peripheral pixels at each time-step ) . Cells respond to suprathreshold levels of fibronectin by shrinking their area from seven pixels to five pixels ( corresponding to rounding up into a virtual third dimension; Figure 2C ) and increasing the rate at which they move and change shape . Once a cell ventures onto fibronectin it has the tendency to remain there , with a low probability of leaving the condensation . Seven pixels is the smallest multipixel representation that allows for both shape change events and appropriate cross-sectional area change when cells round up while maintaining a multipixel kernel . The relatively small cell-size representation in the model allows us to run extensive simulations with large numbers of cells for a wide range of different parameters . Smaller representations essentially reduce to a particle system; it is straightforward to add more pixels if greater cell shape fidelity is desired . ( See Materials and Methods for implementation of all the above . ) The degrees of freedom built into our model allowed us to calibrate some of the simulation parameters with experimentally determined values obtained in related or analogous systems . In particular , the diffusion rate of the activator morphogen and that of mesenchymal cells correspond well to experimental values , and they both play an important role in the resultant behavior of the model . Lander et al . [33] calculated the effective diffusion coefficient for a molecule the size and shape of Dpp , a morphogen of the same superfamily as TGF-β , to be 10 μm2/s . We have used this value , although the actual value may differ due to varying capacities of members of the superfamily to bind to variable microenvironments [34] . In the present model , the diffusion rate for the activator morphogen in the reaction–diffusion system was found to be a key parameter for determining the size of the resultant patterns ( see below ) . If the diffusion rate is too slow , the activator does not spread out across a sufficiently large area to produce broad condensations; in contrast , if the diffusion rate is too fast , the activator spreads out too much , thus preventing patterns from even forming . The effective “diffusion” rate for cells is considerably slower than that of morphogens , and cells do not move significant distances over the time period of precartilage condensation formation [31 , 35 , 36] . ( Diffusion of cells is of course not due to Brownian motion , as is molecular diffusion , but rather results from randomly directed cell locomotion based on internally generated surface protrusive forces . ) Tracking of cells in time-lapse videos of developing chicken limb precartilage mesenchyme showed that they move with an average diffusion coefficient of ∼0 . 5 μm2/min in noncondensed regions and slightly faster in condensations [32] . We have therefore incorporated these experimental values into our model , increasing the cell diffusion rate by 50% in the presence of a threshold level of fibronectin ( see Materials and Methods ) . In addition , we cause the area of cells situated on fibronectin to shrink in the presence of suprathreshold levels of fibronectin , corresponding to the observed rounding-up of cells in mesenchymal condensations . Using the experimental values for activator and cell diffusion coefficients greatly facilitated choosing other parameters so that appropriately sized and spaced condensations formed in silico . This contrasted with parameter searches performed with nonbiological choices of activator and cell diffusion coefficients . In those cases , no realistic patterns formed in scores of simulations . The inhibitor morphogen , elicited when cells in incipient condensations are exposed to one or more ectodermally produced fibroblast growth factors [30] , must spread at a faster rate than the activator morphogen for stable patterns to be generated according to the reaction–diffusion dynamics . We performed a number of simulations that varied the ratio between the activator diffusion rate , which was kept constant , and the inhibitor diffusion rate . Consistent pattern formation was obtained when the inhibitor diffuses at a rate four to eight times faster than the activator . At a slower than 4-fold ratio , the patterns degraded in consistency until the point where no patterns were produced at all , which occurred when both diffusion rates were almost equal ( Table 1 ) . Beyond the 8-fold ratio , consistent patterns were still produced ( results not shown ) . The relatively small ratio between the two diffusion rates makes the hypothesis of a diffusible inhibitor of condensation formation [30 , 37] biologically plausible . Experimental evidence indicated that limb mesenchymal cells in vitro respond to transient elevation in TGF-β concentration early during the culture period by upregulating fibronectin production for at least a day [2] . We have therefore assumed for the simulations described below that cells are induced to produce and secrete fibronectin by their first suprathreshold exposure to activator and become unresponsive to later exposures to activator . Furthermore , we have assumed that cells that are not exposed to inducing levels of activator during a critical period follow an alternative fibroblastic differentiation pathway [38] , rendering them similarly unresponsive to later exposure to activator . Consistent with the experimental constraints described above , we searched for a parameter set in the model that reproduces the formation of precartilage condensation patterns . We calculated the average interval of the centroids ( “peak interval” ) [39] and the average island size of the fibronectin patches [26] for five simulation images and compared the values with those obtained from 12 in vitro condensation images such as that in Figure 3A . The results ( Figure 4 ) indicate that our enhanced model reproduces the pattern of precartilage condensations equally as well as the model of Kiskowski et al . [26] . Different views of one simulation with parameters chosen within the “standard” range are shown in Figure 3B–3D . The distribution of condensations ( Figure 3D ) conforms very well to the photograph of the 72-h culture ( Figure 3A ) , although the cells in the individual in silico condensations are not tightly packed as they are in the in vitro ones . This is not unexpected , since the model at present lacks representation of a cell–cell adhesion molecule , several of which are upregulated at condensation sites in limb mesenchyme [9 , 10] . The shape change of the model cells once they encounter fibronectin does nonetheless lead to a realistically higher cell density in condensed versus noncondensed regions of the simulated cultures . The simulated distribution of fibronectin ( Figure 3C ) conforms to the distribution of condensations , as expected from immunolocalization studies [5] . The distribution of activator peaks at the time-point shown in Figure 3D maps out the set of eventual condensations . Previous experimental studies show TGF-β localization to anticipate the formation of condensations by up to a day [2] , and to trigger the subsequent production of fibronectin after a brief , transient exposure [2] . The model , with different parameter choices , leads to realistic condensation patterns with either transient ( as in the simulation shown in Figure 3B–3D ) or stable activator patterns ( see below ) . We explored the robustness of the parameter set by varying key parameters independently ( ±5% ) ; results can be seen in Figure 5 . Minor variation of the inhibitor strength on activator ( k2 ) by either +5% or −5% produced little change in the resulting condensation patterns . Instead , the temporal dynamics were modified , causing an increase and decrease in the period of the morphogen oscillations , respectively , with the +5% and −5% changes . For a decrease of 5% in the activator self-regulation ( k1 ) or an increase of 5% in the activator regulation of inhibitor ( k3 ) , smaller condensations were produced with the condensations spaced further apart from one another . For a 5% increase in k1 or a decrease of 5% in k3 , condensation patterns greatly expanded in size such that the condensations touched one another , producing a pattern of interconnected stripes instead of spots . Similar results were also obtained if the inhibitor decay ( k4 ) was increased by 5% . For a 5% decrease in k4 , the chemical reaction was effectively damped , and no patterns were produced . Consistent with observations of limb precartilage development in vitro and in vivo , our simulation results indicate that cells can form condensation patterns by undergoing small displacements of less than a cell diameter , packing more closely by changing their shapes , while maintaining a relatively uniform cell density across the entire spatial domain . Given the possibility that choices of spatial domain and boundary conditions could lead to simulation artifacts , we sampled various alternatives in combination and investigated changes in the resulting condensation patterns . With respect to the spatial domain , we ran simulations with rectangular grids of various widths and heights ( unpublished data ) ; this produced no noticeable effects on the size , shape , or distribution of the condensations . We conclude that the total area of the spatial domain determines only the number of condensations . We also ran simulations with periodic and no-flux conditions . In periodic conditions , grid boundaries are connected together simulating a continuous space , whereas the no-flux boundary acts as a barrier . Both types of boundary conditions produced similar results for the size , shape , and distribution of the condensation patterns apart from the expected pattern truncations under no-flux conditions ( unpublished data ) . Our simulations disclosed two regimes of behavior in the reaction–diffusion system of morphogens ( Figure 6 ) . In one regime , the maximum concentration levels for the two morphogens are characterized by a stationary value; this regime appears when the chemical reaction is slow ( i . e . , the production rate of the activator morphogen is balanced with the production rate of the inhibitor morphogen; Figure 6B ) . In the other regime , the concentrations levels for the two morphogens had an oscillatory behavior; concentrations increase up to a peak value , decrease back down to almost zero , and then continually repeat that cycle ( Figure 6A ) . The oscillatory regime occurs when the chemical reaction is fast but a cap exists for the maximum amount of morphogen produced for a single reaction step . Both regimes for the reaction–diffusion system can produce condensations patterns in the range of experimental values for size and distribution . ( See Figures 3 and 4 for the oscillatory regime and Figures S1 and S2 for the stationary regime ) . The limits on morphogen production ( MAXU and MAXV in Table 2 ) induce the oscillatory regime by restricting production of activator while still allowing production of inhibitor , whose concentration has not yet reached the limit . The result is that inhibitor concentrations build up in the system; the inertia of inhibitor concentration dampens activator production throughout the whole system , which quickly accelerates and reduces the activator concentration down to basal levels . Cells continue to produce a basal amount of activator , so over time conditions are reproduced for the onset of morphogen pattern formation . The dynamics repeat , with transient patterns being formed , though the spatial arrangement of the peaks varies unpredictably from one oscillation to the next . Variations in the limits on morphogen production in the oscillatory regime produced minimal changes in the average peak interval and average island size of the fibronectin patch distribution ( Table 3 ) . The oscillatory regime is more robust for higher limits and breaks down when the concentrations are low . In contrast , the stationary regime operates in the lower concentration levels of the morphogens . The oscillatory regime is robust to a noisy threshold level for cell differentiation . Simulations where each cell's threshold is randomly assigned from a normal distribution , N ( 9 , 000; 1 , 000 ) , instead of a constant value , produce only slight variation in the average peak interval and average island size despite the large deviation in the threshold levels . However , the stationary regime is sensitive to the threshold level for cell differentiation as a modest variation , N ( 2 , 400; 170 ) , completely disrupts the spatial distribution of the fibronectin patches ( unpublished data ) . The formation of patterns in the stationary regime is sensitive to the period that cells are exposed to activator morphogen and to the threshold level for cell differentiation . If the exposure time is too short , small , irregularly spaced condensations are produced . If the exposure is too long , irregularly shaped condensations are produced . Although the stationary regime produces stable activator peaks , those peaks tend to wander spatially over time due to the underlying cell diffusion . The oscillatory regime is less sensitive to the threshold level for cell differentiation , and a single transient pulse provides a well-defined exposure period . While the focus of our model has been on producing the spot patterns typically seen in leg-cell cultures [5 , 40] , the exact same model can produce stripe patterns with a slight adjustment to parameters ( Figure 7 ) . This is significant because uncontrolled variations in the preparation of cultures grown under the same conditions as the spot-producing ones occasionally give rise to stripe patterns ( Figure 7A ) . When the reaction–diffusion system progresses to spot patterns , it goes through a brief period of partial stripe formation until dominant activator peaks stabilize the system into spot patterns . Reducing the limits on morphogen production ( MAXU and MAXV in Table S1 ) prevents peaks of activator morphogen from dominating , and stable stripe patterns are maintained . This corresponds to theoretical analysis by Shoji and coworkers of reaction–diffusion systems with linear kinetics and constant constraints [41]; they show that stripe patterns are generated instead of spot patterns if the upper and lower constraints are equal distances from the equilibrium . Similar to the formation of spot patterns in the stationary regime , the formation of stripe patterns is sensitive to the duration of the period in which cells are exposed to activator morphogen .
We have demonstrated that parameter choices can be found for our quasi–3-D discrete model that reproduce the experimental distribution and size range of precartilage condensations in experimental micromass cultures . The performance of the model was equal to that of Kiskowski et al . [26] , despite the imposition of realistic scaling and experimentally determined constraints . The new model has allowed us to study the interplay between reaction–diffusion processes , fibronectin production , and cell–fibronectin interaction in greater detail than previously possible . In particular , our simulations disclosed two regimes in the interplay of the reaction–diffusion system of morphogens with fibronectin production and cell behavior . In one regime , stationary morphogen patterns were produced , followed by cell rearrangement into patterns of condensation . In the second regime , morphogen patterns were transient and oscillatory in time , and the induced fibronectin production ( and consequent cell rearrangement ) occurred with a delay . In addition , the dynamic characteristics of the second regime provide a natural explanation for apparent oscillatory effects of limb precartilage cell responses to TGF-β seen in previous experimental studies [2] . As mentioned in Results , the transient regime also exhibits less sensitivity than the stationary regime to several key system parameters , giving it plausibility as the more robust pattern-forming mechanism . However , in order to suppress “second-generation” condensation patterns due to the recurrence of activator peaks in this regime , we assumed that cell differentiation to a morphogen-nonresponsive state occurs rapidly relative to the period of oscillation . This assumption is obviously not needed for simulations in the stationary case; indeed , stable pattern formation in this regime would be consistent with extended ( i . e . , over a period of a day or more ) susceptibility to perturbation by exogenous TGF-β . We are currently performing in vitro experiments analogous to earlier studies on the first day of development [2] to test this predicted difference , as well as some others . Our model generates realistic patterns of precartilage condensation in high-density culture without the need to postulate direct cell–cell adhesive interactions . This feature appears to reflect biological reality . First , although the separation of condensing from noncondensing cells superficially resembles sorting out by differential adhesion ( see [42] for a recent model of the latter process based on a free-energy minimization principle ) , haptotactic binding to fibronectin is sufficient to recruit limb precartilage mesenchymal cells , or even inert particles , into condensations [6] . Second , while as mentioned above , several cell–cell adhesive proteins , including N-CAM [9] and N-cadherin [10] , are expressed at sites of condensation , their loss does not impair condensation-dependent skeletogenesis [43 , 44] . We note that in both the oscillatory and stationary cases , the region of parameter space that leads to realistic fibronectin patch and condensation patterns corresponds to activator morphogen peaks that are on the spatial scale of the condensations themselves . For the oscillatory regime , a small number of those peaks ( see Videos S1 and S2 ) have relatively high and possibly nonphysiological activator and inhibitor concentrations ( assuming morphogen units represent one or more protein molecules ) . If morphogen dynamics in these cultures is indeed oscillatory [2] , this may represent an inauthentic aspect of our model , resulting from the use of the classic diffusion-dependent Turing-type morphogen scheme . We are therefore exploring alternative embodiments of the model using juxtacrine signaling , the role of which is suggested by recent demonstration of involvement of the Notch signaling pathway in the inhibitory branch of the condensation-patterning network [45] . Recent analyses have suggested that introducing juxtacrine signaling into the dynamics can bring reaction–diffusion pattern-forming systems that are otherwise biochemically implausible into more realistic parameter domains [46] . We note that our multipixel representation will enable the incorporation of cell asymmetry and polarity ( a known feature of limb mesenchymal cells [47] ) in future models using cell relay mechanisms . The capacity of our model to generate both spots and stripes of precartilage condensation under slightly different parameter choices corresponds well to experimental results in which either morphotype may be generated under similar initial conditions . Because the developing limb itself generates its skeleton in the form of spots and stripes of precartilage condensation ( Figure 1; see also [1] ) , this result of our simulations supports the applicability of the core molecular–genetic mechanism we have used to the understanding of both in vitro and in vivo chondrogenic pattern formation . Moreover , the flexibility and generality of the framework presented here makes it suitable for representing and testing other experimentally motivated models for periodic patterning in which cell movement and shape change is involved , such as the formation of feathers and hair [25 , 48] and teeth [49 , 50] .
Cell cultures were prepared using precartilage mesenchymal tissue isolated from the myoblast-free distal 0 . 3 mm [51] of Hamburger and Hamilton stages 24–25 [52] leg buds of 5-d White Leghorn embryos ( Moyer's Chicks , http://www . moyerschicks . com ) under the conditions described for the standard cultures in Kiskowski et al . [26] ( 1 . 75 × 105 cell per 10-μl spot in serum-free defined medium [53] ) . Living cultures were photographed using Hoffman Modulation Contrast optics ( 4× objective lens; Modulation Optics , Inc . , http://www . modulationoptics . com ) with condenser and polarizer adjusted to visualize cell condensations [7] . The spatial environment that cells and molecules occupy is modeled on a 2-D plane . The implementation provides support for multiple superimposed discrete grids of various spatial scales . In our current model , we use two scales: one for the cellular level and another finer-resolution scale for the molecular level . The coarsest resolution spatial scale is considered to be the base spatial scale , which is the cellular level for our model; all other grids are an integer ratio size of that base grid . The base spatial grid can be defined as a square or rectangular grid of any height and width , and all of the grids overlay one another and cover the same physical area . Each cell is represented as a set of seven contiguous pixels operating on the base spatial grid as shown in Figure 2A . We maintained four pixels in a two-by-two square ( kernel ) configuration that represents the portion of the cell that contains the nucleus and allowed the remaining pixels to occupy the border region around the nucleus . Cells that round up shrink their spatial extent to five pixels ( Figure 2C ) . Cell diffusion was implemented as a random walk . If the cell moves , then all of its seven ( or five ) pixels move one pixel in the appropriate direction ( up , down , left , right ) . Cells can also fluctuate in shape , yet such fluctuations maintain a structural representation of the central region containing the nucleus by preserving intact a two-by-two square block of pixels . Therefore , shape fluctuations are restricted to the motion of the three ( or one ) border pixels around the nucleus which either move to new border pixels or displace nucleus pixels; Figure 2B gives an example of both types of fluctuations for a cell changing shape . Cells are prevented from overlapping each other when they move or change shape . In our discrete representation of the Turing mechanism , a discrete number of activator and inhibitor molecules occupy each pixel on the grid , and each molecule is considered to have a spatial representation of just one pixel . We modeled the reaction dynamics of the activator and inhibitor molecules at each pixel as follows: let Ut and Vt be the concentration of the activator and inhibitor , respectively , at time t , and let ϕt be an indicator function for the existence of a cell ( ϕt = 1 if cell is at the pixel; otherwise , ϕt = 0 ) at time t . Equation 1 shows the change over time for each pixel on the grid of the activator morphogen concentration based upon a proportion ( as defined by chemical reaction rates ) of the current activator and inhibitor concentrations . Equation 2 shows the corresponding change over time for each pixel on the grid of the inhibitor morphogen . The activator morphogen is considered a positive self-regulating molecule and a positive regulator of the inhibitor; thus , chemical rate parameters k1 and k3 both have positive values . The inhibitor morphogen is considered a negative regulator of activator that decays over time; thus , chemical rate parameters k2 and k4 are both negative values . In our model , production of the activator and inhibitor molecules , as represented by the parameters k1 and k3 , can only occur in the presence of a cell as enforced by ϕt; however , the decay of activator and inhibitor , as represented by the parameters k2 and k4 , is considered to occur independent of cell presence . Cells are initially randomly distributed on the grid and secrete a small basal amount ( BU ) of activator morphogen that provides the initial concentration of activator; cells continue this basal production throughout the simulation , and inhibitor concentration starts at zero . In keeping with the biology , we considered cells to respond to low concentrations of morphogens and thus represent morphogen molecules as discrete entities . Consequently , the morphogen concentrations ( Ut , Vt ) are whole numbers , and changes in the concentrations at a time-step are rounded to the nearest integer and prevented from becoming negative . Nonetheless , we treated the chemical rate parameters ( k1 , k2 , k3 , k4 ) for the two morphogens as averages of the reaction rates and allowed them to assume real number values . In any physicochemical reaction there is limitation on how much reagent a single cell can realistically produce during any period of time . For this reason , our model provides separate parameters ( MAXU , MAXV ) for the maximum amount of activator and inhibitor that can be produced during a single reaction step . The maximums are imposed on individual pixels of the molecular grid rather than across the entire cell , consistent with polarization of limb mesenchymal cells [47] . This allows for small morphogen gradients to be present across the spatial extent of an individual cell through the spatially polarized secretion of morphogens . The peaks of activator concentration produced by the reaction–diffusion dynamics define a large-scale prepattern , equal in spatial area to the fibronectin patches , containing around 30 cells within a single patch . Polarization plays a role for the cells on the border region of the patch , whereas cells in the patch interior perceive a relatively constant morphogen concentration across their entire spatial extent . Molecular diffusion from any pixel can occur randomly toward any of the four neighboring pixels ( up , down , left , right ) . The diffusion rates ( DU , DV ) are scaled into a probability factor 0 < p < 1 and a time-step n such that D = pn . The probability determines the chance that a molecule will diffuse , and the time-step indicates how many opportunities a molecule has to diffuse for a single simulation iteration; if the molecule diffuses , then one of the four neighboring pixels is picked with equal probability . The chemical reaction operates at a much slower rate than molecular diffusion , so the time scales are separated with diffusion calculated at a small time-step and the reaction calculated at a longer time-step . Fibronectin is a nondiffusing ECM molecule that forms the template for precartilage condensations . As the concentration levels of the activator morphogen increases in the presence of a cell , that cell produces fibronectin mRNA , which can then be translated into actual fibronectin protein molecules . The model supports a simple threshold level ( CDT ) such that once the sum of activator concentration across the entire spatial area of a cell exceeds that threshold value , the cell differentiates into a fibronectin-producing cell . Because we did not directly describe the level of fibronectin mRNA within the cell , the trigger for cell differentiation is separated from the actual production of fibronectin , and a model parameter defines the delay between cell differentiation and secretion of fibronectin . We assumed that there is a critical period during which exposure , or lack of exposure , of cells to activator morphogen , causes them to either differentiate into fibronectin-producing cells or follow an alternative differentiation pathway . For purposes of simulation , we disabled the reaction–diffusion dynamics after this critical period ( adjustable by a parameter ) and prevented additional cells from differentiating . For the oscillatory regime , a single transient pulse ( Figure 6B ) defines the exposure period; reaction–diffusion dynamics are disabled when the activator morphogen returns to basal concentration levels . For the stationary regime , reaction–diffusion dynamics are disabled after 500 simulation iterations ( see Figures S1 and S2 ) . When a cell produces fibronectin , a single unit representing a multimolecular complex is secreted with random probability for each of the pixels on the molecular grid in the cell's spatial domain , and each unit is allowed to perform an initial small diffusion of at most one pixel [26] . Production of fibronectin units continues until a maximum concentration level is reached at a pixel , although cells may still continue to produce fibronectin on pixels that have not yet reached the maximum . The production rate of fibronectin , the duration of such production , and the maximum amount of fibronectin allowed per pixel can be adjusted with model parameters . In attempting to calibrate our model parameters with known empirical parameters , we wanted to correlate the spatial and temporal patterns produced by computer simulation results with in vivo and in vitro experiments . For spatial patterns , we considered the size , shape , and distribution of the fibronectin-rich spatial domains; for temporal patterns , we considered the reaction rates of activator and inhibitor production , the diffusion rates of both cells and molecules , the onset of fibronectin production , the production rate of fibronectin , and the shape and movement fluctuations of cells on fibronectin . The actual values for the set of key parameters used in the simulation and their corresponding physical measurements , if known , are shown in Table 2 . Whereas the previous model of Kiskowski et al . [26] was written in Matlab ( http://www . mathworks . com ) , we rewrote the current model in the C programming language for efficiency , and then migrated it to the Objective-C programming language to take advantage of object-oriented features . We still used Matlab for visualizing data produced by simulation runs . The original source code is available as Dataset S1 accompanying this article , and can be obtained directly from the authors . We intend to continue developing the software by expanding the capability to add molecular and cellular detail to models of the limb micromass culture system and allied 2-D and quasi–3-D developmental systems .
|
The development of an organism from embryo to adult includes processes of pattern formation that involve the interactions over space and time of independent cells to form multicellular structures . Computational models permit exploration of possible alternative mechanisms that reproduce biological patterns and thereby provide hypotheses for empirical testing . In this article , we describe a biologically motivated discrete stochastic model that shows that the patterns of spots and stripes of tightly packed cells observed in cultures derived from the embryonic vertebrate limb can occur by a mechanism that uses only cell–cell signaling via diffusible molecules ( morphogens ) and cell substratum adhesion ( haptotaxis ) . Moreover , similar-looking patterns can arise both from stable stationary dynamics and unstable transient dynamics of the same underlying core molecular–genetic mechanism . Simulations also show that spot and stripe patterns ( which also correspond to the nodules and bars of the developing limb skeleton in vivo ) are close in parameter space and can be generated in multiple ways with single-parameter variations . An important implication is that some developmental processes do not require a strict progression from one stable dynamic regime to another , but can occur by a succession of transient dynamic regimes tuned ( e . g . , by natural selection ) to achieve a particular morphological outcome .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"chicken",
"in",
"vitro",
"vertebrates",
"computational",
"biology"
] |
2007
|
Patterns of Mesenchymal Condensation in a Multiscale, Discrete Stochastic Model
|
Although a considerable proportion of serum lipids loci identified in European ancestry individuals ( EA ) replicate in African Americans ( AA ) , interethnic differences in the distribution of serum lipids suggest that some genetic determinants differ by ethnicity . We conducted a comprehensive evaluation of five lipid candidate genes to identify variants with ethnicity-specific effects . We sequenced ABCA1 , LCAT , LPL , PON1 , and SERPINE1 in 48 AA individuals with extreme serum lipid concentrations ( high HDLC/low TG or low HDLC/high TG ) . Identified variants were genotyped in the full population-based sample of AA ( n = 1694 ) and tested for an association with serum lipids . rs328 ( LPL ) and correlated variants were associated with higher HDLC and lower TG . Interestingly , a stronger effect was observed on a “European” vs . “African” genetic background at this locus . To investigate this effect , we evaluated the region among West Africans ( WA ) . For TG , the effect size among WA was the same in AA with only African local ancestry ( 2–3% lower TG ) , while the larger association among AA with local European ancestry matched previous reports in EA ( 10% ) . For HDLC , there was no association with rs328 in AA with only African local ancestry or in WA , while the association among AA with European local ancestry was much greater than what has been observed for EA ( 15 vs . ∼5 mg/dl ) , suggesting an interaction with an environmental or genetic factor that differs by ethnicity . Beyond this ancestry effect , the importance of African ancestry-focused , sequence-based work was also highlighted by serum lipid associations of variants that were in higher frequency ( or present only ) among those of African ancestry . By beginning our study with the sequence variation present in AA individuals , investigating local ancestry effects , and seeking replication in WA , we were able to comprehensively evaluate the role of a set of candidate genes in serum lipids in AA .
The role of the distribution of serum lipids in influencing disease risk is well-established . Serum lipids are under the influence of genetic and non-genetic ( e . g . , dietary ) factors . Lipids are routinely evaluated in the screening for and monitoring of metabolic disorders . Effectively controlling serum lipids is a key intervention for metabolic disorders , providing a compelling motivation for investigating the genetic determinants of these traits , as new understanding of biology and potential drug targets can be achieved using this approach . Heritability estimates for these traits suggest that they are highly heritable , with a range of 43–76% for high-density lipoprotein cholesterol ( HDLC ) and 28–71% for triglycerides ( TG ) among those of European ancestry [1]–[6] ( with overlapping estimates among African ancestry individuals [6]–[9] ) . While large-scale efforts have made considerable progress in identifying genetic factors underlying the distribution of serum lipids ( for instance [10] ) , the focus of the majority of reports of the genetic epidemiology of serum lipids in diverse populations has been on replication or fine-mapping of variants that were identified in European ancestry individuals [10]–[15] . Although agreement between findings in samples of different ancestries does provide support for the significance of specific variants , this approach can only give a limited understanding of the genetic factors that influence trait distribution in diverse populations as it ignores variation of importance in the replication sample that would not be identified in the initial sample ( due to interethnic frequency differences , for example ) . The existence of interethnic differences in distribution of serum lipids between African Americans ( AA ) and individuals of European ancestry is known [16] . AA individuals generally have healthier lipid profiles than those of non-African ancestry , counter to expectation based on distributions of lifestyle factors that influence serum lipids . In nationally-representative data , for instance , mean serum triglycerides were 113 mg/dl in AA and 143 mg/dl in European Americans ( EA ) , and high-density lipoprotein cholesterol ( HDLC ) was higher in AA compared to EA ( 54 vs . 50 mg/dl ) [12] . The fact that these differences are seen in children [17]–[19] and that low TG has also been observed among those of similar genetic ancestry but widely divergent environments ( for instance , among African Americans and West Africans [16] ) provide strong evidence for a role of genetic factors . Further support for this inference comes from the observation that HDLC level increases with increasing proportion of genome-wide African ancestry in AA; this proportion is associated inversely with TG [16] , [20] . Taken together , these observations suggest the contribution of genetic variation that is highly differentiated or not shared between populations in influencing serum lipids , motivating African-ancestry focused analyses . Although there are many genes that have been associated with serum lipids that could have been selected for this study , we focused on the following 5 genes because of their potential ( based on literature review ) to provide novel insights into the well-documented differences between lipid profile of EA and AA: ATP-binding cassette A1 ( ABCA1 ) , lecithin-cholesterol acyltransferase ( LCAT ) , lipoprotein lipase ( LPL ) , paraoxonase 1 ( PON1 ) , serpin peptidase inhibitor E1 ( SERPINE1 ) . ABCA1 is a membrane-associated protein that is central to reverse cholesterol transport , acting as an efflux pump to facilitate the removal of cellular lipid to apolipoprotein A-I . Sequence variants in ABCA1 have consistently been associated with HDLC concentration [10] , [12] , [15] , and variations in this gene lead to Tangier disease , defined by extremely low levels of HDLC [21] . LCAT converts free cholesterol into cholesterol ester , a key step in the formation of HDL , and sequence variants in LCAT are associated with HDLC concentration [10] , [12] , [15] , [22] , [23] . LPL hydrolyzes TG and releases fatty acids . Sequence variants in LPL are associated with TG [10] , [12] , [15] , [20] , [22] and HDLC [10] , [12] , [15] . PON1 hydrolyzes a wide range of substrates and protects against lipid oxidation , being largely responsible for the antioxidative properties of the HDL particle . PON1 is of particular interest based on a protective role for atherosclerosis and related outcomes ( reviewed in [24] ) . Additionally , a linkage analysis for HDLC in West Africans identified a region that includes PON1[7] . SERPINE1 encodes plasminogen activator inhibitor-1 , an important regulator of fibrinolysis . PAI-1 concentration is associated with both CVD [25] and Metabolic Syndrome [26] , and sequence variation in SERPINE1 is associated with TG [27] , [28] . We sought to take a comprehensive , in-depth look at the association between these selected lipid candidate genes and serum lipids in African Americans . We sequenced these genes in 48 individuals from the extremes of the distribution of serum lipids in a population-based sample and genotyped the identified variants in the full cohort ( n = 1694 ) . The association between these variants and serum lipids was evaluated using separate rare and common variant analyses . We were able to identify lipids-associated variants that were African ancestry-specific ( or at much higher frequency in African ancestry ) and variants with a different effect depending on ancestry .
We sequenced 98 , 901 base pairs across selected regions of the five candidate genes ( including known or predicted exons , flanking introns , 5′ untranslated region ( UTR ) /promoter region , and evolutionarily conserved regions of each gene ) , distributed as follows: ABCA1 64 , 516; LCAT 4 , 432; LPL 10 , 131; PON1 9 , 999; and SERPINE1 9 , 823 . The frequency and type of SNPs discovered in this sequencing stage were not different from the distribution of variants identified in the same regions in the 1000 Genomes data ( AFR ) [29] when data was limited to variants with MAF≥0 . 01 ( our power to detect variants with a MAF<0 . 01 was less than 60% given the number of chromosomes interrogated [n = 96] ) . Given that our interest was in describing all the variation in these genes , including variants that might not have been detected in our sequencing stage , the 675 variants identified were supplemented with imputation using 1000 Genomes data as reference [29] for a total of 1 , 918 variants carried forward for genotyping in the larger cohort . There were 110 variants that were only present among those in the extreme group with a favorable lipid profile and 115 variants that were only found in the extreme group with an unfavorable lipid profile ( Table S2 ) . Of these variants that were exclusive to one extreme group that were not successfully genotyped or imputed for follow-up in the full study population , there were two of note . rs268 in LPL has been previously associated with serum lipids and related disease outcomes [30]–[32]; this variant was excluded from analysis in the full study population because it was not in HWE . ABCA1 variant rs35819696 was previously identified in AA from the Dallas Heart Study who had low HDLC [33]; this variant was found only in the unfavorable lipid profile group , but it was monomorphic in the full study population , precluding further analysis . Of the 1 , 918 variants identified in the sequencing and imputation stages , 1 , 415 were successfully genotyped or imputed in the full sample of 1 , 694 individuals . In the rare variant ( RV ) analysis , one association was found between gene-defined SNP set and serum lipids that remained statistically significant after correction for multiple comparisons . RVs in ABCA1 were associated with logTG ( p = 0 . 0095 ) . No associations were observed for any of the other genes with logTG or for any of the genes with HDLC in the RV analysis . Haplotype analyses were conducted by gene but no haplotypes were statistically significant after permutation testing . Statistically significant results from the common variant ( CV ) analysis , after correction for multiple hypothesis testing , are presented in Table 2 and reviewed below by gene . All effects are described in terms of the minor allele . CVs in LPL were associated with serum lipids . The well-known missense variant rs328 and variants that are in linkage disequilibrium ( LD ) with it were associated with increased HDLC and decreased logTG . Across this region , individuals who were homozygous for the risk allele had ∼15 mg/dl higher HDLC compared to those with other genotypes ( p = 0 . 0002 for rs328 ) , while each minor allele was associated with a ∼5% decrease in TG ( p = 0 . 0001 for rs328 ) . A variant in LD ( R2 = 0 . 6 ) , rs12679834 , was also associated with logTG ( p = 0 . 00009 ) , but not HDLC ( p = 0 . 2 ) . For further understanding of the association between variants in this region and serum lipids , the influence of local ancestry on the reported associations was evaluated . As with all analyses in this study , models were adjusted for genome-wide average proportion of African ancestry; thus , local ancestry effects observed should not be confounded by genome-wide African ancestry . Nearly all of the associations in this region were significantly modified by local ancestry ( Table S3 ) . As illustrated for rs328 ( Figure 1 ) , the variant was associated with a larger effect size on the European- compared to the African-ancestry background in this admixed sample of AA . Statistically significant interactions were observed between rs328 carrier status and local ancestry for both HDLC ( pinteraction = 0 . 008 ) and logTG ( pinteraction = 0 . 01 ) . Notably , among AA with only European ancestry at this locus , carriers of the minor allele had 12 . 7 mg/dl higher HDLC than those homozygous for the major allele ( p = 0 . 02 ) ; among those with African ancestry at this locus , rs328 was not associated with HDLC ( difference between carriers and non-carriers of the minor allele <2 mg/dl ) . Similarly , among those with either 1 or 2 copies of the European ancestry allele at this locus , rs328 carriers of the minor allele had 10 . 5% lower TG levels than non-carriers ( p = 0 . 00002 ) , while this difference was reduced substantially to 3 . 3% ( p = 0 . 06 ) among those with two African-ancestry alleles . As expected , genome-wide proportion of African ancestry varied within category of local ancestry: mean genome-wide African ancestry was 68% , 75% , and 82% among those with 0 , 1 , and 2 copies of the African ancestry allele , respectively . There was evidence for replication of the association between some of these variants and logTG among WA . The association between rs328 and logTG was much smaller and not statistically significant among the WA: the minor allele was associated with 2% lower TG ( p = 0 . 3 ) . There was insufficient power to evaluate the rs328-HDLC relationship in WA . Nearby variant rs12679834 ( R2 with rs328 = 0 . 81 ) was associated with a 5% lower TG in WA ( p = 0 . 04 ) . The stronger association with rs12679834 may result from the slightly higher MAF of this variant compared to rs328 ( 0 . 09 vs . 0 . 07 ) . Of note is the consistency of the ancestry effects for both HDLC and logTG: the association observed for these LPL variants among WA matched very closely with what was observed among the admixed AA with primarily African ancestry at this locus ( Figure 1 ) . Another LPL variant , rs1059611 ( and rs149865365 , R2 = 0 . 99 with rs1059611 ) , which was not in LD with the other LPL variants ( R2<0 . 3 for all comparisons ) , was associated with 2 mg/dl higher HDLC; there is also evidence of significant modification of this association by local ancestry ( pinteraction = 0 . 004 ) . As above with rs328 , greater differences in HDLC by minor allele carrier status were observed on the European ancestry background compared to the African ancestry background ( 14 . 6 for 2 copies of the European ancestry allele vs . <2 mg/dl with any copy of the African ancestry allele ) . Despite reasonable power ( 0 . 74 ) to detect an association of the magnitude observed in AA among WA , the rs1059611 finding did not replicate ( −0 . 6 mg/dl , p = 0 . 6 ) ; this result is consistent with the observed local ancestry effect ( with association in AA seen only on the European ancestry background ) . Several intronic CVs in ABCA1 that are predicted to affect a variety of regulatory motifs [34] were associated with altered serum lipids ( Table 2 ) . Four of these CVs are in much higher frequency among African ancestry individuals , with rs78294949 , rs114851717 , and rs73521828 only found among those with African ancestry , and clear differentiation by ethnicity for rs115763221 ( 1000 Genomes [29] MAF: 0 . 17 ( AFR ) , 0 . 003 ( EUR ) ) . LCAT variant rs13306496 was associated with 5 mg/dl lower HDLC . This intronic variant does not alter LCAT protein sequence , but it is predicted to alter regulatory motifs ( HaploReg [34] ) . This variant is present at a higher frequency among African-ancestry populations compared to other ethnicities ( 1000 Genomes [29] ) : MAF 0 . 16 ( AFR ) and 0 ( EUR ) . Variants in this gene were not associated with TG . In PON1 , two CVs were associated with serum lipids . A common PON1 variant , rs2049649 , was associated with HDLC . This intronic variant alters regulatory motifs and promoters [34] . A small deletion , rs3917549 , was associated with lower logTG . rs3917549 alters regulatory motifs and is in much higher frequency among African ancestry vs . European ancestry populations ( AFR: 0 . 62 , EUR: 0 . 18 ) . For SERPINE1 , one CV was associated with serum lipids . SERPINE1 CV rs2227674 was inversely associated with logTG . This variant has been previously associated with plasminogen activator inhibitor-1 levels [35] .
In this study , we undertook a comprehensive evaluation of the sequence variation present in 5 serum lipids candidate genes and their association with HDLC and TG to shed light on the consistently observed , yet unexplained , different lipid profiles seen in African Americans ( AA ) compared to European Americans ( EA ) . Genes ABCA1 , LCAT , LPL , PON1 , and SERPINE1 were each sequenced in 48 AA individuals , and the variants identified were genotyped in a population-based sample of AA . Using a variety of analytical techniques , we were able to describe the genetic architecture of these traits at these loci in AA . Notable among our findings , in terms of underscoring the importance of African ancestry-focused studies of serum lipids , are the discovery of lipids-associated variants that are in higher frequency ( or only found ) among African ancestry individuals , loci for which effect size differed by genetic ancestry in this admixed population , and an opportunity to take advantage of interethnic LD differences . One of the motivations for conducting population-specific work is that some risk variants may be absent or at different frequencies by ethnicity . For example , variants of minimal impact among European ancestry individuals may have a greater significance in AA because of a higher frequency . Of those variants that were analyzed in this study , all of the LCAT variants were in significantly higher frequency ( χ2 test , p<0 . 05 ) among the 1000 Genomes African vs . European ancestry samples ( AFR vs . EUR ) ; for the other genes , the majority of variants followed this pattern ( ABCA1 74% , LPL 70% , PON1 64% , SERPINE1 52% ) . Notably higher minor allele frequencies , among 1000 Genomes African vs . European ancestry samples ( AFR vs . EUR ) , were observed for some of the associated CVs including ABCA1 variant rs115763221 ( AFR: 0 . 17 , EUR: 0 . 003 , p<0 . 0001 ) , LCAT variant rs13306496 ( AFR: 0 . 16; EUR: 0 . 003 , p<0 . 0001 ) , and LPL variant rs201109344 ( AFR: 0 . 08 , EUR: 0 . 001 , p<0 . 0001 ) . Three ( rs78294949 , rs114851717 , and rs73521828 ) of the associated ABCA1 CVs were not present in the 1000 Genomes European ancestry samples . Many of the RVs in this study were African ancestry specific , and 65% of the RVs analyzed were found in AFR and not in EUR . Clearly , these associations could not be evaluated in a non-African ancestry sample . Interethnic differences in the genetic architecture of serum lipid traits extend beyond simple frequency differences as highlighted with this study's observation of variants with differing effect sizes by local ancestral background . A common LPL nonsense variant , rs328 ( S447X ) , and variants in LD with rs328 were associated with both HDLC and TG in this study . This variant is associated with increased LPL mRNA [36] , increased LPL activity [37] , decreased TG [20] , [38]–[40] , and increased HDLC [41]–[44] . In this study of admixed AA , much larger associations were observed among those with European ancestry at this locus as compared to those with predominantly African ancestry at this locus , as has been previously observed [20] . The local ancestry-stratified logTG outcomes in AA are remarkably consistent with observations in the respective parental populations: the size of the observed association was 2–3% in AA with only African local ancestry and WA , while the larger effect size among AA with local European ancestry ( 10 . 5% ) is nearly identical to what was reported in a meta-analysis of over 43 , 000 predominantly European ancestry individuals ( 10% ) [45] . The local ancestry-stratified HDLC outcomes also show a larger effect size with local European ancestry , but the comparisons with parental populations are more complex . While the effect size among AA with local African ancestry was consistent with what was seen in WA , the effect size among AA with local European ancestry was more than twice what was observed in those with European ancestry ( 12 . 7 vs . ∼5 mg/dl higher for GG vs . CC ) [41]–[44] . Thus , this variant is showing a larger association in AA than has been previously reported for European ancestry populations , even when limiting to AA individuals with local European ancestry . This inconsistency suggests the presence of a genetic or environmental factor that influences the rs328-HDLC association and differs by ethnicity . Effect modification of the rs328-HDLC association has been observed with dietary fat parameters [39] , [44] . Intriguingly , in one analysis , an interaction that differed by ethnicity was observed for HDLC , and a similar interaction was not observed for TG , in agreement with our TG findings , which were consistent within genetic ancestry categories . The generally reduced LD in the genomes of African ancestry individuals compared to European ancestry individuals can be used to narrow the region of interest around an association signal ( trans-ethnic fine-mapping ) . Similar associations have been observed among those of European and Asian ancestry for rs1059611 and rs328 with serum lipids [39] , [46] , [47]; this similarity is unsurprising given the high LD between these variants among these populations ( R2 = 0 . 96 ) . In this study of AA , however , these variants were not in LD ( R2 = 0 . 28 ) , the rs328-HDLC effect size was nearly 9-fold larger than the rs1059611-HDLC effect size , and no association was observed for rs1059611-logTG . It is expected that interethnic differences in relevant genetic variants may contribute to the observed differences in the distribution of serum lipids in African ancestry individuals compared to those of other ethnicity ( with more favorable lipid profiles , higher HDLC and lower TG among AA ) . The most compelling case to be made for a genetic contribution to the interethnic serum lipid differences may be with rs328 ( and variants in LD with rs328 ) and HDLC . These variants are associated with a favorable lipid profile , but the associated variants in this region are consistently in ∼5% higher minor allele frequency among EA than AA , and less common among WA . However , for HDLC , the effect size in AA was 15 mg/dl for variants in this region , while among EA the effect size is 4–6 mg/dl . Interestingly , though TG is consistently low across African ancestry populations , HDLC is generally much higher among AA than WA ( mean HDLC: AA 53 . 0 , WA 39 . 3 mg/dl ) . This HDLC distribution is consistent with the possibility of variants from European ancestry influencing HDLC among AA , but with a larger effect in AA , perhaps due to some gene × environment interaction ( as has been reported for rs328 [39] , [44] ) . A similar approach to ours was undertaken in the Dallas Heart Study [33] . In this study , Cohen et al found that nonsynonymous sequence variants were more common among individuals in the lowest vs . the highest 5% of the distribution of HDLC in their population-based study . While we did not find an excess of nonsynonymous variants in those sequenced compared to what was found in the 1000 Genomes data , one of the variants identified among African Americans in their low HDLC group was also observed in our unfavorable lipid profile group , providing further support for the role of ABCA1 variant rs35819696 in influencing serum lipids . Some strengths of this study deserve mention . First , the fact that this project began with sequencing AA at the extremes of the lipid distribution is significant . Selecting variants in this way , as opposed to simply trying to assess the replication of findings identified in other populations , allowed us to address the question of what variation influences these traits in African-ancestry individuals instead of evaluating how similar results are across ethnicities . This distinction is important given the evidence that the distribution of serum lipids differs by ancestry and the evidence of a different relationship between serum lipids and metabolic disorders ( reviewed in [16] ) . Additionally , with the increasing emphasis in complex disease research on rare variation , which is less shared across ethnicities than common variation , beginning with variants identified directly in an AA sample is appropriate . Another strength of this study was the inclusion of supporting information from WA individuals . Given the genetic similarity between these populations ( ∼80% genome-wide shared ancestry with AA ) and the widely divergent lifestyle , diet , and environmental contexts between them , such comparisons are invaluable in disentangling the basis of complex traits in admixed individuals . In this analysis , these data were particularly informative for evaluating the loci for which there was a different association by local ancestry . This study had a few limitations which should be considered . The selection of these candidate genes on which to focus our study is certainly insufficient to describe all regions for which genetic factors may play a role in the interethnic differences in serum lipids and it was not our intention to do so . As we sequenced 48 individuals ( 96 chromosomes ) and excluded variants that were only observed once in these individuals , it is expected that the variants identified do not fully describe the variation present , although a serious effort to address this lack was made by supplementing our lab-generated data with variants imputed using the 1000 Genomes samples as a reference [29] . Additionally , the number of WA samples that matched our inclusion criteria and could be used for replication was relatively limited , precluding comprehensive evaluation in WA for the majority of variants that were associated among the AA . The lack of ethnic diversity in genetic research has come to prominence , with large-scale efforts to reduce this disparity underway ( for example , the Human Heredity and Health in Africa [H3Africa] initiative , http://www . h3africa . org/ ) . In this analysis , we targeted the genetic architecture of specific candidate genes , focusing on the variation that was identified directly in an AA sample , yielding useful insights into the interethnic differences in the genetic determinants of these traits . Particularly informative was an association between the LPL locus and serum lipids that differed by local genetic ancestry , with data from parental populations supporting inferences and demonstrating the complexity of ancestry effects . African ancestry-focused work is an important part of understanding the role of genetics in the distribution of serum lipids .
Participants included in these analyses were from the Howard University Family Study ( HUFS ) , which has been described in detail previously [48] . Briefly , the HUFS followed a population-based selection strategy designed to be representative of African American families living in the Washington , DC metropolitan area . Ethnicity was ascertained by self-report . The HUFS was approved by the Howard University Institutional Review Board , and was conducted in accordance with the Declaration of Helsinki . All participants provided written informed consent . There were 1694 individuals included in the study after application of exclusion criteria: extreme phenotype values ( HDLC<20 or >100 mg/dl; TG<20 or >500 mg/dl ) and missing covariate data ( age , body mass index [BMI] , and gender ) . Given known perturbations to serum lipids that co-occur with Type 2 Diabetes , subjects with fasting blood glucose ≥126 mg/dl or taking physician-prescribed diabetes medication were excluded . These data include related participants . Family members were included in the common variant analysis , where adjustment for the random effect of family was possible ( further description below ) . For the RV analysis , only unrelated family members were included ( with one randomly-selected individual from each family , total n = 919 ) . Serum measurements were made on fasting samples . HDLC and TG were determined enzymatically with the COBAS Integra 400 Plus Analyzer ( Roche Diagnostics , Indianapolis , IN ) . Methods were standardized to in-house and other appropriate reference methods: CDC reference methods for HDLC , Isotope dilution-mass spectrometry ( ID-MS ) for TG by the manufacturer . Forty-eight unrelated subjects were selected from the HUFS , with 24 from the “favorable lipid profile” category ( lowest TG quartile , highest HDLC quartile ) and 24 from the “unfavorable lipid profile” category ( highest TG quartile , lowest HDLC quartile ) . A sample of 48 individuals ( 96 chromosomes ) provides a 99% probability of finding a sequence variant with a minor allele frequency ( MAF ) of 0 . 05 , 86% probability of finding a variant with an MAF of 0 . 02 and 62% probability of finding a variant with an MAF of 0 . 01 . The sequencing strategy applied methods as previously described [49] . Briefly , we sequenced known or predicted exons ( including 5 bp of the flanking introns ) , ∼1 kb of the 5′ untranslated region ( UTR ) /promoter region , and the 3′ UTR if it is evolutionarily conserved . We also sequenced up to three of the most evolutionarily conserved regions of each gene that are not captured by the above features . Sequencing primers were designed to allow for sufficient overlap in individual sequencing reads . Florescent dye-terminator chemistry was used for bi-directional DNA sequencing , and sequence delineation was performed by automated ABI Prism 3730xl DNA sequencers , which typically give >650 bp Q20/Phred20 read lengths . Mutations and heterozygotes were scored by automated comparative analysis against the provided reference sequence . All mutations were confirmed by manual curation . Sequence variants were selected based on frequency ≥2% ( to exclude variants that were only observed once , which may reflect error ) , for genotyping in the full HUFS sample . 174 of these identified variants were available from previous genotyping in this sample using the Affymetrix Genome-Wide Human SNP Array 6 . 0 [48] . Primers were designed for an additional 103 SNPs ( Table S4 ) , and genotyping was performed using the iPLEX Gold assay on the MassArray platform ( Sequenom , San Diego , CA ) as previously described [50] . Briefly , the PCR and extension primers were designed using MassArray designer Software . SNPs were excluded for assay failure ( n = 19 ) , lack of variation ( n = 13 ) , and genotype success rate <90% ( n = 24 ) . After these exclusions , none of the variants failed the filter for departure from Hardy-Weinberg equilibrium ( p-value<0 . 000001 ) . As our goal was to analyze the variation present in these genes in the individuals with extreme lipid profiles as comprehensively as possible , the list of variants identified by sequencing strategies was augmented by variants identified for these individuals from previous GWAS and by imputation based on 1000 Genomes dataset . Imputation was performed using MaCH-Admix [51] , an imputation tool specifically designed for use in admixed samples , using a cosmopolitan reference panel based on 1000 Genomes data . Imputed variants were filtered on an R2 of 0 . 3 . Rare variants with a MAF<0 . 01 were excluded due to diminished confidence in imputation in variants below this threshold . With previous ( n = 174 ) and new ( n = 47 ) genotyping , and imputation ( n = 1 , 194 ) , there were a total of 1 , 415 variants were analyzed ( Table S5 ) . These included 921 common variants ( CVs; MAF ≥ 0 . 05 ) and 494 less common or rare variants ( RVs; MAF < 0 . 05 ) distributed as follows within the candidate genes: ABCA1 ( 636 CVs , 389 RVs ) , LCAT ( 6 CVs , 2 RV ) , LPL ( 139 CVs , 48 RVs ) , PON1 ( 110 CVs , 33 RVs ) , and SERPINE1 ( 30 CVs , 22 RVs ) . TG was log-transformed , but the distribution of HDLC was approximately normal and left untransformed . Principal component analysis to assess population structure in the admixed African Americans was conducted using EIGENSOFT [52] , and , as reported previously [53] , the first principal component was retained on the basis of Velicer's minimum average partial test , and included in all analyses as a covariate representing overall proportion of African ancestry . All analyses were also adjusted for age , body mass index ( BMI ) , and gender , and P<0 . 05 after multiple test correction was considered statistically significant . All SNP effects are described in terms of the minor allele ( thus , an “inverse association” indicates that the minor allele was associated with decreasing phenotypic values relative to the major allele ) . Separate analytical strategies were employed for the analysis of common variants ( CVs ) and rare variants ( RVs ) . Variants with MAF ≥ 0 . 05 were included in the common variant ( CV ) analysis . For the CV analysis , the associations between variants and phenotypes were assessed in linear mixed models ( Proc Mixed ) in SAS 9 . 3 ( SAS Institute , Cary , NC ) with adjustment for age , BMI , gender , and the overall proportion African ancestry , as well as random clustering within families . For each variant , models were run for additive , dominant , recessive , and heterosis coding . To correct for the number of SNPs tested , P-values were adjusted for the effective number of SNPs ( based on LD>0 . 6 ) within the gene evaluated , as previously described [54] . Briefly , this method involves conducting a covariance matrix for all of the variants within a gene , using this covariance to determine the LD-adjusted number of independent tests that were conducted when analyzing all of the variants in that gene , and multiplying the individual P-values by that correction factor . All variants with MAF<0 . 05 were included in the less common and rare variant analysis ( hereafter referred to as the rare variant [RV] analysis ) . For the RV analysis , SKAT was used ( http://www . hsph . harvard . edu/skat/ ) [55] . SKAT aggregates individual score tests statistics for a set of SNPs , returning a P-value for the set ( in our implementation , each gene ) . Only unrelated participants were included ( all unrelated and one randomly-selected individual from each family , n = 919 ) . SKAT accommodates adjustment for covariates , and all analyses were adjusted ( as in the CV analysis ) for age , BMI , gender , and overall proportion African ancestry . Bootstrap resampling under the null model ( considering covariates ) was conducted , and statistical significance for the RV analysis was declared after correction for a family-wise error rate of 0 . 05 . For follow-up in the LPL region , local ancestry at the locus was estimated as previously described [56] . Briefly , ancestry at each locus was categorized as having 0 , 1 , or 2 chromosomes of African ancestry as estimated based on nearly 800 , 000 markers using LAMPANC version 2 . 3 [57] and HapMap Phase II+III CEU and YRI reference allele frequencies ( http://hapmap . ncbi . nlm . nih . gov/downloads/frequencies/2010-08_phaseII+III/ ) . A difference in genotype-phenotype association by local ancestry was evaluated in SAS using linear mixed models ( PROC MIXED ) with a genotype by local ancestry interaction term and evaluating models stratified by local ancestry . Haplotype analysis was conducted in all candidate genes using a sliding window approach in PLINK [58] , with up to 5 SNPs included in each haplotype and adjustment for covariates ( age , gender , BMI , and overall proportion African ancestry ) . Permutation testing ( 1000 permutations ) was employed to evaluate statistical significance . Identified loci in African Americans were assessed for replication in a West African sample obtained from the African American Diabetes Mellitus study ( AADM; described previously [59] ) . Briefly , AADM is a large-scale case-control study designed to explore the genetic and environmental determinants of T2D from West Africa , but only non-diabetic controls were included in this analysis . All participants provided written , informed consent . Variants were genotyped using the Affymetrix Axiom Genome-Wide Pan-African Array Set ( ∼2 . 2 million markers ) , which is optimized for coverage of African-ancestry populations . Imputation was also conducted in this sample ( as described above ) . A limited number of participants with genotype data remained after applying the exclusions described above ( n≤536 ) . This sample had 80% power to detect an effect of 7% for logTG and 5 mg/dl for HDLC when the variant was common ( MAF = 0 . 05 ) , and 80% power to detect effects of 15% and 11 mg/dl when the variant was less common ( MAF = 0 . 01 ) . The variants for which there was at least moderate power ( >60% , QUANTO [60] ) to detect an association of the magnitude observed in African Americans are described in the text .
|
Most of the work on the genetic epidemiology of serum lipids in African Americans ( AA ) has focused on replicating findings that were identified in European ancestry individuals . While this can be very informative about the generalizability of lipids loci across populations , African ancestry-specific variation will be missed using this approach . Our aim was to comprehensively evaluate five lipid candidate genes in an AA population , from the identification of variants of interest to population-level analysis of high-density lipoprotein cholesterol ( HDLC ) and triglycerides ( TG ) . We sequenced five genes in individuals with extreme lipids ( n = 48 ) drawn from a population-based study of AA . The variants identified were genotyped in 1 , 694 AA and analyzed . Notable among the findings were the observation of ancestry specific effect for several variants in the LPL gene among these admixed individuals , with a greater effect observed among those with European ancestry in this region . These associations were further elucidated by replication in West Africans . By beginning with the sequence variation present among AA , investigating ancestry effects , and seeking replication in West Africans , we were able to comprehensively evaluate these candidate genes with a focus on African ancestry individuals .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"biochemistry",
"lipids",
"epidemiology",
"genetics",
"biology",
"human",
"genetics",
"genetic",
"epidemiology"
] |
2014
|
Gene-Based Sequencing Identifies Lipid-Influencing Variants with Ethnicity-Specific Effects in African Americans
|
Histone H3 lysine-4 ( H3K4 ) methylation is associated with transcribed genes in eukaryotes . In Drosophila and mammals , both di- and tri-methylation of H3K4 are associated with gene activation . In contrast to animals , in Arabidopsis H3K4 trimethylation , but not mono- or di-methylation of H3K4 , has been implicated in transcriptional activation . H3K4 methylation is catalyzed by the H3K4 methyltransferase complexes known as COMPASS or COMPASS-like in yeast and mammals . Here , we report that Arabidopsis homologs of the COMPASS and COMPASS-like complex core components known as Ash2 , RbBP5 , and WDR5 in humans form a nuclear subcomplex during vegetative and reproductive development , which can associate with multiple putative H3K4 methyltransferases . Loss of function of ARABIDOPSIS Ash2 RELATIVE ( ASH2R ) causes a great decrease in genome-wide H3K4 trimethylation , but not in di- or mono-methylation . Knockdown of ASH2R or the RbBP5 homolog suppresses the expression of a crucial Arabidopsis floral repressor , FLOWERING LOCUS C ( FLC ) , and FLC homologs resulting in accelerated floral transition . ASH2R binds to the chromatin of FLC and FLC homologs in vivo and is required for H3K4 trimethylation , but not for H3K4 dimethylation in these loci; overexpression of ASH2R causes elevated H3K4 trimethylation , but not H3K4 dimethylation , in its target genes FLC and FLC homologs , resulting in activation of these gene expression and consequent late flowering . These results strongly suggest that H3K4 trimethylation in FLC and its homologs can activate their expression , providing concrete evidence that H3K4 trimethylation accumulation can activate eukaryotic gene expression . Furthermore , our findings suggest that there are multiple COMPASS-like complexes in Arabidopsis and that these complexes deposit trimethyl but not di- or mono-methyl H3K4 in target genes to promote their expression , providing a molecular explanation for the observed coupling of H3K4 trimethylation ( but not H3K4 dimethylation ) with active gene expression in Arabidopsis .
Histone lysine methylation regulates chromatin structure and gene transcription in eukaryotes . Various lysine residues on histones can be methylated and the ε-amino group of lysines can be mono- , di- and tri-methylated . Lysine methylation is linked with transcriptional activation or repression depending on the particular residue that is methylated and the degree of methylation [1] . For instance , H3 lysine-27 trimethylation ( H3K27me3 ) is exclusively involved in transcriptional repression , whereas H3K4 trimehtylation is associated with actively transcribed genes [2] . Recent genome-scale analyses of H3K4 methylation have revealed that different H3K4 methylation states are often associated with distinct transcription states in a gene . In the well-studied Saccharomyces cerevisiae , H3K4 trimethylation is a mark for actively transcribed genes , whereas mono- or di-methylation of H3K4 is not linked with active gene expression [3] . In Drosophila and mammals , both di- and tri-methylation of H3K4 are associated with gene activation [4] , [5] . In contrast to animals , in the higher plant Arabidopsis thaliana , only H3K4 trimethylation , but not mono- or di-methylation of H3K4 , is implicated in transcriptional activation [6] . H3K4 methylation is catalyzed by various methyltransferases . In Saccharomyces cerevisiae , the COMPASS ( for Complex Proteins Associated with Set1 ) H3K4 methyltransferase complex catalyzes H3K4 methylation [7] . This complex contains an H3K4 methyltransferase called Set1 , the only known H3K4 methyltransferase in yeast . By itself , Set1 is unable to catalyze H3K4 methylation , and requires other structural components in the complex for its catalytic activity [8] , [9] . COMPASS-like complexes have been identified in mammals , and so far , five such complexes known as hSet1 and MLL1 , MLL2 , MLL3 and MLL4 have been biochemically purified [for a review , see [2]] . All of these complexes contain four core components including an H3K4 methyltransferase and three structural core components known as WDR5 , Ash2 and RbBP5 , homologs of the yeast SWD3 , BRE2 and SWD1 , respectively [2] . WDR5 , Ash2 and RbBP5 together form a stable core subcomplex that provides a structural platform for H3K4 methylation [10] . The Ash2-RbBP5-WDR5 subcomplex interchangeably associates with different H3K4 methyltransferases such as hSet1 , MLL1 and MLL2 to form different catalytic complexes , and is essential for both di- and tri-methylation of H3K4 [9] , [10] . An in vitro reconstituted four-component mini-complex composed of WDR5 , Ash2 , RbBP5 and MLL1 methylates H3K4 specifically [9] . So far , COMPASS-like complexes have been identified in mammals , but still remain to be identified in other multicellular organisms such as plants . H3K4 trimethylation typically occurs concomitantly with active gene transcription , and trimethyl H3K4 ( H3K4me3 ) predominantly accumulates in the 5′ transcribed regions [2] , [11] . In yeast , the RNA Polymerase II Associated Factor 1 complex ( Paf1c ) recruits the COMPASS complex to the initiating and early-elongating RNA Polymerase II ( Pol II ) , resulting in H3K4 trimethylation in the 5′ genic regions [12] . A similar mechanism might exist in multicellular organisms as Paf1c appears to be evolutionarily conserved in animals and plants [13]–[15] . Although H3K4me3 is a chromatin mark for actively transcribed genes , whether it plays an active role in transcriptional activation remains unclear because the accumulation of H3K4me3 in the 5′ transcribed regions might merely result from active transcription by Pol II [16] . Recent studies have revealed a few components involved in H3K4 methylation in Arabidopsis . It has been shown that the Arabidopsis Paf1c is required for the accumulation of H3K4me3 in the 5′ regions of actively transcribed genes in Arabidopsis genome [15] . Bioinformatic and phylogenetic analyses reveal that there may be up to ten putative Arabidopsis H3K4 methyltransferases , among which are ARABIDOPSIS TRITHORAX1 ( ATX1 ) , ATX2 , SET DOMAIN PROTEIN25 ( SDG25; also known as ATXR7 ) , SDG14 and SDG16 [17] , [18] . Loss of ATX1 function causes a slight reduction in global H3K4me3 and accelerated developmental transition from a vegetative to a reproductive phase ( i . e . , flowering ) [19] . In addition , atx1 mutation leads to floral organ abnormalities in a particular Arabidopsis ecotype [20] . Recently , it has been shown that loss of ATX2 or ATXR7 function causes moderately early flowering [21]–[23] . Besides these putative H3K4 methyltransferases , Arabidopsis has two homologs of the human COMPASS-like complex core component WDR5 , namely WDR5a and WDR5b [24] . WDR5a binds histone H3 tails , is involved in H3K4 methylation in its target gene chromatin , and represses Arabidopsis flowering [24] . The timing of floral transition in Arabidopsis is genetically controlled by a network of flowering genes , among which FLC plays a central role . FLC , a MADS box transcriptional factor , is a key floral repressor that quantitatively inhibits the floral transition in Arabidopsis [for a review , see [25] , [26]] . Besides FLC , there are five FLC homologs in Arabidopsis: FLOWERING LOCUS M ( FLM ) /MADS BOX AFFECTING FLOWERING 1 ( MAF1 ) and MAF2-MAF5 [27] , [28] . FLM , MAF2 and MAF4 moderately repress flowering [27]–[29] , whereas the roles for MAF3 and MAF5 remain unclear [28] . FLC expression is under complex control and chromatin modification plays a critical role in FLC regulation which has become a model for understanding plant gene regulation by chromatin-based mechanisms [26] , [30] , [31] . FLC is expressed at low levels in many early-flowering Arabidopsis accessions because of the repression by ‘autonomous-pathway’ genes , among which are a few histone modifiers including CURLY LEAF ( a putative H3K27 methyltransferase ) , HISTONE DEACETYLASE6 , FLOWERING LOCUS D ( FLD , a putative H3K4 demethylase ) and several Type I and II arginine methyltransferases [32]–[36] . These proteins mediate ‘repressive’ histone modifications in FLC chromatin and repress FLC expression to promote flowering . In addition to repressive histone modifications , FLC chromatin can also be modified by H3K4 methylation . Recent studies have revealed that H3K4 trimethylation is linked with activation of FLC expression . H3K4me3 accumulates in the region around the transcription start site ( TSS ) of FLC including the 5′ end of transcribed region [13] , [21] . This modification requires Paf1c and WDR5a , and in addition , ATX1 and ATXR7 are partly required [13] , [15] , [21] , [22] , [24] , [37] . WDR5a can interact with ATX1 , binds to FLC chromatin , is required for H3K4 trimethylation in FLC and for FLC activation , and has been proposed to act in the context of an H3K4 methyltransferase complex [24] . So far , although several components involved in H3K4 methylation have been characterized in Arabidopsis , it remains essentially unknown whether H3K4 methyltransferase complexes exist in Arabidopsis . In addition , although it has been well documented that H3K4me3 accumulation is associated with actively transcribed FLC chromatin , whether H3K4 trimethylation can activate FLC expression is unknown . Here , we report that Arabidopsis homologs of the human COMPASS-like complex core components Ash2 and RbBP5 , together with WDR5a , form a nuclear subcomplex for H3K4 methylation during vegetative and reproductive development . The subcomplex component WDR5a can associate with several putative H3K4 methyltransferases , suggesting that multiple COMPASS-like complexes exist in Arabidopsis . Loss of ASH2R function causes a great decrease in genome-wide H3K4 trimethylation , but not di- or mono-methylation . In addition , we found that the ASH2R subcomplex mediates H3K4 trimethylaiton , but not H3K4 dimethylation , in FLC and FLC homologs to activate their expression resulting in delayed flowering . Our findings suggest that the ASH2R-containing COMPASS-like complexes ( ASH2R-COMPASS ) deposit H3K4me3 , but not H3K4me2 or H3K4me1 in Arabidopsis genome . Furthermore , we found that a null lesion in ASH2R causes arrested embryo development at globular stage and that ASH2R is also required for proper leaf growth and development , suggesting that the ASH2R-COMPASS-mediated H3K4 trimethylation plays important roles for multiple Arabidopsis developmental processes .
Recently we have identified two homologs of the human WDR5 , namely WDR5a and WDR5b in Arabidopsis genome [24] . To explore whether homologs of the other core COMPASS-like complex components Ash2 and RbBP5 , exist in Arabidopsis , we searched the Arabidopsis protein database with the amino acid sequences of these proteins and identified a single homolog for each protein , namely ASH2R ( At_1g51450 ) and RbBP5 LIKE ( RBL; At_3g21060 ) ( Figure S1 and Figure S2 ) . The sequence similarity between ASH2R and the human Ash2 over the entire ASH2R is 46% ( Figure S1 ) . A recent phylogenetic analysis of Ash2 homologs from several animals and plants also indicates that ASH2R ( also known as TRO ) is a clear homolog of the human Ash2 [38] . We carried out phylogenetic analysis of RbBP5 homologs from representative animal and plant species . As shown in Figure S2 , in each animal or plant species examined , there is only a single RbBP5 homolog; animal RbBP5 homologs form one clade , and plant homologs constitute another clade . These findings show that there are Ash2 and RbBP5 homologs in Arabidopsis . The evolutionary history of WDR5 homologs is complex . We found that there are multiple WDR5 homologs in land plants , for instance , three in the basal land plant moss , two in the monocot rice and three in the eudicot poplar , whereas in most animals , there is only a single WDR5 homolog ( Figure 1 ) . We further performed phylogenetic analysis of WDR5 homologs from representative species . The phylogenetic tree indicates that all WDR5 homologs from plants form a single clade with 64 bootstrap support , whereas all animal WDR5 homologs form another clade with 67 bootstrap support ( Figure 1 ) . There is no strong bootstrap support for an animal-plant clade . These results indicate that the animal WDR5 homologs might function like the human WDR5 , but raise a question on biochemical functions of the multiple WDR5 homologs in plants . It was of great interest to determine biological functions of ASH2R and RBL . First , we employed a double-stranded RNA interference ( dsRNAi ) approach to knock down RBL expression in wildtype Col ( knockout mutants in RBL were not publically available ) . Briefly , a 245-bp RBL-specific fragment from the 3′ transcribed region was used to create a dsRNAi cassette driven by the constitutive 35S promoter . Eight independent transgenic lines were generated , from which homozygous RBL RNAi-1 and RBL RNAi-2 lines with a single T-DNA locus , were identified . In long days ( LD ) , five out of the eight lines flowered earlier than parental Col , but otherwise were normal; no noticeable phenotypes were observed in the remaining three lines ( Figure 2A and 2B , and data not shown ) . RBL transcript levels were quantified in RBL RNAi-1 and -2 seedlings by real-time quantitative PCR , and indeed , RBL expression was knocked down compared to Col ( Figure 2C ) . Together , these results show that RBL represses the floral transition . FLC plays a central role in floral repression in Arabidopsis , and as described in the Introduction , several components involved in H3K4 methylation promote FLC expression . Hence , it was of interest to examine whether FLC expression was suppressed upon RBL knockdown . FLC transcript levels were reduced in RBL RNAi seedlings compared to Col ( Figure 2D ) . We further examined the expression of FLC homologs upon RBL knockdown , and found that MAF4 and MAF5 transcript levels were reduced , whereas levels of FLM , MAF2 and MAF3 transcripts in the RBL RNAi lines were similar to those in Col ( Figure 2D ) . Together , these results suggest that RBL promotes the expression of FLC , MAF4 and MAF5 , but not FLM , MAF2 or MAF3 , to repress the floral transition . To explore the biological function of ASH2R , we identified an ash2r-1/+ heterozygous line from the SAIL collection [39] , which carries an insertional T-DNA in the middle of ASH2R . We did not recover any ash2r-1 homozygotes from a large population of the selfed progeny of ash2r-1/+ heterozygotes . Subsequently , we found that in developing siliques from ash2r-1/+ plants , about one-fifth of the seeds ( 167 out of 788 seeds ) were aborted white and brown seeds ( Figure S3A ) , whereas in wild-type siliques only less than four thousandth of the seeds ( 3 out of 798 ) were aborted . Thus , we conclude that ash2r-1 homozygous seeds are not viable . Next , we examined ash2r-1 embryo development under a differential-interference-contrast microscope . Up to globular stage , ash2r-1 embryos developed similarly as wild-type ones , however , their further development was arrested ( Figure S3B-S3H ) . When wild-type embryos reached heart stage ( around 6 days after pollination; DAP ) , the ash2r-1 embryos were still at globular stage ( Figure S3E and S3H ) . These findings are in line with a very recent study by Aquea et al . [38] showing that ASH2R/TRO is essential for Arabidopsis early embryogenesis . To explore the role of ASH2R in vegetative development , we exploited a dsRNAi approach to knock down ASH2R expression . A 223-bp ASH2R-specific fragment was used to create a dsRNAi cassette driven by the 35S promoter . Four independent homozygous transgenic lines with a single T-DNA locus , ASH2R RNAi-1 to -4 , were created . These lines all flowered earlier than Col , but otherwise were normal in LD ( Figure 3A , 3B ) . In addition , these lines set seeds normally . ASH2R expression was examined in ASH2R RNAi seedlings; indeed , it was knocked down and the ASH2R transcript levels in these lines were about 30–50% of those in Col ( Figure 3C ) . Furthermore , we quantified the transcript levels of FLC and FLC homologs in these RNAi lines , and found that FLC , MAF4 and MAF5 expression was reduced , whereas FLM , MAF2 and MAF3 expression was not affected upon ASH2R knockdown ( Figure 3D ) . Thus , ASH2R , like RBL , is required for the expression of FLC , MAF4 and MAF5 and represses Arabidopsis flowering . We further investigated the role for ASH2R in vegetative development using two mutant lines in which ASH2R function is severely disrupted as detailed below . In an effort to create glucocorticoid-inducible ASH2R expression lines , we unintentionally introduced two point mutations T136 to A and K187 to R , into the ASH2R coding sequence . Phylogenetic analysis suggests that T136 is conserved among Ash2 relatives in eudicots , whereas K187 appears not to be conserved ( Figure S4B ) . The point-mutated ASH2R transgene was inserted downstream a synthetic promoter known as pOp6 ( Figure S4A ) , which is responsive to glucocorticoid treatment [40] . In the absence of glucocorticoid induction , we identified five independent lines homozygous for the null ash2r-1 mutation and carrying the point-mutated ASH2R transgene , in which the transgene expression is leaky ( see description below ) . Individual T3 populations for these five lines were obtained ( note that only transgenic progeny can grow into seedlings due to the embryo lethality of ash2r-1 ) ; subsequently , we examined a representative T3 population from each line phenotypically , and found that at seedling stage two lines ( Near-Normal #1 and #2 ) developed like wildtype Col ( WT ) , whereas the other three lines including ash2r-2 ( hypomorphic; thereafter ash2r-2hyp ) and ash2r-3 ( hypomorphic; thereafter ash2r-3hyp ) displayed leaf abnormalities ( Figure 4A , and data not shown ) . Of note , all of the progeny within each examined population displayed nearly identical phenotypes that were stably inheritable . In terms of flowering time , the two near normal lines flowered moderately earlier than Col , whereas the other three mutant lines flowered much earlier than Col in LD ( Figure 4B , and data not shown ) . Using real-time quantitative PCR , we further examined ASH2R transgene expression in shoot apices from 10-day ( d ) old seedlings , and found that the ASH2R transcript levels in both near-normal lines were higher than those in Col , whereas they were much lower in ash2r-2hyp and -3hyp compared to Col ( Figure 4C ) . This suggests that the increased transcript levels of point-mutated ASH2R can partly ameliorate the partial functionality of point-mutated ASH2R . Interestingly , in all of these five lines the ASH2R transgene was expressed in the absence of the chemical inducers glucocorticoids; most likely this is because the strong 35S promoter located at the 3′ end of ASH2R may promote its expression ( Figure S4A ) . To confirm that the early-flowering phenotypes of ash2r-2hyp and -3hyp , like those of the ASH2R RNAi lines , were caused by reduced expression of FLC and its homologs , we quantified the transcript levels of FLC , MAF4 and MAF5 in shoot apices of these lines . Indeed , FLC expression was greatly reduced , and both MAF4 and MAF5 expression was nearly eliminated in ash2r-2hyp and -3hyp compared to Col ( Figure 4D ) . Furthermore , we found that only about two-fifth of the siliques or silique-like structures of ash2r-2hyp and -3hyp bore seeds ( typically only several viable seeds in each of these siliques ) ( Figure S5 ) . As described earlier in the text , the ASH2R RNAi lines , unlike ash2r-2hyp and -3hyp , are normal except for early flowering as ASH2R expression is only partially suppressed in the RNAi lines . Together , these observations led us to conclude that both ash2r-2hyp and -3hyp are strong hypomorphic loss-of-function mutants , largely resulting from the point mutations of T136 to A and/or K187 to R and an expression reduction in the point-mutated ASH2R . The role of ASH2R in leaf growth and development was further explored . We observed that all of the ash2r-2hyp and -3hyp mutants ( T3 ) displayed leaf abnormalities: small , narrow , and sometimes curled leaf blades compared to wildtype Col ( Figure 4A ) , suggesting that ASH2R is required for proper leaf growth and development . In an effort to create transgenic lines with the FLAG-tagged ASH2R transgene ( ASH2R:FLAG ) , we identified one line named as ASH2R:FLAG-1a ( T3 generation ) homozygous for the null ash2r-1 mutation and carrying a single-locus ASH2R:FLAG fusion driven by the 35S promoter . The majority of ASH2R:FLAG-1a ( T3 ) seedlings were near normal as Col , but about one-third of them displayed leaf phenotypes: typically small and narrow leaf blades as exemplified in Figure 4A . These phenotypes were due to the partial functionality of ASH2R:FLAG , not caused by the increased levels of ASH2R:FLAG because ASH2R overexpression did not give rise to any leaf phenotypes ( see Figure 5A , next section ) . The leaf phenotype of ASH2R:FLAG-1a was similar to , but typically weaker than that of the strong hypomorphic ash2r-2hyp and -3hyp ( Figure 4A ) . Together , these findings demonstrate that besides flowering repression and seed development , ASH2R is also required for proper leaf growth and development . Loss of ASH2R function in vegetative development causes accelerated floral transition and leaf abnormalities . We sought to further explore the effects of gain of ASH2R function on Arabidopsis development . ASH2R coding region was overexpressed by the constitutive 35S promoter in Col . Eight independent transgenic lines were generated , among which three homozygous lines ( T3 generation ) with a single T-DNA locus , ASH2Rox-1 to -3 were further characterized . First , we confirmed that indeed ASH2R mRNA levels in these ASH2Rox lines were higher than those in parental Col ( Figure S6 ) . In opposite to the ASH2R RNAi lines , these three lines flowered later than Col ( Figure 5A , 5B ) . In addition , we also observed that three out of the other five lines flowered later than Col ( data not shown ) . Of note , except late flowering , these lines developed normally at vegetative phase ( Figure 5A ) , and no noticeable phenotypes were observed in seed development . We further quantified transcript levels of FLC and its homologs in ASH2Rox seedlings . In opposite to ASH2R knockdown , both FLC and MAF4 expression was upregulated in the ASH2Rox lines , and surprisingly , MAF5 expression was significantly activated upon ASH2R overexpression ( Figure 5C , 5D ) . Consistent with that ASH2R is not involved in the regulation of FLM , MAF2 and MAF3 , their transcript levels in the ASH2Rox lines were similar to those in Col ( Figure 5C and data not shown ) . Taken together , these results show that ASH2R plays an important role in activation of the expression of FLC , MAF4 and MAF5 to inhibit flowering . Interestingly , the significant activation of MAF5 expression upon ASH2R overexpression appears to have a limited effect on the floral repression . This could be attributed to that MAF5 may be a weak floral repressor . It was of interest to examine whether the spatial expression patterns of ASH2R , RBL and WDR5a overlap in vegetative and reproductive phases . To uncover the ASH2R spatial pattern , a 1 . 5-kb 5′ promoter region plus part of the genomic coding region of ASH2R was translationally fused with the β-GLUCURONIDASE ( GUS ) gene . In addition , a 1 . 1-kb promoter plus 0 . 6-kb the genomic coding region of RBL was translationally fused with GUS . Both transgenes were introduced into Col , and using histochemical staining we found that in vegetative phase both ASH2R and RBL were strongly expressed in root tips , shoot apices and vascular tissues ( Figure 6A ) . This pattern overlaps with those of WDR5a and FLC [21] , [24] . Next , we analyzed the spatial patterns of ASH2R-GUS , RBL-GUS and WDR5a-GUS in seed development . ASH2R-GUS was strongly expressed in developing embryos and endosperms ( Figure 6C , 6D ) , consistent with ASH2R playing an essential role in seed development . In addition , both WDR5a and RBL , like ASH2R , were expressed in developing embryos and endosperms at globular and heart stages ( Figure 6E–6H ) . Together , these findings are consistent with the notion that ASH2R , RBL and WDR5a might act as part of a complex in vegetative and seed development . We further explored whether ASH2R , RBL and WDR5a form a stable complex . First , yeast two-hybrid assays were carried out to examine direct associations among these three proteins . In yeast , ASH2R interacted with RBL , not with WDR5a , whereas RBL interacted with WDR5a ( Figure 7A and 7C ) . Next , to examine whether RBL associated with ASH2R and WDR5a in plant cells , we performed bimolecular fluorescence complementation ( BiFC ) experiments using non-fluorescent N-terminal and C-terminal EYFP ( for Enhanced Yellow Fluorescent Protein ) fragments ( named as nEYFP and cEYFP , respectively ) fused to the full-length WDR5a , ASH2R or RBL proteins . nEYFP-ASH2R and RBL-cEYFP were simultaneously expressed in onion epidermal cells , and fluorescence was observed in the nuclei of onion cells ( Figure 7B ) , demonstrating that ASH2R associates with RBL . Similarly , we also found that RBL associated with WDR5a in the nuclei of onion cells ( Figure 7D ) , consistent with that ASH2R , RBL and WDR5a form a complex . Our phylogenetic analyses of the WDR5 homologs show that there are multiple WDR5 homologs in land plants ( Figure 1 ) . Besides WDR5a , Arabidopsis has another WDR5 homolog , WDR5b . We sought to determine whether WDR5b could form a complex with RBL and ASH2R . Using yeast two-hybrid approach , surprisingly , we found that WDR5b did not interact with either RBL or ASH2R in yeast ( Figure S7 ) . These results suggest that in Arabidopsis there may be only a single core subcomplex for H3K4 methylation , namely ASH2R-RBL-WDR5a . We further performed co-immunoprecipitation experiments to determine if ASH2R , RBL and WDR5a form a complex in vivo using the ASH2R:FLAG-1a line ( T3 ) . Indeed , anti-FLAG specifically immunoprecipitated WDR5a from young seedlings and developing siliques ( Figure 7E–7F ) . As described above , ASH2R physically associates with RBL which directly associates with WDR5a ( Figure 7B and 7D ) , and there is no direct association between ASH2R and WDR5a ( Figure 7C ) . Thus , ASH2R may form a complex with WDR5a via RBL . To determine whether RBL is required for the in vivo complex formation of ASH2R with WDR5a , we crossed the RBL RNAi-1 into the ASH2R:FLAG-1a line , and subsequently , the F1 seedlings were used for co-immunoprecipitation . Upon RBL knockdown , the anti-FLAG recognizing ASH2R:FLAG failed to pull down WDR5a in RBL RNAi-1;ASH2R:FLAG-1a seedlings ( Figure 7G ) . Thus , RBL is required for the ASH2R-WDR5a complex formation . These results together led us to conclude that ASH2R , RBL and WDR5a form a complex during vegetative and reproductive development . The physical association of RBL with ASH2R and WDR5a in the onion nuclei suggests that the ASH2R-RBL-WDR5a complex acts in the nucleus . Recently , it has been reported that a transiently expressed fusion protein of ASH2R with GFP ( for Green Fluorescent Protein ) in onion cells is localized into the nucleus [38] . Using an ASH2R-GFP transgenic line , we confirmed that the ASH2R fusion protein was indeed localized specifically in the nucleus ( Figure S8 ) . Together , these results are consistent with that the ASH2R subcomplex functions as a transcriptional regulator . The human Ash2-RbBP5-WDR5 core subcomplex associates with MLL1 , MLL2 and other H3K4 methyltransferases to form distinct catalytic complexes that deposit both di- and tri-methyl H3K4 [9] , [10] , [41] , [42] . In Arabidopsis genome , there are may be up to ten putative H3K4 methyltransferases [17] . Recently we have found that WDR5a can associate with ATX1 [24] . Using yeast two-hybrid approach , we further explored whether WDR5a could associate with other putative H3K4 methyltransferases including SDG14 and SDG16 . Full-length SDG14 and SDG16 proteins fused to the GAL4 DNA-Binding Domain ( BD ) were co-expressed with WDR5a fused to the GAL4 Activation Domain ( AD ) in yeast , and subsequently , we found that both SDG14 and SDG16 physically interacted with WDR5a ( Figure 8A and 8C ) . Next , using BiFC we examined whether WDR5a could interact with SDG14 and SDG16 in plant cells . Upon simultaneous expression of WDR5a-cEYFP with nEYFP-SDG14 or nEYFP-SDG16 in onion epidermal cells , fluorescence was observed in the nuclei ( Figure 8B and 8D ) . Hence , WDR5a indeed can physically interact with SDG14 and SDG16 in plant cells . In addition , we found that neither RBL nor ASH2R interacted with SDG14 , SDG16 or ATX1 ( Figure S9 ) . Based on these results , we infer that the ASH2-RBL-WDR5a subcomplex associates with different H3K4 methyltransferases via WDR5a to form multiple functional COMPASS-like H3K4 methyltransferase complexes in Arabidopsis . In yeast and mammal cells , Ash2-containing COMPASS and COMPASS-like complexes catalyze di- and tri-methylation of H3K4 [8] , [9]; removal of BRE2 ( the Ash2 homolog ) function in yeast causes a great reduction in genome-wide H3K4me2 and H3K4me3 [8] . Recent studies reveal that the deficiency of MLL2 ( the H3K4 methyltransferase of MLL2 COMPASS-like complex ) in the mouse oocytes leads to a reduction in global H3K4me2 and H3K4me3 [42] . As described above , the Arabidopsis ASH2R core subcomplex can associate with multiple putative H3K4 methyltransferases to form catalytic COMPASS-like complexes . We sought to investigate the role of Arabidopsis ASH2R-COMPASS in H3K4 methylation . First , genome-wide H3K4 methylation was examined upon loss of ASH2R function . Total histones were extracted from seedlings of wildtype Col and the strong hypomorphic ash2r-2hyp and -3hyp , and levels of monomethyl H3K4 ( H3K4me1 ) , H3K4me2 and H3K4me3 were measured by western blotting with antibodies specifically recognizing these modifications . We found that H3K4me3 , but surprisingly not H3K4me2 , was strongly reduced upon loss of ASH2R function ( Figure 9A ) . In addition , H3K4 monomethylation was not affected in ash2r-2hyp or -3hyp compared to the wildtype ( Figure 9A ) . Next , we examined global H3K4 methylation in ASH2R-overexpression seedlings including ASH2Rox-2 and -3 , and found that in opposite to loss of ASH2R function , H3K4me3 levels were increased in both ASH2Rox-2 and -3 lines , whereas levels of H3K4me1 and H3K4me2 remained unchanged compared to parental Col ( Figure 9B , 9C ) . Together , these results suggest that in Arabidopsis , the ASH2R-COMPASS complexes are responsible for H3K4 trimethylation , but not for di- or mono-methylation of H3K4 . To investigate whether ASH2R directly interacted with its target the FLC locus to activate its expression , we performed chromatin immunoprecipitation ( ChIP ) with anti-FLAG using the wildtype-like seedlings of ASH2R:FLAG-1a ( T3 ) in which FLC expression was only moderately suppressed ( FLC transcript levels in ASH2R:FLAG-1a WT-like seedlings were about 60% of those in Col ) . ASH2R was enriched around the FLC TSS ( FLC-P ) , but not in a distal region upstream of the TSS ( FLC-U ) or in the middle of FLC ( FLC-M ) ( Figure 10B ) . Thus , we infer that the ASH2R core subcomplex directly binds to FLC chromatin to regulate its expression . Next , we investigated the effect of ASH2R knockdown on di- and tri-methylation of H3K4 in FLC chromatin in ASH2R RNAi seedlings by ChIP . Previously , it has been shown that H3K4 trimethylation predominantly occurs around the TSS in the actively transcribed FLC locus [13] , [21] . We found that upon ASH2R knockdown , H3K4me3 levels were reduced in the FLC TSS region ( Figure 10C ) , consistent with the ASH2R binding to this region . Furthermore , we found that levels of H3K4me2 in FLC chromatin remained unchanged upon ASH2R knockdown ( Figure 10D ) . Hence , ASH2R is required for H3K4me3 , but not for H3K4me2 in FLC chromatin , in line with that ASH2R mediates only genome-wide trimethylation , but not dimethylation of H3K4 . We further examined the association of ASH2R with the MAF4 and MAF5 loci , and found that ASH2R bound to the chromatin of 5′ transcribed regions of both MAF4 and MAF5 , but not to the 5′ region of MAF3 that is located immediately upstream of MAF4 ( Figure 10B ) , consistent with that ASH2R promotes the expression of MAF4 and MAF5 , but not MAF3 . Furthermore , ChIP assays show that levels of H3K4me3 in both MAF4 and MAF5 in their 5′ transcribed regions were strongly reduced , whereas H3K4me2 levels remained unchanged in both genes in ASH2R RNAi seedlings relative to Col ( Figure 10C , 10D ) . Thus , ASH2R directly mediates H3K4 trimethylation not only in FLC , but also in MAF4 and MAF5 . As described in Introduction , H3K4 trimethylation is a chromatin mark for actively transcribed eukaryotic genes , but whether H3K4me3 can cause transcriptional activation is unclear . We have found that ASH2R overexpression causes activation of FLC , MAF4 and MAF5 expression . Increased ASH2R expression is expected to increase the availability of ASH2R protein for assembly of ASH2R-COMPASS complexes . It was of great interest to determine whether ASH2R overexpression would cause elevated H3K4 trimethylation in FLC , MAF4 and MAF5 , consequently leading to their activation . First , we performed ChIP experiments to examine tri-methyl H3K4 levels in these loci in Col and ASH2Rox-2 seedlings . At the FLC locus , levels of H3K4me3 in FLC-P , but not in FLC-U , were increased in ASH2Rox-2 relative to Col ( Figure 11A ) . Interestingly , H3K4me3 levels in FLC-M were also increased upon elevated ASH2R expression ( Figure 11A ) . In addition , we found that H3K4me3 levels in MAF4 were increased in ASH2Rox-2 seedlings relative to Col ( Figure 11A ) . At the MAF5 locus , a strong increase of H3K4me3 was observed upon ASH2R overexpression ( Figure 11A ) . Noteworthily , upon ASH2R overexpression in ASH2Rox-2 seedlings , a two-fold increase of H3K4me3 in MAF5 chromatin appears to cause an around 30-fold increase in MAF5 mRNA levels ( Figure 5D and Figure 11A ) , indicating that at the MAF5 locus , H3K4 trimethylation state exerts a strong effect on its activation . We further examined H3K4 dimethylation state in FLC , MAF4 and MAF5 . Consistent with that the ASH2R core subcomplex is required for genome-wide H3K4 trimethylation , but not for H3K4 dimethylation , levels of H3K4me2 in these three loci remained unchanged in ASH2Rox-2 seedlings compared to Col ( Figure 11B ) . Thus , consistent with ASH2R's role for H3K4 trimethylation , ASH2R overexpression leads to increased levels of H3K4me3 , but not H3K4me2 , in its target genes . To further explore the role for ASH2R in the promotion of H3K4 trimethylation in its target genes , we carried out timed activations of ASH2R expression using a transgenic line ( in the Col background ) harboring a stringent two-component glucocorticoid-inducible transgene expression system that has been widely used for gene expression induction in plants [40] , [43] . In this line ( harboring pOp6-ASH2R;p35S-GR:LhG4 ) , the expression of ASH2R transgene ( free of mutations ) was directed by the pOp6 promoter ( Figure S10A ) , that is recognized by the trans-acting transcriptional activator GR:LhG4 ( GR for Glucocorticoid Receptor ) . Upon binding of GR to glucocorticoids such as dexamethasone ( DEX ) , GR:LhG4 is expected to specifically binds to pOp6 to turn on ASH2R expression . As noted earlier , we observed leaky expression of the point-mutated ASH2R transgene , likely caused by proximity of the 35S promoter driving GR:LhG4 expression; hence , we re-oriented p35S to maximize its distance to pOp6 ( Figure S10A ) . Because ASH2R expression levels exert a much stronger effect on MAF5 expression than on FLC or MAF4 expression ( Figure 5C , 5D ) , we followed MAF5 induction upon DEX-activated ASH2R expression as described next . We applied DEX to 7-d-old seedlings , and found that with a single application , MAF5 expression was slightly induced in 72 hour ( h ) ( Figure S10C ) . Next , we applied DEX twice ( the second application occurred 36h after the initial treatment ) ; ASH2R expression was strongly activated upon DEX applications ( still up to about 20 fold in 96h; see Figure 11C ) . Following ASH2R expression activation , MAF5 expression in shoot apices including newly emerged leaves , was induced to 2 . 0 fold over the mock in 72h and about 4 . 0 fold in 96h ( Figure 11D ) . In addition , we found that MAF5 expression was not induced in the first pair of rosette leaves even though ASH2R expression was highly activated upon DEX applications ( Figure S10D-S10E ) , suggesting that cell division activity is required for MAF5 induction . Interestingly , it took at least about 48h to induce MAF5 expression upon ASH2R expression activation ( Figure 11D ) . Arabidopsis shoot apical meristematic cells divide once in 1–2 d [44] . It is likely that one to two rounds of cell division ( or DNA replication ) in the shoot apices may be required for MAF5 induction by ASH2R . We performed ChIP experiments to examine H3K4 methylation state in MAF5 chromatin at 48h and 96h after the initial DEX application . Levels of H3K4me3 , but not H3K4me2 in MAF5 , increased in 96h , consistent with the induction of MAF5 expression ( Figure 11E , 11F ) . Thus , the timed ASH2R induction causes increased H3K4me3 in MAF5 chromatin and turns on MAF5 expression . These findings together with the elevated levels of H3K4me3 in FLC , MAF4 and MAF5 upon ASH2R overexpression , strongly suggest that increased H3K4 trimethylation in these loci causes activation of their expression .
The human COMPASS-like complexes contain three structural core components , one H3K4 methyltransferase and several non-conserved components [2]; the three core components together with the methyltransferase form a functional ( catalytic ) core complex for H3K4 methylation [10] . In this study , we have found that the Arabidopsis ASH2R , RBL and WDR5a , homologs of the three structural core components of the COMPASS-like complexes , form a core subcomplex during vegetative and reproductive development . Although there are two WDR5 homologs in Arabidopsis , our study suggests that Arabidopsis only has a single core subcomplex for H3K4 methylation , namely ASH2R-RBL-WDR5a . The biochemical and biological functions of WDR5b remain elusive . Our phylogenetic analyses show that in contrast to most animals , there are multiple WDR5 homologs in land plants . Biochemical functions of these proteins cannot be directly inferred simply based on the homology to WDR5 , and have to be determined experimentally . In Arabidopsis there are may be up to 10 putative H3K4 methyltransferases based on similarity of the catalytic SET domains to the known H3K4 methyltransferases in yeast and animals . We have found that WDR5a can associate with multiple putative H3K4 methyltransferases including ATX1 , SDG14 and SDG16 . Thus , we infer that multiple COMPASS-like H3K4 methyltransferase complexes exist in the higher plant Arabidopsis , for instance , the ASH2R-RBL-WDR5a-ATX1 complex . Of note , this ASH2R-ATX1 complex might contain other co-components beside the four-component catalytic core complex . These COMPASS-like complexes are expected to methylate H3K4 in various target genes and control multiple developmental processes including leaf growth and development , the floral transition and seed development . The ASH2R-ATX1 COMPASS-like complex activates FLC expression to repress the floral transition in the rapid-cycling accession Col . Recently it has been shown that the ATX1 protein binds to FLC chromatin [21]; in this study , we have found that ASH2R , like ATX1 , directly interacts with the FLC locus . These findings are consistent with that these two proteins are part of the ASH2R-ATX1 complex binding to FLC chromatin to activate FLC expression . Our genetic analyses show that loss-of-function ash2r mutants , similar to atx1 , cause a great reduction in FLC expression and consequent early flowering . Furthermore , we have found that knockdown of either WDR5a or RBL expression , like ash2r and atx1 , leads to reduced FLC expression and early flowering . These genetic findings are consistent with that WDR5a and RBL are part of the ASH2R-ATX1 COMPASS-like complex that activates FLC expression . FLC is expressed at a level relatively low in the Col background ( the wild-type accession used in this study and also commonly used in other Arabidopsis studies ) , compared to FRIGIDA ( FRI ) -containing accessions ( note that FRI is mutated in Col ) [45] . FRI , encoding a coiled-coil domain protein , upregulates the expression of FLC to a higher level that significantly delays flowering [45] . Previously , it has been shown that ATX1 is required for FRI-dependent FLC upregulation [21] . We have recently revealed that a functional FRI causes an increase in the amount of WDR5a protein bound to FLC chromatin , resulting in elevated H3K4 trimethylation in FLC [24] . WDR5a knockdown strongly suppresses FRI-dependent FLC upregulation , but not FLC upregulation/de-repression upon loss of FLD function ( note that FLD , a putative H3K4 demethylase , represses FLC expression ) [24] . Similarly , we have observed that ASH2R knockdown strongly suppresses FRI-dependent FLC upregulation , but not FLC de-repression in fld mutants ( data not shown ) . Given the association of WDR5a with ATX1 and ASH2R ( via RBL ) , these findings suggest that in the presence of a functional FRI , the ASH2R-WDR5a-ATX1 COMPASS-like complex is further enriched at FLC chromatin to upregulate FLC expression . Recently it has been reported that a histone methyltransferase known as EFS , that can methylate both H3K4 and H3K36 in vitro , is required for FRI recruitment to the FLC locus [46] . This recruitment is expected to cause the enrichment of ASH2R-WDR5a-ATX1 complex at FLC , which may function in concert with EFS leading to H3K4 and H3K36 methylation at FLC and FLC upregulation in the FRI background . We and others have found that a null ASH2R/TRO lesion causes arrested embryogenesis at globular stage [see Figure S3 and [38]]; in addition , we found that ASH2R forms the core subcomplex with RBL and WDR5a in developing siliques and is required for genome-wide H3K4 trimethylation . These findings , together with the strong expression of ASH2R in developing seeds ( Figure 6C , 6D ) , lead us to conclude that the ASH2R core subcomplex ( presumably ASH2R-COMPASS ) mediates H3K4 trimethylation in seed chromatin to control seed development . Interestingly , we noticed that knockdown of RBL or WDR5a , like ASH2R knockdown , does not disrupt seed development . There are two possible explanations . Firstly , the dsRNAi targeting RBL , ASH2R or WDR5a driven by the 35S promoter only partially suppresses target gene expression , and the remaining transcripts may be sufficient for proper embryogenesis . Secondly , WDR5a , ASH2R or RBL expression in the RNAi lines may not be affected in early embryogenesis because the 35S promoter has been shown to be inactive in the early seed development [47] . In yeast , the intact COMPASS complex is required for di- and tri-methylation of H3K4 [8] . Mammalian Ash2-COMPASS complexes are capable of catalyzing di- and tri-methylation of H3K4 [9] , [41]; for instance , MLL2 deficiency in mouse oocytes causes a genome-wide reduction in H3K4me2 and H3K4me3 [42] . In mammals , both di- and tri-methylation of H3K4 are associated with actively transcribed genes [5] , whereas recent genome-scale studies have revealed that only H3K4me3 , but not H3K4me2 , is correlated with active transcription and implicated in transcriptional activation in Arabidopsis [6] . Because H3K4me3 and H3K4me2 have distinct distribution patterns in Arabidopsis genome , they are expected to be deposited by different players . In this study , we have found that loss of ASH2R function causes a strong global reduction in H3K4me3 , but not in H3K4me2 or H3K4me1 in Arabidopsis , whereas ASH2R overexpression leads to a genome-wide increase in H3K4me3 , but not in H3K4me2 or H3K4me1 . Furthermore , we have revealed that ASH2R knockdown causes a strong reduction in H3K4me3 , but not in H3K4me2 , whereas ASH2R overexpression leads to an increase in H3K4me3 , but not in H3K4me2 , in its direct target genes FLC , MAF4 and MAF5 to activate their expression . These findings further support the notion that the Arabidopsis ASH2R-COMPASS complexes specifically deposit H3K4me3 to promote target gene expression , providing a molecular explanation for the observed coupling of H3K4me3 ( but not H3K4me2 ) with active gene expression in Arabidopsis . This indicates a difference in the role of Ash2-COMPASS in H3K4 methylation between Arabidopsis and yeast/mammals . Both constitutive overexpression of ASH2R and the timed chemical induction of ASH2R expression give rise to increased H3K4me3 in its target genes such as MAF5 , and target gene activation . Interestingly , overexpression of WDR5a and RBL by the constitutive 35S promoter , unlike ASH2R overexpression , did not give rise to any noticeable phenotypes ( data not shown ) . Together , these observations indicate that the availability of ASH2R protein is a rate-limiting factor in Arabidopsis H3K4 trimethylation . As noted in Introduction , H3K4 trimethylation is a chromatin mark for actively expressed genes in eukaryotic organisms that so far have been examined , but whether it can activate gene expression has been under much debate [see [16]] as the accumulation of H3K4me3 ( predominantly in the 5′ transcribed regions ) in actively transcribed chromatin may merely result from active transcription . In this study , we have found that the ASH2R core subcomplex , presumably ASH2R-COMPASS , binds to FLC , MAF4 and MAF5 chromatin , and that ASH2R knockdown leads to a reduction in H3K4me3 and suppresses the expression of these three genes . In addition , ASH2R overexpression causes increased H3K4me3 and activation of these gene expression . Furthermore , we have found that timed chemical induction of ASH2R expression causes increased H3K4 trimethylation in MAF5 ( a direct target of ASH2R ) in shoot apices and turns on its expression . These results strongly suggest that H3K4 trimethylation in FLC , MAF4 and MAF5 can activate their expression , providing concrete evidence for the notion that H3K4 trimethylation can activate eukaryotic gene expression . Ash2-COMPASS complexes are expected to be actively recruited to target gene chromatin to catalyze H3K4 trimethylation; subsequently , the evolutionarily conserved ATP-dependent chromatin remodeling factors such as NURF in human that recognizes and binds to H3K4me3 [11] , may be recruited to target genes to actively mobilize/remodel nucleosomes , resulting in transcriptional activation or active gene expression .
All of the Arabidopsis thaliana lines used in this study were in the Col background . The ash2r-1 allele ( SAIL_851_H01 ) was isolated from the SAIL collection [39] . Plants were grown under cool white fluorescent lights in long days ( 16h light/8h dark ) . The Matchmaker GAL4 Two-Hybrid System 3 ( Clontech ) was adapted for the yeast two-hybrid assay . The full-length coding sequences for the tested genes were cloned into the pGADT7 and/or pGBKT7 vectors ( Clontech ) . The experiments were performed according to the manufacturer's instructions using the strain AH109 ( Clontech ) . To test the interactions , yeast cells were spotted on the highly selective SD media lacking of leucine , tryptophan , histidine and adenine . The full-length coding sequences for WDR5a , ASH2R , RBL , SDG14 and SDG16 were translationally fused with either an N-terminal EYFP fragment in the pSAT1A-nEYFP-N1/pSAT1-nEYFP-C1 vectors and/or a C-terminal EYFP fragment in the pSAT1A-cEYFP-N1/pSAT1-cEYFP-C1-B vectors ( www . bio . purdue . edu/people/faculty/gelvin/nsf/index . htm ) . Onion epidermal cells were transiently co-transformed by appropriate plasmid pairs using the Helium biolistic gene transformation system ( Bio-Rad ) following the manufacturer's instructions . Within 24–48 hrs after bombardment , YFP fluorescence was observed and imaged using a Zeiss LSM 5 EXCITER upright laser scanning confocal microscopy ( Zeiss ) . For the examination of seed development , seeds were cleared in 8∶1∶3 ( W/V/V ) chloral hydrate:glycerol: water for 1–2 hrs . The cleared seeds were examined using differential-interference-contrast optics on a Leica DM4500B microscope , and images were acquired with a Nikon DXM1200F digital camera . To knock down RBL expression , two copies of a 245-bp RBL specific fragment ( from +1853 to +2097 of the RBL cDNA; TSS as +1 ) were inserted into the pB7GWIWG2 vector [48] for hairpin RNA production . For ASH2R knockdown , two copies of a 223-bp ASH2R-specific fragment ( from +1456 to +1678 of the ASH2 cDNA; TSS as +1 ) , were inserted into the pB7GWIWG2 vector for hairpin RNA production . For overexpression of ASH2R and the rescue of ash2r-1 mutant , the full-length coding sequence for ASH2R was inserted downstream of the CaMV 35S promoter in the pMDC32 vector [49] via gateway technology ( Invitrogen ) , resulting in the 35S-ASH2R construct . For ASH2R:FLAG construction , the full-length ASH2R coding sequence except the stop codon was first fused in frame with a FLAG tag ( three copies ) , and subsequently the ASH2R-FLAG fragment was inserted downstream of the 35S promoter in the pMDC32 vector via gateway technology . For ASH2R subcellular localization , the full-length ASH2R coding sequence except the stop codon was inserted between the 35S promoter and GFP in the pMDC85 vector [49]; the ASH2R coding sequence was in frame with the downstream GFP reporter gene . For ASH2R-GUS construction , an 1816-bp ASH2R genomic fragment ( from −1537 to +279; A of the start codon as +1 ) including a 1537-bp native promoter and a 279-bp genomic coding region was inserted upstream of the GUS reporter gene in the pMDC162 vector [49]; the genomic coding sequence was in frame with GUS . To construct RBL-GUS , we inserted a 1665-bp RBL genomic fragment ( from −1081 to +584; A of the start codon as +1 ) upstream of the GUS reporter gene in pMDC162; the genomic coding sequence of RBL was in frame with GUS . Total RNAs were extracted from aerial parts of 10-d-old seedlings or DEX-treated tissues as described previously [50] . The total RNAs were subsequently used as templates to synthesize cDNAs by reverse transcription . Real-time quantitative PCR was carried out on an ABI Prism 7900HT sequence detection system as previously described [24] , [50] . Primers used to amplify the cDNAs of FLC , FLM , MAF2-MAF5 and TUB2 ( At_5g62690 ) have been previously described [29] , [50] . The primer pair , 5′-AGGAAGGGTACAAGGAAGGTGATG-3′ and 5′-AACGATATTTCACTGCCTGGTACAAC-3′ , was used for ASH2R amplification; the primer pair , 5′-CGAAGATGAATTTGATTTGATACCTG-3′ and 5′-TGTCTCACCCATTTCTTCTGCTTGT-3′ , was used to amplify RBL cDNAs . Each sample was quantified in triplicate and normalized to the endogenous control TUB2 . Bars indicate standard deviations of triplicate measurements . Plant tissues were fixed in 80% acetone at −20°C for 1 hr , washed by a staining buffer ( 5 mM EDTA pH 8 . 0 , 0 . 05% Triton X-100 , 2 mM potassium ferrocyanide , 2 mM potassium ferricyanide , 100 mM NaH2PO4 and 100 mM Na2HPO4 ) , and incubated in the staining buffer with 0 . 5-mg 5-bromo-4-chloro-3-indolyl-β-d-glucuronic acid ( X-Gluc ) at 37°C for 4 to 12 hrs . The ASH2R-GUS seedlings were stained for 4 hrs , and other stainings were performed for 12 hrs . Seeds were cleared after staining for observation under a Leica DM4500B microscope . Immunoprecipitation experiments were performed as described previously [51] with minor modifications . Briefly , about 0 . 5-g 10-d-old seedlings were harvested and ground in liquid nitrogen , and subsequently , total proteins were extracted . 1 . 0-ml protein extracts were incubated with 60-µl slurry of anti-FLAG M2 affinity gel ( Sigma , Cat#: A2220 ) , and the immunoprecipitated proteins were washed three times and subsequently boiled in the SDS-PAGE loading buffer , followed by western blotting with anti-FLAG ( Sigma , Cat#: A8592 ) or anti-WDR5 [Abcam , Cat#: ab75439 ( Batch #: 639695 ) ; note that this antibody recognizes WDR5a; see [24]] . In the immunoprecipitation experiments with 5-DAP siliques , the samples were cross-linked with 1% formaldehyde for 30 minute before protein extraction as described previously [51] . Histone protein extraction and western analysis were performed as previously described [29] . Briefly , total histones were extracted from 10-d-old seedlings , separated in an SDS-PAGE gel , and transferred to a 0 . 2-µl nitrocellulose membrane ( Bio-Rad ) . The protein blots were probed with anti-monomethyl histone H3K4 ( Abcam , Cat#: ab8895 ) , anti-dimethyl H3K4 ( Millipore , Cat#: 07-030 ) , anti-trimethyl H3K4 ( Millipore , Cat#: 05-745R ) and anti-histone H3 ( Millipore , Cat#: 07-690 ) . ChIP experiments were performed with 10-d-old seedlings or DEX-treated shoot apices including newly emerged leaves as previously described [52] . Immunoprecipitations were carried out using Rabbit polyclonal anti-dimethyl H3K4 ( Millipore , Cat#: 07-030 ) , anti-trimethyl H3K4 ( Millipore , Cat#: 05-745R ) or anti-FLAG ( Sigma , Cat#: A8592 ) . The real-time quantitative PCR and the primers used to amplify FLC-U , FLC-P , FLC-M and TUB2 were described previously [29] , [50] . The primer pair , 5′-CGGCGAGTTATGCAGACATCACA-3′ and 5′-GTGGCAGAGATGATGATAAGAGCG A-3′ , was used to amplify MAF4; the primer pair , 5′-CAGGATCTCCGACCAGTTTATACAGAC-3′ and 5′-GAGGAGTTGTAGAGTTTGCCGGT-3′ , was used to amplify MAF5 . The MAF3 region was amplified by the primer pair: 5′-GTCTAGCCCAAAAGAAGAAGATAGAAACG-3′ and 5′-GGAGGCAGAGTCGTAGAGTTTTCC-3′ . The relative fold changes are the averages of two independent biological repeats . The initial treatments were carried out by applying 10 µM DEX plus 0 . 015% Silwet L-77 to 7-d old seedlings of Col and the pOp6-ASH2R;p35S-GR:LhG4 line grown in long days ( the second treatment was carried out 36h after the initial treatment ) . In the mock control , DEX was omitted . The hour of initial DEX application was designated as 0h . Shoot apices including newly emerged leaves or the first pair of rosette leaves were collected at 0 , 24 , 48 , 72 and 96h for RNA analyses . For ChIP experiments , shoot apices with newly emerged leaves were collected at 0 , 48 and 96h .
|
Histones can be covalently modified and histone modifications regulate chromatin structure and gene transcription . One such modification is histone H3 lysine-4 ( H3K4 ) methylation , which can be mono- , di- , or tri-methylated . In animals such as fruitfly and mammals , both di- and tri-methylation of H3K4 are associated with active gene expression . In contrast to animals , in the flowering plant Arabidopsis only H3K4 trimethylation has been implicated in gene transcriptional activation . H3K4 methylation is catalyzed by the H3K4 methyltransferase complexes known as COMPASS-like in mammals . Here , we report that COMPASS-like H3K4 methyltransferase complexes exist in Arabidopsis . Loss of function of a core complex protein causes a great decrease in Arabidopsis genome-wide H3K4 trimethylation , but not in di- or mono-methylation . Our analyses of several direct target genes of these COMPASS-like complexes show that they mediate deposition of trimethyl but not dimethyl H3K4 in these loci to activate their expression , providing concrete evidence for the notion that H3K4 trimethylation accumulation can activate eukaryotic gene expression . Furthermore , our findings provide a molecular explanation for the observed coupling of trimethylation but not dimethylation of H3K4 with active gene expression in Arabidopsis . In addition , we found that H3K4 trimethylation regulates leaf growth and development , flowering , and embryo development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"biology/plant",
"growth",
"and",
"development",
"plant",
"biology/plant",
"genetics",
"and",
"gene",
"expression",
"genetics",
"and",
"genomics/plant",
"genetics",
"and",
"gene",
"expression",
"genetics",
"and",
"genomics/gene",
"expression"
] |
2011
|
Arabidopsis COMPASS-Like Complexes Mediate Histone H3 Lysine-4 Trimethylation to Control Floral Transition and Plant Development
|
Burkholderia pseudomallei , the causative agent of melioidosis , is an important public health threat due to limited therapeutic options for treatment . Efforts to improve therapeutics for B . pseudomallei infections are dependent on the need to understand the role of B . pseudomallei biofilm formation and its contribution to antibiotic tolerance and persistence as these are bacterial traits that prevent effective therapy . In order to reveal the genes that regulate and/or contribute to B . pseudomallei 1026b biofilm formation , we screened a sequence defined two-allele transposon library and identified 118 transposon insertion mutants that were deficient in biofilm formation . These mutants include transposon insertions in genes predicted to encode flagella , fimbriae , transcriptional regulators , polysaccharides , and hypothetical proteins . Polysaccharides are key constituents of biofilms and B . pseudomallei has the capacity to produce a diversity of polysaccharides , thus there is a critical need to link these biosynthetic genes with the polysaccharides they produce to better understand their biological role during infection . An allelic exchange deletion mutant of the entire B . pseudomallei biofilm-associated exopolysaccharide biosynthetic cluster was decreased in biofilm formation and produced a smooth colony morphology suggestive of the loss of exopolysaccharide production . Conversely , deletion of the previously defined capsule I polysaccharide biosynthesis gene cluster increased biofilm formation . Bioinformatics analyses combined with immunoblot analysis and glycosyl composition studies of the partially purified exopolysaccharide indicate that the biofilm-associated exopolysaccharide is neither cepacian nor the previously described acidic exopolysaccharide . The biofilm-associated exopolysaccharide described here is also specific to the B . pseudomallei complex of bacteria . Since this novel exopolysaccharide biosynthesis cluster is retained in B . mallei , it is predicted to have a role in colonization and infection of the host . These findings will facilitate further advances in understanding the pathogenesis of B . pseudomallei and improve diagnostics and therapeutic treatment strategies .
B . pseudomallei , an environmental saprophyte , is the etiological agent of melioidosis and has been traditionally described as being endemic to Northern Australia and Southeast Asia [1] . However , an increasing body of evidence indicates B . pseudomallei is more widely distributed than previously thought [2–4] . As diagnostics and clinical awareness improve , melioidosis cases and their bacterial cause are increasingly detected worldwide [5] . B . pseudomallei is an important global pathogen , as indicated by a recently published study that predicts approximately 165 , 000 human cases of melioidosis with greater than 50% mortality annually in 79 countries where the pathogen is probably endemic [6] . Due to the lack of vaccines , the intrinsic resistance to numerous antibiotics , and high mortality rate associated with acute infections , in addition to its potential use as an agent for biological warfare and bioterrorism , B . pseudomallei is currently designated as a Tier 1 select agent by regulatory agencies in the United States [7 , 8] . B . pseudomallei is well known for its ability to produce biofilm , which may be critical to the increased persistence of this pathogen in the environment [9] . Bacteria growing as a biofilm are embedded in a matrix comprised of self-produced extracellular polymeric substances ( EPS ) that include polysaccharides , proteins , lipids , and nucleic acids . This matrix is thought to serve as a scaffold to hold biofilm cells together and protect from some antimicrobials ( see [10] for recent review ) . Despite the importance of the EPS components that comprise the biofilm matrix , we know surprisingly little about it [11] . EPS from B . pseudomallei has been described for capsular polysaccharides , O-polysaccharides , and exopolysaccharides ( for review [12] ) . However , the contribution and characterization of these EPS components to B . pseudomallei pathogenesis has not been conclusively evaluated in chronic models of melioidosis . Additional capsular polysaccharides and exopolysaccharides also remain to be identified and characterized . In the absence of information linking the identity , structural composition , and expression of these EPS components , it will not be possible to determine their role in the establishment and progression of disease . Multiple polysaccharides associated with the surface of B . pseudomallei have been characterized based on structure and antigenicity . Two of the best characterized EPS components are the primary capsule ( CPSI , Bp1026b_I0499-Bp1026b_I0524 ) [13 , 14] and >150 kDa acidic exopolysaccharide [15 , 16] . Additional biosynthetic clusters have also been identified that are predicted to encode three additional capsules ( CPSII-IV ) [12] . However , the composition , structure , and role during pathogenesis is not well understood for all of these polysaccharides . Capsule III gene expression has been shown to be increased in water as compared to relatively low levels of expression in vivo , which is proposed to contribute to the survival of B . pseudomallei in the environment [17] . However , the role of this capsule in B . pseudomallei is unknown . CPSI was originally described as O-antigenic polysaccharide ( O-PS I ) and is an unbranched homopolymer consisting of monosaccharide repeats having the structure [→3 ) -2-O-acetyl-6-deoxy-β-D-manno-heptopyranose- ( 1→] [13 , 14] . The structure of an acidic exopolysaccharide has also been reported to be a unique linear tetrasaccharide repeating unit consisting of three galactose residues and one 3-deoxy-D-manno-2-octulosonic acid ( Kdo ) residue [15 , 16] . A number of published studies have identified genes involved in the production of the B . pseudomallei biofilm matrix [18–24] . However , the contribution of these genes to pathogenesis in B . pseudomallei has been complicated by the use of multiple strains and different genetic approaches , which has resulted in conflicting reports of the role of key biofilm matrix components [25–27] . To gain a more comprehensive understanding of the genes that contribute to biofilm formation , we screened a sequence-defined two-allele library of transposon mutants comprising approximately 81% coverage of ORFs in B . pseudomallei 1026b , which is a clinical isolate from a diabetic patient afflicted with disseminated melioidosis [28] . This strain has become a model strain for B . pseudomallei studies of pathogenesis and antibiotic resistance because the genome is fully sequenced , publicly available , amenable to genetic analysis , and naturally transformable [28–30] . Numerous animal models have also been developed to study acute and chronic disease associated with melioidosis [10 , 31–34] . In this systematic analysis of genes that contribute to biofilm production , we identified 59 transposon insertion mutants in unique genetic loci that have an integral role in B . pseudomallei 1026b biofilm formation . These loci encode polysaccharide biosynthesis , fimbriae , motility , cellular homeostasis , transport , and hypothetical genes . One of the key EPS components discovered in this study is synthesized by a novel 28 kb biosynthesis gene cluster ( Bp1026b_I2907-Bp1026b_I2927 ) , which we have designated as becA-R ( biofilm exopolysaccharide gene cluster ) . In addition to identifying these genetic loci as requirements for biofilm formation , we evaluated transposon insertion mutants in B . pseudomallei genes previously described to contribute to biofilm formation . This is the first report that describes the multiple genetic components that contribute to biofilm formation on a genome-wide scale in B . pseudomallei .
All experiments were performed in the BSL3 facility at Colorado State University except for studies conducted with the select agent excluded B . pseudomallei strain Bp82 , which was handled at BSL2 . Transposon ( T24 ) mutant derivatives described in these studies ( S1 Table ) were generated during the production of a comprehensive two-allele sequence defined transposon mutant library of B . pseudomallei 1026b ( manuscript in preparation ) . Briefly , the B . pseudomallei 1026b two-allele library contains two mutants per gene for which the transposon locations were confirmed by resequencing . Two different representative mutants were chosen from the primary library with transposon insertion sites between 5% and 80% of the respective predicted open reading frame , and for which the precise transposon-genome junctions have been determined by sequencing . T24 is a Tn5-derived transposon containing a select agent approved kanamycin resistance selection marker that was constructed in the laboratory of Colin Manoil ( University of Washington ) ( http://www . gs . washington . edu/labs/manoil/transposons/transposons . pdf ) . Transposon mutants in the primary biofilm screen were grown in 1 . 2 mL LB with 10% glycerol and 35 μg/mL kanamycin . For all subsequent assays , overnight cultures of selected transposon mutants were grown in LB ( 10 g/L tryptone , 5 g/L yeast extract , and 5 g/L NaCl ) with 300 μg/mL kanamycin . Location of the transposon insertion was reconfirmed by sequencing . Swimming motility and growth curve assays were conducted as previously described [35] . For growth on NAP-A plates [36] , overnight cultures were either pin replicated or spotted ( 3 μL ) and incubated at 37°C for two days . All strains and plasmids are described in S2 Table . The primary biofilm screen was conducted using deep 96 well plates ( Simport #T110-10S ) . Plates were inoculated with a 96-well pin replicator and statically incubated at 37°C for two days . Transposon insertional mutants that visually appeared to have reduced or no pellicle formation were selected and plated on LB kanamycin plates for colony isolation . Biofilm phenotypes were further evaluated in static microtiter biofilm assays as previously described [35] . All transposon insertions were confirmed via Sanger sequencing . The DOOR 2 . 0 operon database was used to identify the first gene in each operon using B . pseudomallei K96243 as the reference genome [37] . A combination of bioinformatics tools and open-access genomic databases was used to compare the putative exopolysaccharide gene clusters from the sequenced genomes of B . pseudomallei 1026b ( taxid: 884204 ) , B . cenocepacia J2315 ( taxid: 216591 ) , B . vietnamiensis G4 ( taxid: 269482 ) , B . mallei ATCC 23344 ( taxid: 243160 ) , and B . thailandensis E264 ( taxid: 271848 ) . Regions of homology were initially identified using BLASTN ( BLAST , NCBI ) using default parameters and the Burkholderia Orthologous Groups classification system from the Burkholderia Genome Database ( http://www . burkholderia . com , [38] ) . Genome sequences for B . pseudomallei 1026b chromosome I ( accession number: NC_017831 . 1 ) and B . cenocepacia J2315 chromosome II ( accession number: NC_011001 . 1 ) were downloaded from the GenBank sequence database ( NCBI , NIH ) and regions of interest were extracted using Geneious version 7 . 1 . 7 ( http://www . geneious . com , [39] ) . Comparative analysis of bce-I and bce-II gene clusters was conducted for B . pseudomallei and B . vietnamiensis G4 . Previously published research on cepacian production in several Burkholderia spp . strains has linked polysaccharide production and structural characterization to specific biosynthetic gene clusters that were annotated in the B . vietnamiensis G4 genome [40] . GenBank sequence files were visualized with EasyFig version 2 . 2 . 3 [41] and Python programming language version 2 . 7 ( http://www . python . org ) . Homology and inversions among gene loci were calculated using BLASTN with the EasyFig default parameters of a minimum identity cutoff of 60% . Individual percent identities for each locus were calculated using Multiple Sequence Comparison by Log-Expectation ( MUSCLE ) tool provided by the European Bioinformatics Institute ( EMBL-EBI ) , which creates percent identity matrices using Clustal 2 . 1 [42] . To calculate and visualize sequence homology , we used a threshold E-value of 1e-3 and minimum identity value of 0 . 60 for blast hits drawn . Cut-off thresholds were validated using the command line BLAST+ application to generate a frequency distribution of E-values for all predicted homologous alignments and false-positive non-homologs . Construction of EPS biosynthetic cluster deletions was accomplished by amplification and fusion of flanking genomic sequences external to the region of interest using SOEing PCR and cloning into the allelic exchange vector , pEXKm5 [43] . Introduction of the suicide vector for allelic exchange was accomplished by conjugation of E . coli RHO3 with pEXKm5 constructs into B . pseudomallei [43] . The mutations were verified using internal and external PCR primers to the gene of interest , after counterselection and screening for B . pseudomallei kanamycin-sensitive clones containing the putative deletion mutation . PCR primers were designed using genomic sequence obtained from the Burkholderia Genome Database [38] . SOEing PCR used the following primers: left flank ( 5’- NNCCCGGGCGAACAGGTTGCGCGGACGGT-3’ ) and ( 5’- ACGAACGACGACAGCCGCCGTCCCGCGCGGACCTCAGAAGC-3’ ) and right flank with ( 5’- GCTTCTGAGGTCCGCGCGGGACGGCGGCTGTCGTCGTTCGT-3’ ) and ( 5’- NNNCCCGGGAAGAGCCTCGCGACCGCGCAC-3’ ) to amplify the regions flanking Bp1026b_I2907-Bp1026b_I2927 . Primers incorporated XmaI sites as indicated in bold text . The 1 . 5 kb flanking region was cloned into pEXKm5 and introduced into E . coli RHO3 for allelic exchange in B . pseudomallei 1026b and B . pseudomallei 1026b ΔwcbR-A::FRT-Zeo and into the attenuated B . pseudomallei Bp82 and B . pseudomallei Bp82 ΔwcbR-A::FRT-Zeo [44] . Complementation studies utilized a transposase-mediated integration approach for conditional expression of full length Bp1026b_I1954 , Bp1026b_I2907 , and Bp1026b_II2527 under the control of an inducible Ptac promoter to allow for controlled expression in B . pseudomallei as previously described [35] . Gene expression constructs were introduced into the chromosome using pUC18T-mini-Tn7T-Km-LAC [45] . Full length genes were amplified from genomic B . pseudomallei Bp82 DNA using the following primers: 5’-NNNCCCGGGATGGATTTCGTTTTGCGGG-3’ and 5’-NNAAGCTTTCACTCCGCGTCCCCCTG-3’ for Bp1026b_I1954 , 5’-NNNCCCGGGAGGAGGATATTCATGAATCTGTCTTCCCCGTTATCC-3’ and 5’-NNNAAGCTTTCAATCGAGCGCGCGCA-3’ for Bp1026b_I2907 ( becA ) , and 5’-NNNCCCGGGATGACGCCCGAACGGCCCGACGCTT-3’ and 5’-NNNAAGCTTTCAATCCTCATGCCCCGCGA-3’ for Bp1026b_II2527 with engineered XmaI and HindIII sites ( in bold ) and a ribosomal binding site in italics for Bp1026b_I2907 . Amplified PCR products were cloned into pUC18T-miniTn7T-Km-LAC and sequences were verified by Sanger sequencing . The T24 transposon which contains a kanamycin resistance cassette was removed from I1954::T24 , I2907::T24 ( becA ) , and II2527::T24 using a FLP-mediated recombinase strategy that leaves behind a FRT signature sequence in the disrupted gene . Triparental mating , Tn7 integration , and validation of complementation constructs was done as previously described [35] . Mini-Tn7 insertion was confirmed to be at the glms2 neutral site . Conditional expression of Bp1026b_I1954 , Bp1026b_I2907 , and Bp1026b_II2527 in B . pseudomallei was achieved by adding 1mM isopropyl-β-D-thio-galactopyranoside ( IPTG ) to the growth medium . B . pseudomallei Bp82 cultures were grown in LB supplemented with 80 μg/mL adenine for 16 h . 400 μL of the overnight cultures was spotted onto a polycarbonate 47 mm 0 . 2 μm ( Poretics ) membrane disk placed on NAP-A plates for a total of four membranes per culture and incubated at 37°C for 24 h . The membranes were transferred onto fresh NAP-A plates and incubated at 37°C for another 24 h . Membranes were placed in 50 mL conical tubes containing 20 mL 1X PBS and vortexed for 30 min . Exopolysaccharide extraction was performed as described by Steinmetz et al . with the exception of the initial growth of cells on NAP-A plates [15] . Briefly , cells were centrifuged for 4 h at 20 , 000g at 4°C , and then the supernatant was heated at 80°C for 30 min and centrifuged again for 4 h at 20 , 000g at 4°C . Supernatant was precipitated with 80% ( vol/vol ) ethanol overnight at -20°C . Precipitate was centrifuged for 30 min at 3 , 000g at 4°C , washed with 80% ethanol , centrifuged , and washed with 96% ethanol . Precipitate was solubilized in PBS and treated with RNaseA and DNaseI for 2 . 5 h and then centrifuged for 30 min at 20 , 000g at 4°C . Supernatants were precipitated with 80% ethanol and centrifuged . Precipitate was dissolved in LC-MS grade water . The protein concentrations from harvested cells were quantified with the 660nm protein assay kit ( Pierce ) to estimate biomass prior to exopolysaccharide isolation . Western blot analysis was performed on semi-purified extracts of polysaccharides from the select-agent exempt B . pseudomallei Bp82 . Exopolysaccharides were purified as described above from cultures of Bp82 , Bp82 ΔbecA-R , Bp82 ΔwcbR-A , and Bp82 ΔbecA-R ΔwcbR-A . Exopolysaccharide samples were diluted to normalize loading amounts equivalent to 2 μg of the original biomass based on total protein in samples prior to extraction . To obtain purified B . pseudomallei CPSI , culture media was inoculated with B . pseudomallei RR2683 ( O-polysaccharide mutant; select agent-exempt strain ) and incubated overnight at 37°C with vigorous shaking [14] . Cell pellets were obtained by centrifugation and extracted using a modified hot aqueous-phenol procedure [13] . Purified CPSI was obtained as previously described [14] . Semi-purified exopolysaccharide extracts and purified CPSI were added to Laemmli buffer containing β-mercaptoethanol . Samples were run on 4–15% Criterion gradient gels ( Bio-Rad ) . Gels were transferred to a 0 . 2 μm PVDF membrane ( Bio-Rad ) and blocked in 1X TBST containing 5% ( w/v ) skim milk . Immunoblots were probed with either 1:2 , 000 primary B . pseudomallei-specific CPSl ( 4C4 IgG1 ) [46] or 1:2 , 000 acidic exopolysaccharide ( mAb 3015 ) [15] and detected with goat anti-mouse poly HRP secondary ( 1:50 , 000 ) ( Pierce ) . The immunoblots were visualized using a Clarity Western ECL Blotting Substrate . Images were taken with a Biorad ChemiDoc XRS+ . Carbohydrate analysis of B . pseudomallei polysaccharide extracts was done following a similar approach as previously described with modification [47] . Aliquots of the partially purified B . pseudomallei Bp82 exopolysaccharide extracts and a mixture of monosaccharide standards ( 5 μg each , rhamnose , arabinose , ribose , fucose , mannose , glucose and galactose , 100 μg/mL stock solution in water ) were spiked with 5 μg of internal standard , 3-O-methylglucose ( 100 μg/mL stock solution in water ) and dried under nitrogen without applying any heat . The dried samples were hydrolyzed for 2 h at 120°C with 2M TFA and alditol acetate derivatives of the resulting monosaccharides were generated . After cooling to room temperature , the hydrolysates were dried under a gentle stream of nitrogen to remove TFA . The resultant monosaccharides were reduced with sodium borodeuteride ( 10 mg/mL in 1M ammonium hydroxide-ethanol , 1:1 ) overnight at room temperature . The reaction was terminated with 3–4 drops of glacial acetic acid . The reduction products were dried under nitrogen in the presence of methanol to remove excess borodeuteride and subsequently per-O-acetylated with acetic anhydride at 100°C for 2 h to convert each monosaccharide to its corresponding alditol acetate . The alditol acetates were extracted with a biphasic partition of chloroform and water . The organic phase was dried under nitrogen without applying heat , reconstituted in chloroform and analyzed by GC-MS . The GC-MS analyses of the alditol acetates were performed with a CP 3800 gas chromatograph coupled with an MS3200 mass spectrometer ( Varian Inc . Palo Alto , CA ) . Helium was used as carrier gas at a constant flow of 1 mL/min . The alditol acetates were chromatographically separated on a VF- 5ms column ( 30 mm x 0 . 25 mm i . d . x 0 . 25 μm film thickness , Agilent J & W ) . Chromatographic separation of the alditol acetates was achieved with the temperature gradient: 100°C for 1 min . , increased to 150°C at 20°C/min . , increased to 200°C at 2 . 5°C/min , and finally increased to 275°C at 30°C/min . Total chromatographic time was 37 min . The mass spectrometer was operated in the EI mode 70 eV with a source temperature 250°C , transfer line temperature 250°C , scan range 50–450 amu . Data acquisition and analysis were ascertained by Varian MS workstation software . Identification of the alditol acetate derivatives was carried out by comparing retention time and mass spectra with authentic standards . The peak area of individual alditol acetates was calculated from the total ion chromatogram , normalized to peak area of the internal standard and finally normalized to the total protein content of sample . All statistical analyses were performed using GraphPad Prism ( GraphPad Software , Inc . ) . Mutant strains were normalized to wild type . The biofilm data met the assumption of normally distributed values , and therefore a paired Student’s t-test was utilized to analyze the data . The data was not distributed normally for the analysis of swimming motility , thus the Mann-Whitney test was used to analyze the motility data . Significance in all analyses was defined by a calculated p-value less than or equal to 0 . 001 , which was determined using the Bonferroni correction to account for multiple comparisons . Error bars indicate standard error of the mean .
Key genes involved in B . pseudomallei biofilm formation were identified by screening a 1026b two-allele transposon library for mutants defective in pellicle biofilm formation . This approach allows for efficient genome-scale phenotypic screening of nearly all the non-essential genes to identify putative gene function . During the primary biofilm screen , 118 unique transposon insertional mutants exhibited reduced or no pellicle formation by visual inspection . For secondary screening purposes , we performed a quantitative static biofilm assay with the 118 transposon insertional mutants identified in the primary screen . Sequence analysis was performed to eliminate duplicate transposon insertions in the same ORF . Additionally , establishing a cut-off of >20% decrease in biofilm formation in the quantitative biofilm assay resulted in the retention of 59 mutants for further analyses ( S1 Table ) . As a means to reduce the number of transposon insertion mutants in follow-up assays , we also conducted bioinformatics analyses with DOOR 2 . 0 operon analysis [37] . Based on this analysis , 37 transposon insertional mutants which represented the first gene in each putative operon were selected ( S1 Table ) . These transposon insertion mutants exhibited an approximately 40–60% decrease in biofilm formation as compared to the wild type ( Fig 1 ) . A majority of the transposon insertions with a biofilm-deficient phenotype were within genes located in clusters on chromosome I and included genes predicted to produce flagella , fimbriae , regulators , polysaccharides , and an assortment of housekeeping and hypothetical proteins . Genes involved in the production of flagella and pili are key structural components that are known to contribute to the establishment and initial phases of bacterial biofilm formation in some bacteria ( for review [48] ) . As predicted , the screen identified 17 transposon insertions in genes involved in either motility or chemotaxis that were abrogated in biofilm formation ( S1 Table ) . We chose to characterize ten transposon mutants based on the criteria described above . All of the motility/chemotaxis mutants I0029::T24 ( fliM ) , I0032::T24 ( fliP ) , I3239::T24 ( flgE ) , I3242::T24 ( flgB ) , I3243::T24 ( flgA ) , I3283::T24 ( fliF ) , I3529::T24 ( flhA ) , I3544::T24 ( motA ) , I3545::T24 ( flhC ) , and II2012::T24 were decreased in biofilm formation as compared to the wild type ( Fig 1 ) . The transposon screen also identified pili biosynthesis genes , which are known to play a role in establishing biofilms in some bacteria . The fimbriae gene cluster identified in this screen is one of six predicted type I fimbriae biosynthesis clusters in B . pseudomallei [49] and is homologous to the cupE gene cluster that has been implicated in P . aeruginosa biofilm formation [50] . All of the genes ( Bp1026b_I1992-I2000 ) in this 11 kb gene cluster were identified in this transposon screen ( S1 Table ) . All four of the representative transposon insertion mutants from this gene cluster were significantly impaired in biofilm formation ( Fig 1 ) . Furthermore , we also identified transposon insertion mutants in four hypothetical genes ( Bp1026b_I0138 , Bp1026b_I0743 , Bp1026b_II1577 , and Bp1026b_II2160 ) , one transporter gene ( Bp1026b_II1233 ) , two genes involved in purine synthesis ( Bp1026b_I1351 , purL , and Bp1026b_I2392 , purD ) , RNA polymerase factor sigma 54 ( Bp1026b_I2962 , rpoN ) , a DNA response regulator ( Bp1026b_I1954 ) , a tRNA modification GTPase ( Bp1026b_I0084 , trmE ) , a DnaA regulator inactivator Hda ( Bp1026b_I0492 ) , a sigma-54 interacting response regulator protein ( Bp1026b_I0895 ) , and a sensor histidine kinase/response regulator ( Bp1026b_II2527 ) ( S1 Table ) that resulted in decreased biofilm formation ( Fig 1 ) . The I1954::T24 mutant exhibited a >80% decrease in biofilm formation ( Fig 1 ) . Based on bioinformatics analysis with SMART , Bp1026b_I1954 is predicted to possess REC and transcriptional regulator domains and is most likely part of a two-component regulatory system . One of the more striking findings from this study was the near saturation of a putative 28 kb polysaccharide biosynthesis gene cluster ( Bp1026b_I2907-Bp1026b_I2927 ) which has been designated becA-R ( biofilm exopolysaccharide gene cluster ) . Only three genes within this novel 18 gene biosynthetic cluster were not identified during the primary phenotypic screen of the transposon library . Of these three , Bp1026b_I2922 ( becM ) and Bp1026b_I2924 ( becO ) are not represented in the two-allele library and were unavailable to test . Bp1026b_I2915 ( becF ) is represented in the library but was not identified in this screen for biofilm deficient mutants . The transposon screen also identified duplicate alleles for eight out of 18 genes in this cluster . This biosynthetic cluster contains seven independent operons as predicted by DOOR 2 . 0 operon analysis [37] . The ORFs within this cluster are predicted to encode a glycosyl transferase , glycosyl hydrolase , capsular polysaccharide biosynthesis/export periplasmic proteins , UDP-glucose lipid carrier transferase , and a mannose-1-phosphate guanylyl transferase/mannose-6-phosphate isomerase ( S1 Table ) . We chose to study seven transposon insertional mutants ( Bp1026b_I2907 ( becA ) , Bp1026b_I2910 ( becD ) , Bp1026b_I2914 ( becE ) , Bp1026b_I2916 ( becG ) , Bp1026b_I2921 ( becL ) , Bp1026b_I2923 ( becN ) , and Bp1026b_I2925 ( becP ) ) from this polysaccharide gene cluster that exhibited reduced biofilm formation ( Fig 1 ) . Interestingly , the transposon screen also identified four transposon insertional mutants in genes predicted to contribute to polysaccharide biosynthesis that are not associated with the exopolysaccharide biosynthetic gene cluster ( Bp1026b_I2907-Bp1026b_I2927 , becA-R ) . Two of these genes are Bp1026b_I0648 , a glycosyl transferase family protein , and Bp1026b_I0649 , a UDP-glucose 4-epimerase , both of which are part of a five gene cluster adjacent to the wbiA gene cluster responsible for lipopolysaccharide biosynthesis [51 , 52] . The I0649::T24 mutant had the greatest reduction ( >60% ) in biofilm formation as compared to all of the other transposon insertional mutants in genes predicted to participate in polysaccharide biosynthesis ( Fig 1 ) . Additional transposon insertions that resulted in decreased biofilm formation ( Fig 1 ) were identified in Bp1026b_II1959 , a predicted glycosyltransferase that is part of the capsule III biosynthetic cluster [53] , and Bp1026b_II2123 , a predicted tyrosine-protein kinase . Exopolysaccharides are a key component of many bacterial biofilms [10] , and the biofilm-associated exopolysaccharide biosynthetic gene cluster is conserved in other closely related species of the B . pseudomallei phylogenetic complex which includes B . mallei and B . thailandensis [54] . The exopolysaccharide biosynthesis genes are encoded by 18 loci and 3 pseudogene remnants spanning Bp1026b_I2907-Bp1026b_I2927 on chromosome I of B . pseudomallei 1026b and share high sequence identity with other clustered genes in B . mallei and B . thailandensis ( S1 Fig and S3 Table ) . Nearly identical gene clusters were identified in the sequenced genomes of B . mallei ATCC 23344 and B . thailandensis E264 spanning the loci BMA0027-BMA0048 and BTH_I0520-BTH_I0537 , respectively . Genetic alignments for B . pseudomallei and B . mallei revealed greater than 99% sequence identity at the nucleotide level or almost full conservation of this cluster . Alignments for B . pseudomallei and B . thailandensis identified an average sequence identity of 93 . 2% among the 18 loci , indicating a similarly high level of conservation . In particular , the gene locus predicted to encode for the mannose-1-phosphate guanylyltransferase , manC ( Bp1026b_I2925 , becP ) , shares 99 . 8% identity with BMA0029 in B . mallei and 94 . 1% identity with BTH_I0522 in B . thailandensis . These observations indicate a high degree of genetic conservation for this EPS cluster among closely related species despite differences in human and animal pathogenicity and environmental niche adaptation . Beyond the closely related B . pseudomallei complex of bacteria , the becA-R biosynthetic cluster ( Bp1026b_I2907-Bp1026b_I2927 ) was most highly conserved with a B . cenocepacia J2315 gene cluster ( Fig 2 and S3 Table ) . Bioinformatics analyses comparing the putative exopolysaccharide gene clusters of B . pseudomallei 1026b and B . cenocepacia J2315 revealed high sequence conservation amid genetic rearrangement between the closely related pathogens . The exopolysaccharide cluster of B . cenocepacia J2315 has been previously reported to be encoded by loci BCAM1330-BCAM1341 on chromosome II and experimentally validated as a major structural component of biofilms [55] . We characterized the genetic sequence similarity based on the common ancestral origin of these strains in order to make comparisons with the biofilm-associated exopolysaccharide gene cluster identified from B . pseudomallei 1026b ( Fig 2 ) . Local pairwise alignments of genomic sequences using BLASTN showed high sequence homology within a region on chromosome I of B . pseudomallei 1026b and the cluster on chromosome II of B . cenocepacia J2315 . Of the 18 predicted coding sequences that comprise the exopolysaccharide biosynthesis cluster , 14 are directly homologous to the exopolysaccharide cluster in B . cenocepacia J2315 , spanning BCAM1334-BCAM1350 ( Fig 2 ) . The DNA sequences from 14 coding regions are 75 . 8% identical altogether and represent gene cluster homologues between the two Burkholderia species . Interestingly , a majority of the homologous coding regions have flipped directional arrangements while maintaining high sequence identity ( Fig 2 ) . The coding regions of the B . pseudomallei 1026b cluster fully or almost fully align to homologous sequences in B . cenocepacia J2315 with percent identities ranging from 66–84% at the nucleotide level except for Bp1026b_I2920 ( becK ) , which has an additional unique sequence in the middle of the gene interrupting alignment . A notable locus , Bp1026b_I2925 ( becP ) , predicted to encode for mannose-1-phosphate guanylyltransferase ( manC ) that is required to catalyze the formation of nucleotide sugar GDP-mannose , shares 82 . 3% identity with BCAM1340 . Additionally , Bp1026b_I2910 ( becD ) , shares 74 . 9% identity with BCAM1349 , the proposed transcriptional regulator of the exopolysaccharide gene cluster in B . cenocepacia J2315 [55] . These results indicate functional conservation of the exopolysaccharide cluster; however , our bioinformatics analysis revealed some crucial differences . Four loci in the 1026b predicted EPS cluster showed no homology to the J2315 cluster . The loci Bp1026b_I2907 ( becA ) , I2908 ( becB ) , I2922 ( becM ) , and I2923 ( becN ) , are predicted to encode a glycosyltransferase protein , a polysaccharide export periplasmic protein , a PAP2 superfamily protein , and a glycoside hydrolase family protein , respectively . Bp1026b_I2907 ( becA ) shares similarity to two predicted glycosyltransferases in the B . cenocepacia J2315 EPS cluster , BCAM1337 and BCAM1338 with 60 . 15% and 62 . 56% respective nucleotide identities . Likewise , Bp1026b_I2908 ( becB ) shares 61 . 89% nucleotide identity to BCAM1330 , which is predicted to encode a putative polysaccharide export protein . However , the sequence correlations of Bp1026b_I2907 ( becA ) and Bp1026b_I2908 ( becB ) to B . cenocepacia J2315 do not pass our E-value threshold of 1e-3 , representing a 0 . 001 chance of random sequence alignment , indicating that these correlations are not biologically significant . Bp1026b_I2922 ( becM ) and Bp1026b_I2923 ( becN ) also do not show significant sequence correlations to the B . cenocepacia J2315 genome; however , the flanking coding sequences of Bp1026b_I2921 ( becL ) and Bp1026b_I2924 ( becO ) appear homologous to BCAM1342 and BCAM1341 , respectively , amid directional inversions ( Fig 2 ) . Interestingly , Bp1026b_I2921 ( becL ) and Bp1026b_I2924 ( becO ) are flanked by large noncoding intergenic regions totaling 1176bp , which co-localizes to an intergenic region spanning 1319bp in B . cenocepacia J2315 . One explanation for this disparity involves the acquisition of the two coding sequences by B . pseudomallei 1026b for a species-specific fitness advantage . Interestingly , we identified a transposon mutant insertion in Bp1026b_II1959 , which is part of the capsule III biosynthetic cluster [53] . Our bioinformatics analysis revealed that this biosynthetic cluster ( Bp1026b_II1956-Bp1026b_II1966 ) is homologous to the bce-I cluster that has been previously reported to synthesize the exopolysaccharide cepacian [40] in bacteria from the Burkholderia cepacia complex ( S2A Fig and S4 Table ) . However , we only identified a single gene in this biosynthetic cluster in our biofilm screen and we did not identify any transposon insertion mutants in the bce-II cluster of the Burkholderia cepacia complex that shares strong homology with Bp1026b_II1796-Bp1026b_II1807 ( S2B Fig and S4 Table ) . Since motility can contribute to biofilm formation , we sought to address whether impaired motility contributed to decreased biofilm formation . A vast majority of the transposon mutants were not altered in swimming motility ( Fig 3 ) ; however , not surprisingly , all transposon insertion mutants in genes involved in flagella production and assembly exhibited decreased motility ( Fig 3 ) . In addition , five other mutants ( I0649::T24 , I2962::T24 ( rpoN ) , I0492::T24 , I0084::T24 ( trmE ) , I and 0138::T24 ) were minimally decreased in swim motility with the exception of the I0084::T24 ( trmE ) mutant , which was significantly impaired in swim motility ( Fig 3 ) . In order to rule out the deleterious effect of potential growth defects on biofilm formation , we assayed all the strains used in these studies for rates of growth . A majority of the biofilm-defective mutants exhibited growth rates comparable to wild type when grown in LB medium for 48 h with the exception of I2962::T24 ( rpoN ) , I0084::T24 ( trmE ) , I1351::T24 ( purL ) , I2392::T24 ( purD ) , and II1233::T24 ( S3 Fig ) . To provide an additional means to evaluate exopolysaccharide production , bacterial strains were cultivated on NAP-A agar which contains neutral red and crystal violet dyes [36] . Previous studies have reported the association of rugose colony morphology and the production of exopolysaccharides in a variety of Gram-negative bacteria [56–58] . The appearance of rugose ( wrinkled ) colony morphology and pellicle biofilm formation has also been reported to be linked to exopolysaccharide production in B . cenocepacia [55] . Thus , we hypothesized that the smooth appearance on NAP-A agar medium is directly or indirectly related to the loss or decreased production of exopolysaccharide . In this study , all of the transposon insertion mutants in the novel biofilm exopolysaccharide gene cluster ( Bp1026b_I2907-Bp1026b_I2927 , becA-R ) and a majority of the remaining biofilm-deficient transposon mutants were smooth in appearance , while the motility mutants were rugose ( Fig 4 ) . Two transposon insertional mutants , I1351::T24 ( purL ) and I2392::T24 ( purD ) , in genes involving purine biosynthesis appeared to preferentially uptake crystal violet and were smooth in appearance . Interestingly , one transposon mutant , I0649::T24 ( a predicted UDP-glucose-4-epimerase ) , was more rugose ( wrinkly ) and heavily-pigmented as compared to wild type ( Fig 4 ) . A summary of the phenotypes for the biofilm-deficient transposon mutants can be found in Table 1 . Complementation of representative mutant strains was achieved using a select-agent compliant methodology that removes the transposon leaving behind a FRT signature sequence , which still disrupts the reading frame of the targeted gene . The FRT mutants in a DNA response regulator ( I1954::T24 ) , a glycosyl transferase ( I2907::T24 ( becA ) ) , and a sensor histidine kinase ( II2527::T24 ) were complemented with the respective full length genes at a neutral Tn7 site on the chromosome . Corresponding empty vector ( EV ) control strains were created for the wild type and the FRT mutants . Complementation of the I1954::FRT mutant expressing full length Bp1026b_I1954 significantly restored biofilm formation as compared to the I1954::FRT EV control strain ( Fig 5A ) . It should be noted that I1954::T24 mutant insertional mutant exhibited the greatest decrease in biofilm formation from all mutants that were identified in the initial screen ( Fig 1 ) . The biofilm-defective phenotypes of the II2527::FRT and I2907::FRT mutants were also significantly complemented with full length clones as compared to their respective FRT EV control strains ( Fig 5A ) . Additional evidence for complementation of all three FRT insertion mutants was also observed on NAP-A agar plates supplemented with IPTG . Wild type EV exhibited a rugose ( wrinkly ) phenotype on NAP-A , while all three FRT insertion mutants ( I1954::FRT , I2907::FRT ( becA ) , and II2527::FRT ) exhibited a smooth phenotype that did not bind neutral red dye on these plates ( Fig 5B ) . The rugose ( wrinkly ) phenotype was restored in all three transposon mutants when complemented with their respective full length genes ( Fig 5B ) . Complementation of II2527::FRT not only restored the wrinkly phenotype , but the complemented clone also exhibited enhanced dye binding as compared to the wild type EV control and other FRT complemented strains ( Fig 5B ) . Given the importance of the exopolysaccharide biosynthetic gene cluster ( Bp1026b_I2907-I2927 ) in biofilm formation , we have designated this gene cluster as becA-R . We generated a deletion mutant of the entire becA-R gene cluster ( Bp1026b_I2907-I2927 ) in both the wild type and capsule I deficient ( ΔwcbR-A ) [44] backgrounds to further investigate how this exopolysaccharide contributes to B . pseudomallei biofilm formation . Deletion of the entire becA-R gene cluster in the wild-type background resulted in approximately a 63% decrease in biofilm formation ( Fig 6A ) and pellicle biofilm formation was also impaired ( Fig 6B ) . Interestingly , deletion of capsule I resulted in a 40% increase in biofilm formation which is consistent with a previously published report [53] ( Fig 6A ) . Loss of both capsule I ( ΔwcbR-A ) and the biofilm-associated exopolysaccharide ( ΔbecA-R ) reduced the biofilm 40% , which was an intermediate phenotype as compared to the wild type and the exopolysaccharide deletion mutant ( Fig 6A ) . Motility was not significantly altered in either the ΔwcbR-A , ΔbecA-R or the ΔwcbR-A ΔbecA-R double mutants as compared to wild type ( Fig 6C ) . Interestingly , the wild type and the ΔwcbR-A mutant exhibited rugose colony morphology in contrast to the ΔbecA-R mutant and the ΔwcbR-A ΔbecA-R double mutant that were smooth in appearance on NAP-A agar ( Fig 6D ) . Rugose colony morphology is often associated with the production of exopolysaccharides [55–58] . We conducted western blot analysis on polysaccharide preparations in order to differentiate the biofilm-associated exopolysaccharide encoded by becA-R from the previously described capsular polysaccharide ( CPSI ) and >150kDa acidic exopolysaccharide from B . pseudomallei [13–15 , 59] . Western blot analysis on polysaccharide preparations from ΔbecA-R and ΔwcbR-A ΔbecA-R double mutants indicated that the biofilm-associated exopolysaccharide described in the current study and the previously described acidic exopolysaccharide are not the same using an antibody ( mAb 3015 ) raised against the acidic exopolysaccharide ( Fig 7A ) . Cross reactivity with purified CPSI suggested that the acidic exopolysaccharide-specific antibody reacts with a constituent of CPSI ( Fig 7A ) . We also confirmed that the polysaccharide preparations produced by the becA-R biosynthetic cluster were not CPSI ( Fig 7B ) , since the purified CPSI [14] and preparations from ΔbecA-R are reactive to the CPSI-specific 4C4 IgG1 antibody [46 , 60 , 61] , whereas the preparations from ΔwcbR-A mutant and the ΔwcbR-A ΔbecA-R double mutants are not reactive to the CPSI-specific 4C4 IgG1 antibody . Carbohydrate analysis of polysaccharide preparations from wild-type B . pseudomallei Bp82 indicated that these preparations are comprised primarily of four monosaccharides: glucose , galactose , rhamnose and mannose , in a ratio of 0 . 46:1 . 41:0 . 43:0 . 14 , respectively ( Fig 7C and 7D ) . Comparative analysis suggested that there is a decrease of all four monosaccharides , rhamnose , mannose , glucose , and galactose in the ΔbecA-R or ΔwcbR-A ΔbecA-R double mutant as compared to the wild-type Bp82 ( Fig 7D ) . In addition to characterizing the biofilm-defective mutants identified in our screen , we sought to characterize seven transposon insertion mutants and four deletion mutants in B . pseudomallei genes previously described to contribute to biofilm formation in the literature ( S5 Table ) , since these were candidate genes that we expected to identify in our screen [19–24 , 62] . Seven transposon insertional mutants from the B . pseudomallei 1026b T24 library in addition to four efflux pump deletion mutants were tested in various assays ( S5 Table ) . Under the conditions used in our screen , only one of the seven transposon mutants , II0971::T24 ( bpsl1 ) , recapitulated a reduced biofilm phenotype that had been previously reported [18] and two ( Bp400 ΔbpeAB-oprB::FRT ΔamrRAB-oprA and Bp207 ΔamrRAB-oprA::FRT ΔbpeAB-oprB::FRT ) of the four efflux pump deletion mutants exhibited a significant reduction in biofilm formation ( Fig 8A and Table 2 ) . However , the initial pellicle biofilm screen did identify transposon insertion mutants in a two-component regulator , bfmR ( Bp1026b_I1992 ) , and the corresponding pili biosynthesis genes that it regulates [21] . We also evaluated whether these mutants were altered in swim zone diameter or growth rate under our conditions . Two transposon insertional mutants in quorum sensing , II0971::T24 ( bpsl1 ) and II0974::T24 ( bpsR1 ) , were impaired in swim zone diameter ( Fig 8B ) , while one transposon mutant , I3555::T24 ( fliC ) , was significantly impaired in swim zone diameter ( Fig 8B ) . The swim zone diameters of two efflux pump mutants ( Bp50 ΔamrRAB-oprA and Bp207 ΔamrRAB-oprA ΔbpeAB-oprB ) were reduced by roughly 50% ( Fig 8B ) . Two transposon mutants , II1441::T24 and II0971::T24 ( bpsl1 ) in addition to two efflux pump mutants , Bp207 ΔamrRAB-oprA ΔbpeAB-oprB and Bp400 ΔamrRAB-oprA ΔbpeAB-oprB exhibited delayed or slowed growth over 48 h as compared to wild type ( Fig 8C ) . The transposon insertional mutants in genes previously described to be biofilm defective exhibited a rugose appearance similar to wild type on NAP-A agar ( Fig 4 ) . However , the colony morphology of the efflux pump mutants was not fully assessed due to the sensitivity of these mutants to the antibiotics in the NAP-A medium . A summary of the phenotypes for the previously published biofilm mutants can be found in Table 2 .
B . pseudomallei is the etiological agent of melioidosis , a disease that is often misdiagnosed due to its many clinical manifestations . Inaccurate or delayed diagnoses , lack of a vaccine , evasion of the immune system , and intrinsic antibiotic resistance contribute to the high mortality rate of this disease . Although , the precise role ( s ) of B . pseudomallei biofilm formation in the initiation of an infection and continued persistence in a mammalian host is not fully understood , the identification of the genes that contribute to biofilm formation may provide some insight to address these fundamental questions . It has been previously described that B . pseudomallei can be found in unusual locations in the human body and produce exopolysaccharides that contribute to the evasion of phagocytosis and persistence of chronic infections [63] . A better understanding of the tolerance associated with B . pseudomallei biofilms to antibiotics may help to explain the lack of success in the treatment of the chronic manifestations of melioidosis [64] . In addition , clinical studies suggest that melioidosis relapse may be associated with the biofilm-forming capacity of the primary infecting isolate [65] . To gain a better understanding of the genes that contribute to B . pseudomallei biofilm formation , we utilized a sequence-defined two allele transposon library of B . pseudomallei 1026b to identify 59 genetic loci involved in biofilm formation . This functional based screening approach identified non-essential genes that are directly associated with biofilm formation . The additional phenotypic characterization of these transposon insertional mutants ( Table 1 ) allows for a more detailed analysis of the contribution of these genes to biofilm formation , as opposed to alternative analysis methods that rely on global transcriptional profiling of non-isogenic strains or strains that have an evolutionary relationship to B . pseudomallei . However , these approaches complement the results of the studies reported here . A recent transcriptome analysis of two non-isogenic strains of B . pseudomallei that produce high and low levels of biofilm identified 563 differentially regulated genes using RNA-seq that may contribute to biofilm formation [66] . In this study , we identified and characterized some of the same genes using a functional-based screening approach . Additionally , another RNA-seq based study of a contact-dependent growth inhibition system [67] and the corresponding genes regulated from B . thailandensis identified genes that are homologous to the biofilm exopolysaccharide biosynthesis , fimbriae production , and an exopolysaccharide tyrosine-protein kinase identified in this study . A subset of the genes identified in this study has previously been highlighted in other research studies ( Table 2 ) . The sensor histidine kinase encoded by Bp1026b_I1993 has been previously shown to be up-regulated during an acute infection in Syrian hamsters and a deletion mutant exhibited a higher LD50 as compared to the wild type suggesting a potential role as a virulence factor during infection [68] . The ability to attach to host tissues is a crucial step during infection and biofilm formation . More recently , seven out of the nine genes in this fimbriae gene cluster were reported to be differentially regulated between low and high biofilm producing clinical isolates [66] . In addition , an insertion mutant of BPSL2024 ( Bp1026b_I1992 ) , designated as bfmR ( biofilm formation associated regulator ) , which is a predicted DNA-binding response regulator , exhibited poor growth under iron-limiting conditions , reduced biofilm-forming capacity ( ~70% ) , and reduced fimbriae production , while motility was comparable to the wild type [21] . In this study , we phenotypically characterized four mutants in this fimbriae gene cluster , I1993::T24 , I1998::T24 , I1999::T24 , and I2000::T24 , although we had initially identified all nine genes in this gene cluster for their contribution to biofilm formation . Conflicting reports on the role of biofilm components has further complicated our current ability to determine the role of biofilms in melioidosis [25–27] . As a means to compare our results with the previously published body of literature , we evaluated transposon mutants in the genes previously described to contribute to biofilm formation . A majority of the transposon mutants in published genes with a reported biofilm deficient phenotype that we retested from our library did not exhibit a decrease in biofilm formation under the conditions tested . This could be attributed to variation in the conditions tested , the genetic background of the strains , or the method of gene inactivation used in those studies . A primary example of a potential discrepancy as observed in this study is the transposon insertion in fliC . Under the conditions used in this study , I3555::T24 ( fliC ) was defective in swimming motility , but still competent to form biofilms ( Table 2 ) . We also previously characterized that this transposon insertion mutant did not produce detectable levels of FliC in western blot analyses [35] . Interestingly , other transposon insertion mutations in genes that contribute to flagella biogenesis and function ( e . g . fliM , fliP , flgE , flgB , fliF , motA , flhA , and flhC ) were identified to be defective in both motility and biofilm formation in this study . These results suggest that flagella biogenesis and function contributes to pellicle biofilm formation and static biofilm formation under the conditions tested; however , it is presently unclear as to the exact role of the FliC filament in biofilm formation . Biofilm studies have previously evaluated the role of fliC in an aflagellate transposon insertion mutant designated as MM35 under a variety of conditions in various biofilm formation assays , which resulted in an approximate 40 to 77% reduction in biofilm forming capacity in these published reports [22 , 62] . The variability of the MM35 fliC mutant as reported by those previous studies is an indication that contribution of FliC is conditionally dependent and other factors may share a functionally redundant role with other factors to promote biofilm formation . Future efforts will be directed at further defining the role of FliC under various conditions . Due to the multitude of factors that contribute to biofilm formation and the nature of our initial screen ( qualitative pellicle biofilm assay ) , there are potentially more genes that are involved in biofilm formation that were not identified in our screen . These genes may be detected in more quantitative assays when screened under more diverse growth conditions ( temperature , carbon source , oxygen tension , etc . ) . As of yet , the role of biofilm formation and biofilm-specific components has yet to be elucidated [19] . One of the major issues associated with the conclusions that have been made so far is the lack of studies that look at the role of biofilms in chronic infection models of melioidosis . To address important research questions in a model B . pseudomallei strain as a community , we have validated this sequence-defined transposon insertion library used in this study and deposited it with BEI resources . Deletion of the entire B . pseudomallei biofilm exopolysaccharide cluster led to a significant decrease in biofilm formation and smooth appearance on NAP-A ( Fig 6A and 6D ) , which is indicative of the loss of exopolysaccharide production . Further investigation by Western blot analysis with mAbs specific for the CPSI and the previously described acidic polysaccharide indicated that the product of the becA-R biosynthetic cluster does not contribute to the production of previously characterized polysaccharides identified by these antibodies . The analysis of the CPSI-deficient mutant ( wcbR-A ) and purified CPSI capsular polysaccharide ( [→3 ) -2-O-acetyl-6-deoxy-β-D-manno-heptopyranose- ( 1→] ) also indicated that mAb 3015 detects a polysaccharide component that is produced by the genes previously characterized to contribute to CPSI biosynthesis and may likely be specific to CPSI . One might speculate that the previously reported reactivity of the acidic polysaccharide antibody [59] was potentially due to minor CPSI contamination during the acidic exopolysaccharide extraction . Alternatively , this antibody might be cross-reactive with both polysaccharides and deletion of the CPSI gene cluster might also effect the expression of the acidic exopolysaccharide . The role of CPSI in biofilm formation was also highlighted in our epistatic analysis of polysaccharide biosynthesis clusters . In our study , the deletion of CPSI biosynthesis genes increases biofilm formation , which has been previously observed [53] . This may indicate that CPSI production alters the dynamics of biofilm attachment and subsequent formation . These effects could also be the result of altered cellular levels of nucleotide sugar precursors used in polysaccharide biosynthesis . Interestingly , biofilm formation is also increased in the ΔwcbR-A ΔbecA-R double mutant as compared to ΔbecA-R , indicating that capsule production effects biofilm formation in the absence of the biofilm-associated exopolysaccharide . Thus , CPSI has an overriding contribution and generates interesting complications for the study of B . pseudomallei biofilm formation . Future biofilm studies that aim to understand the biofilm physiology and dynamics of B . pseudomallei biofilm growth will have to be performed in strains that produce CPSI based on the effects that this EPS component has on biofilm formation . Excluding CPSI from future biofilm analyses conducted in closely related Burkholderia species or strains that do not produce CPSI will confound the elucidation of the role of biofilms in melioidosis . We determined through carbohydrate analysis that the biofilm-associated exopolysaccharide is comprised of four primary monosaccharides: glucose , galactose , rhamnose , and mannose , which is consistent with a recent report on a biofilm-associated exopolysaccharide purified from B . pseudomallei [69] . However , the ratio of the four monosaccharides reported differs between our two studies , which may be reflective of differences in growth conditions , method of polysaccharide purification , or strains under investigation . It is also unclear if the exopolysaccharide evaluated previously [69] is the sole product of the becA-R cluster that we have characterized in this study . As Burkholderia species are known to produce a diversity of exopolysaccharides , we conducted additional bioinformatics analyses to further characterize the biofilm-associated exopolysaccharide . Based on our genomics analyses , the becA-R cluster reported here is not part of the evolutionarily conserved cepacian biosynthesis clusters ( bce-I and bce-II ) . However , additional exopolysaccharide biosynthetic clusters , which includes the cepacian cluster , are present in other locations within the B . pseudomallei genome and have yet to be fully characterized . Interestingly , the becA-R is highly conserved between B . pseudomallei , B . mallei and B . thailandensis . The strong conservation of the becA-R cluster during B . pseudomallei genome adaptation and reduction in B . mallei suggests that the biofilm-associated cluster contributes to pathogenesis , as genes that are not necessary for living in an animal host would have been predicted to be lost during genome reduction [70] . This is in contrast to the bce-I cluster ( B . cepacia complex annotation ) that has been lost in B . mallei [40] , which is often referred to as capsule III in B . pseudomallei . Together , these results suggest an evolutionary relationship and differentiation of functional roles for the EPS components in Burkholderia species and a critical need to understand their role in biofilm formation and pathogenesis . The literature continues to expand and modify the structural characterization of surface-associated polysaccharides in B . pseudomallei . However , a critical need exists to link the polysaccharides produced and their corresponding biosynthetic genes . Additional research will be conducted to determine the precise composition and structure of the B . pseudomallei biofilm exopolysaccharide produced by becA-R . Additional studies will also be geared at generating monoclonal antibodies for future diagnostics efforts . The ability of bacterial pathogens to attach , colonize surfaces , and form a biofilm is a key first step in the initiation of pathogenesis and evasion of host defenses . In this paper , we identified and characterized the genetic loci that contribute to B . pseudomallei biofilm formation . We also identified an exopolysaccharide that is essential for biofilm formation , which is confined to a few closely related Burkholderia species that comprise the B . pseudomallei complex . To date , the majority of published studies on B . pseudomallei EPS have focused on capsular polysaccharide I ( CPSI ) . However , the B . pseudomallei genome encodes the capacity for expression of multiple additional capsular polysaccharides and secreted exopolysaccharides . The nature and role of these additional EPS components remains to be characterized in the context of B . pseudomallei tissue tropism , biofilm formation , antibiotic tolerance , and persistence in the host . Our future efforts will be focused on characterizing the growth of B . pseudomallei as a biofilm to understand how biofilm growth contributes to antimicrobial tolerance and the failure of antibiotic treatment in patients with melioidosis .
|
B . pseudomallei , the etiological agent of melioidosis , is an emerging pathogen with limited therapeutic options and no available vaccines . A better understanding of the role of biofilm formation during pathogenesis will aid in melioidosis diagnosis and the development of new therapeutics and vaccines . Melioidosis has both acute and chronic disease manifestations in addition to a highly variable period of latency , which contributes to complications in diagnosis and treatment of the disease . Relapsing melioidosis is correlated with biofilm formation and the role of biofilm growth during chronic human infections has been widely accepted . We utilized a two-allele sequence defined transposon mutant library of B . pseudomallei 1026b to identify genes involved in biofilm formation . This study identified factors that contribute to biofilm formation and included a previously undescribed exopolysaccharide and the genes underlying its biosynthesis . Since antibiotic tolerance in B . pseudomallei has been associated with biofilm formation , the genes identified in this study that contribute to biofilm production are potential targets for therapeutic development . Additionally , the products of these biofilm genes could be used for the development of diagnostics and vaccines .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"biofilms",
"exopolysaccharides",
"gene",
"regulation",
"microbiology",
"genetic",
"elements",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"sequence",
"alignment",
"bioinformatics",
"gene",
"expression",
"genetic",
"loci",
"biochemistry",
"polysaccharides",
"database",
"and",
"informatics",
"methods",
"genetics",
"transposable",
"elements",
"biology",
"and",
"life",
"sciences",
"biosynthesis",
"genomics",
"mobile",
"genetic",
"elements",
"glycobiology"
] |
2017
|
Genome-scale analysis of the genes that contribute to Burkholderia pseudomallei biofilm formation identifies a crucial exopolysaccharide biosynthesis gene cluster
|
In the zebrafish , Fgf and Hh signalling assign anterior and posterior identity , respectively , to the poles of the developing ear . Mis-expression of fgf3 or inhibition of Hh signalling results in double-anterior ears , including ectopic expression of hmx3a . To understand how this double-anterior pattern is established , we characterised transcriptional responses in Fgf gain-of-signalling or Hh loss-of-signalling backgrounds . Mis-expression of fgf3 resulted in rapid expansion of anterior otic markers , refining over time to give the duplicated pattern . Response to Hh inhibition was very different: initial anteroposterior asymmetry was retained , with de novo duplicate expression domains appearing later . We show that Hmx3a is required for normal anterior otic patterning , and that otic patterning defects in hmx3a-/- mutants are a close phenocopy to those seen in fgf3-/- mutants . However , neither loss nor gain of hmx3a function was sufficient to generate full ear duplications . Using our data to infer a transcriptional regulatory network required for acquisition of otic anterior identity , we can recapitulate both the wild-type and the double-anterior pattern in a mathematical model .
The otic placode—precursor of the vertebrate inner ear—has the remarkable ability to generate a mirror-image organ with duplicate structures under some experimental conditions in fish and amphibians , as originally described by R . G . Harrison over eighty years ago ( reviewed in [1] ) . Understanding the generation of such duplicated structures can give us fundamental insights into mechanisms of organ patterning , tissue polarity and symmetry-breaking during embryogenesis . During normal development in the zebrafish , anteroposterior asymmetries in otic gene expression are evident as early as the 4-somite stage ( 11 . 5 hours post fertilisation ( hpf ) ) , when expression of the transcription factor gene hmx3a appears at the anterior of the otic placode [2] . Additional genes with predominantly anterior patterns of expression in the otic placode or vesicle begin to be expressed over the next 10 hours , including the transcription factor genes hmx2 and pax5 [2 , 3] , together with the fibroblast growth factor ( Fgf ) family genes fgf3 , fgf8a and fgf10a [4 , 5] . Later , at otic vesicle stages ( 24 hpf onwards ) , the size and position of the otoliths , together with the position , shape and planar polarity patterns of the sensory maculae , provide landmarks for distinguishing anterior and posterior structures in the ear [6] . In addition , a few markers begin to be expressed specifically in posterior otic tissue ( pou3f3b , bmp7a and fsta ) at otic vesicle stages [3 , 7 , 8] , but these are not reliable posterior markers at earlier otic placode stages . Concomitant with the appearance of anteroposterior asymmetry in the zebrafish otic domain , other early patterning events occur that are symmetrical about the anteroposterior axis . Of relevance for our study , a single sensory-competent domain , marked by the expression of atoh1b , splits into two domains , one at each pole of the ear , by 12 hpf . This process is dependent on Notch signalling and atoh1b function , and defines differences between the poles of the otic placode and a central zone [9] . The two poles express various markers symmetrically , including atoh1a and deltaD , between 14–18 hpf [9] , presaging the appearance of pairs of myo7aa-positive sensory hair cells ( tether cells ) at each pole by 18–24 hpf [10] . Thus , by the completion of otic induction at 14 hpf ( 10 somites ) , the otic domain has two clear poles defined by the symmetric expression of atoh1a and deltaD , with the anterior pole distinguished from the posterior by the asymmetric expression of hmx3a . Although anteroposterior asymmetries in otic gene expression are already apparent by 12 hpf in the zebrafish , these can be disrupted by interfering with extrinsic signalling pathway activity after this time . For example , manipulations of either Fibroblast growth factor ( Fgf ) or Hedgehog ( Hh ) signalling between 14–19 hpf can result in striking double-anterior or double-posterior mirror-image ears . Fgf signalling is both required and sufficient to act as an anteriorising cue , whereas Hh signalling is both required and sufficient for the acquisition of posterior otic identity [6 , 11 , 12] . In these studies , we showed that transient fgf3 mis-expression at 14 hpf or Hh pathway loss-of-function result in the loss of posterior-specific expression domains of fsta at 30 hpf and otx1b at 45–48 hpf , and the gain of anterior-specific gene expression in the posterior of the ear ( hmx2 and pax5 at 24 hpf after fgf3 mis-expression; hmx3a at 30 hpf in Hh loss-of-function mutants ) . These findings suggest that Fgf and Hh signalling normally act to establish and maintain the asymmetric expression of marker genes within the otic epithelium . However , the details of their temporal mode of action in the duplication of anterior otic fates have not been explored . In the current study , we have compared the dynamics of the transcriptional responses that precede the acquisition of a duplicated anterior otic fate in an Fgf gain-of-signalling or a Hh loss-of-signalling context . Although the final duplicated ear structures appear similar after each manipulation , the early transcriptional responses differ for each signalling pathway , progressing in distinct ways to give rise to the double-anterior pattern at larval stages . One gene that shows an early transcriptional response in the zebrafish otic placode to disruption of either Fgf or Hh signalling is the Hmx family homeobox gene hmx3a . We have examined the effects of both loss-of-function and gain-of-function of hmx3a on inner ear patterning . Our data suggest that hmx3a is a key early target for the otic anteriorising activity of Fgf signalling , and that the function of hmx3a is required for the anterior-specific otic expression of fgf3 and pax5 , together with correct positioning and development of the sensory maculae . However , unlike high Fgf levels or low Hh pathway activity , mis-expression of hmx3a was unable to generate full duplications of anterior character at the posterior of the ear . A mathematical model based on our experimental findings can recapitulate both the wild-type and duplicated anterior pattern , allowing us to explore the dynamical principles underlying the generation of a mirror-image duplicated organ system .
To establish optimal conditions for generating a double-anterior ear in the zebrafish embryo , we compared otic phenotypes in transgenic lines for two different fgf genes , fgf3 and fgf8a , with systemic transgene expression driven under the control of the hsp70 heat-shock promoter [13 , 14] . Previously , we showed that a 2-hour heat shock in the Tg ( hsp70:fgf3 ) line at 14 hpf ( 10-somite stage ) resulted in a robust duplication of anterior otic structures [6] . We chose this time point to avoid any interference with otic placode induction , which is also Fgf-dependent , but is complete by 14 hpf [15 , 16] . The 14 hpf time point also follows completion of the Notch-dependent signalling event that distinguishes the otic poles from a central zone of epithelium [9] . For the treatments described here , we reduced the time of heat shock to 30 minutes at 39°C . This shorter heat shock still results in a full ear duplication , but should minimise effects of Fgf mis-expression on other developing organ systems . After heat shock , embryos were then cultured at 33°C for a further 30 minutes , to reduce the incubation temperature gradually , before being returned to 28 . 5°C and incubated until 3 days post fertilisation ( dpf ) for processing and analysis ( Fig 1 ) . This stepwise reduction in temperature is thought to extend transgene activation and reduce cell death following heat shock [17 , 18] . Non-transgenic sibling embryos , subjected to the same heat-shock treatment , served as controls ( S1 Fig ) . In Tg ( hsp70:fgf3 ) embryos , a 30-minute heat shock at 14 hpf gave a robust and complete duplication of anterior otic patterning at 72 hpf , as indicated by the mis-positioning or fusion of the posterior otolith , loss of posterior elements of the saccular macula , and a duplication of anterior ( utricular-like ) sensory elements on the posteroventral floor of the ear ( Fig 1B , 1G and 1G’ ) . The phenotypes seen after mis-expression of fgf8a ( 30-minute heat shock at 14 hpf ) were milder and more variable than those for fgf3 , and included a split and mis-positioned saccular macula rather than a complete duplication of anterior elements ( Fig 1C and 1H ) , and a normal complement of three cristae ( Fig 1H’ ) . A 30-minute heat shock of either transgenic line at a later stage ( 18 hpf ) resulted in only mild effects on ear size and shape , and otolith position ( S2A–S2D Fig ) [6] . We therefore chose to use the Tg ( hsp70:fgf3 ) line , with a 30-minute heat shock ( 39°C ) at 14 hpf followed by 30 minutes at 33°C , in subsequent heat-shock experiments . To optimise our protocols for generating double-anterior duplicated ears through inhibition of Hh signalling , we first examined the ear phenotype in smohi1640Tg/hi1640Tg mutants . The hi1640Tg allele ( a transgenic insertion in the smoothened gene , and a likely null [19] ) is thought to result in a stronger reduction in Hh signalling than the point mutation alleles smob641 and smob577 , both of which predict single amino acid substitutions [20] , and which we used in previous studies [6 , 11] . The smohi1640Tg/hi1640Tg mutants showed a fully penetrant double-anterior duplicated ear phenotype , with two similar-sized small otoliths located ventrally , complete loss of the posterior ( saccular ) macula , and duplication of the anterior ( utricular ) macula at the posterior of the ear , with anterior and posterior elements sometimes present as a contiguous patch of hair cells covering the ventral floor ( Fig 1D and 1I ) . Four cristae , rather than the usual three , were present in all ( 8/8 ) mutant ears imaged ( Fig 1I’ ) . ( For comparison , four cristae were present in only about 50% of ears of smob641/b641 mutant embryos [11] . ) Pharmacological inhibition of the transducer of the Hh pathway , Smoothened , using the small molecule cyclopamine , can also produce double-anterior ear duplications [12 , 21] . This approach enables a conditional inhibition of Hh signalling over a defined time window . For the experiments described here , we treated wild-type embryos with 100 μM cyclopamine from 14–22 . 5 hpf . To examine later stages , we washed out the cyclopamine at 22 . 5 hpf and allowed embryos to develop further until 3 dpf ( 72 hpf ) , when they were fixed for staining and imaging . Stage-matched sibling embryos—either untreated , or treated with vehicle ( ethanol ) only—served as controls . This cyclopamine treatment regime was sufficient to generate the double-anterior ear phenotype , characterised by two ventrally-positioned , small ( utricular-like ) otoliths , loss of the posterior ( saccular ) macula , and a clear duplication of the anterior ( utricular ) macula ( Fig 1E and 1J ) . Ears in 4/8 treated embryos had four cristae ( Fig 1J’ ) ; the remaining 4/8 had the normal complement of three cristae . The size and shape of the ear were less affected than in the smohi1640Tg/hi1640Tg mutant embryos , presumably due to the transient nature of the cyclopamine treatment . Taken together , these data show that either genetic or pharmacological inhibition of Hh signalling in wild-type zebrafish embryos between 14–22 . 5 hpf results in a robust and reproducible double-anterior ear phenotype at 3 dpf . One of the most striking transcriptional changes in response to fgf3 mis-expression is the expansion or duplication of the expression of the anterior otic markers hmx2 and pax5 by 24 hpf [6] . To examine the temporal dynamics of this transcriptional response , we assayed for expression of these and additional anterior otic marker genes following our optimised ‘early’ heat-shock regime ( 14 hpf , 30 min , 39°C ) at three different time points: 16 hpf ( 2 hours post heat shock , to examine any rapid response ) , 22 . 5 hpf ( 8 . 5 hours post heat shock , when anterior otic expression of hmx2 and pax5 is strongly established in wild-type embryos ) and 36 hpf ( 22 hours post heat shock , to examine whether any disruption to the expression pattern resolves or changes over time ) . For hmx2 and pax5 , which showed dynamic expression changes , we subsequently included two additional time points ( 25 . 5 hpf and 30 hpf ) to capture these changes in more detail . We first tested expression of three genes coding for transcription factors ( hmx3a , hmx2 and pax5; Fig 2 ) . At the earliest time point ( 16 hpf , two hours after heat shock ) , hmx3a showed the strongest response: expression had already expanded to cover the entire anteroposterior extent of the otic placode ( Fig 2A and 2B ) . The anterior markers hmx2 and pax5 , not normally expressed at this stage in wild-type embryos , were expressed at very low levels in the anterior of the otic placode of heat-shocked embryos ( Fig 2C–2F ) . We were also able to detect widespread and robust up-regulation of the Fgf target gene etv4 ( formerly pea3 ) in transgenic embryos at this time point ( S3 Fig ) . By 22 . 5 hpf , all three transcription factor genes were strongly expressed in a broad zone across the entire anteroposterior axis of the otic vesicle in heat-shocked embryos , on the medial side , as can be seen in a dorsal view ( Fig 2G–2L ) . Note that the overall size and shape of these otic vesicles were relatively normal . Although there was some variability ( including between both ears of the same fish ) , the vesicles were oval in shape , indicating that otic induction had not been compromised ( compare with the small , rounded vesicles of fgf8ati282/ti282 mutants , in which otic induction is disrupted [4] ) . By 36 hpf , wild-type otic expression of hmx genes was more complex , but a clear difference between anterior expressing and posterior non-expressing regions was evident in ventral otic epithelium ( Fig 2M , 2M’ , 2O and 2O’ ) . By contrast , in heat-shocked embryos , expression of hmx3a remained strong across the entire anteroposterior axis of the ear in ventral regions ( Fig 2N and 2N’ ) ; expression of hmx2 weakened in central regions during intermediate stages , but at 36 hpf was present in a contiguous ventral domain ( Fig 2P and 2P’; S4 Fig ) , while expression of pax5 was lost in central regions , resolving into two discrete ventral domains at the anterior and posterior poles by 25 . 5 hpf ( Fig 2R and 2R’; S5 Fig ) . To test whether the milder ear phenotype caused by a later heat shock reflects a failure to establish the early transcriptional responses described above , we also examined expression of anterior markers in Tg ( hsp70:fgf3 ) embryos after heat shock at 18 hpf ( 30 min , 39°C ) . Unexpectedly , we found that the otic expression of hmx3a and pax5 was very similar to that following an early ( 14 hpf ) heat shock , with a broad band of ectopic expression extending across the entire anteroposterior axis of the otic vesicle by 22 . 5 hpf ( S2E–S2H’ Fig ) . This suggests that the loss of competence to generate a complete double-anterior ear after a late heat shock is not due to an inability to express hmx3a and pax5 ectopically throughout the otic epithelium at otic vesicle stages . To compare the transcriptional response after fgf3 mis-expression at 14 hpf with that following conditional Hh pathway inhibition from the same time point , we examined otic expression of anterior marker genes after treatment of wild-type embryos with cyclopamine ( 100 μM , 14–22 . 5 hpf; Fig 3 ) . To confirm the efficacy of cyclopamine treatment , we also examined the expression of ptch2 , a known target of Hh signalling . Expression of ptch2 was down-regulated throughout the embryo at 22 . 5 hpf , but not abolished ( S6 Fig ) . ( By contrast , ptch2 expression is almost entirely lost at 24 hpf in smohi1640Tg/hi1640Tg mutants [19] ) . We also checked expression of etv4 following cyclopamine treatment , but found no major changes in expression at 22 . 5 hpf ( S6 Fig ) . This result confirmed that there are no strong direct effects of the transient pharmacological inhibition of the Hh pathway on Fgf signalling activity in the ear , in line with our previous findings after genetic abrogation of Hh signalling [6] . We examined otic marker genes at two different time points following cyclopamine treatment ( 22 . 5 hpf and 36 hpf ) . Otic expression of both hmx3a and hmx2 was expanded posteriorly on the medial side of the otic vesicle at 22 . 5 hpf , 8 . 5 hours after the start of the treatment ( Fig 3A–3D’; S7 Fig ) . Expanded otic hmx gene expression was also present by 23 hpf in the smohi1640Tg/hi1640Tg mutant , in which the Hh pathway is constitutively inactive ( S7 Fig ) . Importantly , there was no significant difference in the expansion of hmx3a expression in the ear at 22 . 5–23 hpf between the smo mutants and cyclopamine-treated embryos , indicating that our cyclopamine treatment regime is effective at suppressing Hh signalling relevant to otic patterning at this stage ( S7 Fig ) . The spatial pattern of hmx expansion in response to Hh inhibition was different to that seen after mis-expression of fgf3 . Specifically , in cyclopamine-treated embryos or smohi1640Tg/hi1640Tg mutants , hmx3a and hmx2 were expressed in a graded fashion across the ear at 22 . 5 hpf ( 8 . 5 hours after treatment ) , with higher levels anteriorly , rather than in a uniform broad band ( compare Fig 3B with Fig 2H ) . To examine later time points , treated embryos were transferred to fresh medium at 22 . 5 hpf without cyclopamine , as described above . By 36 hpf ( 13 . 5 hours post wash ) , expression of hmx genes had expanded further posteriorly to cover most of the ventral floor of the otic vesicle in cyclopamine-treated embryos ( Fig 3G–3J’ ) , as observed previously at 30 hpf in contf18b/tf18b and smob641/b641 mutants , both of which have a strong reduction in Hh signalling [11] . Expression of pax5 was slower to respond following cyclopamine-mediated inhibition of Hh signalling , with no apparent expansion of the expression domain within the otic vesicle at 22 . 5 hpf ( Fig 3E–3F’ ) . These results corroborate our previous observations in contf18b/tf18b and smob641/b641 mutants , where there was little change in the otic expression of pax5 at 24 hpf [11] . However , by the later time point ( 36 hpf; 13 . 5 hours post wash ) , a new , discrete domain of pax5 expression appeared within posteromedial otic epithelium of cyclopamine-treated embryos ( Fig 3K and 3L ) . Anteroposterior asymmetry in treated ears was still evident at this stage: the posterior domain of expression was weaker , and in a more medial position , than the anterior expression domain ( Fig 3K and 3L ) . However , the epithelium in posteroventral regions was thicker than normal , indicating development of a duplicate domain of sensory tissue ( Fig 3K’ and 3L’ ) . Taken together , our data indicate that the duplicated anterior domain resulting from either an Fgf gain-of-function or Hh loss-of-function includes a duplication of pax5 expression , but that otic patterning progresses through completely different intermediate states to achieve this duplicated pattern , depending on the signalling pathway that has been disrupted . Expression of fgf3 is itself a marker of anterior otic epithelium from 21 hpf [9] , and so can also be used to indicate the presence of a duplicated anterior otic pattern . We therefore examined the expression of fgf genes to provide additional confirmation of anterior character in the duplicated ears . To distinguish between expression of the fgf3 transgene and endogenous fgf3 expression , we used a probe generated from the fgf3 3’ UTR , which is not included in the transgenic construct . In Tg ( hsp70:fgf3 ) embryos after early ( 14 hpf ) heat shock , expression of endogenous fgf3 now appeared in a new domain at the posterior of the otic vesicle at 22 . 5 hpf ( Fig 4A–4B’ , arrowheads ) . Importantly , expression was not found across the entire anteroposterior axis , but was only present at the poles . Expression of endogenous fgf3 in pharyngeal endoderm beneath the ear was reduced or missing in heat-shocked transgenic embryos ( Fig 4A–4B’ , asterisks ) . We also examined the otic expression of fgf8a and fgf10a ( Fig 4C–4F’ ) . These genes are also normally expressed in the anterior of the otic vesicle , but show a less restricted pattern of expression than that of fgf3 , with weaker expression also normally found in posterior regions [4 , 5 , 22] . Following early heat shock of Tg ( hsp70:fgf3 ) embryos , there was little change in the expression of fgf8a in the otic epithelium , whereas expression of fgf10a was strengthened at both anterior and posterior poles ( Fig 4C–4F’ ) . We also examined the otic expression of fgf genes after pharmacological inhibition of Hh signalling . At 22 . 5 hpf , following cyclopamine treatment from 14 hpf , there was little change in the expression domain or levels of fgf3 or fgf8a in the otic epithelium ( Fig 4G–4J’ ) , although there was loss of an fgf3 expression domain in pharyngeal pouch endoderm ventral to the ear ( Fig 4H’ , asterisk ) . Otic expression of fgf10a was strengthened in about 50% ( 17/29 ) of cyclopamine-treated embryos at this early time point , especially at the anterior otic pole ( Fig 4K–4L’ ) . At 36 hpf ( 13 . 5 hours after cyclopamine wash-out ) , new , discrete domains of fgf3 and fgf8a had appeared at the posterior of the ear , indicating a duplication of anterior otic character ( Fig 4M–4P’ , arrowheads ) . By 48 hpf , the duplicated expression domain of fgf8a persisted ( Fig 4Q’ , arrowhead ) , and loss of the thickened epithelium characteristic of the posterior macula on the medial wall of the otic vesicle was also apparent ( Fig 4Q and 4Q’ , brackets ) . These data demonstrate that in both Fgf gain-of-function and Hh loss-of-function contexts , the duplicated anterior otic character includes expression of fgf genes . Note that the double-anterior ear phenotypes resulting from fgf3 mis-expression or Hh inhibition differ from those resulting from a loss of Notch signalling . In mindbomb1ta52b/ta52b ( mib1 ) mutants , in which the ta52b mutation is known to have strong antimorphic effects on Notch signalling [23] , an expanded expression domain of atoh1 genes marks a sensory-competent area in the otic epithelium at 14 hpf [9] . Hair cells differentiate precociously and in excess; supporting cells are missing; otoliths are small and fail to biomineralise correctly; and supernumerary hair cells are extruded from the ear by 60 hpf [24 , 25] . By contrast , in the ears of embryos that have undergone fgf3 mis-expression or Hh inhibition from 14 hpf , we do not find any evidence for supernumerary or precocious hair cell production , loss of supporting cells , or extrusion of hair cells from the epithelium . Size and shape of otoliths is affected , but they appear to become mineralised normally , unlike in mib1 mutants . The mib1ta52b/ta52b mutant phenotype is therefore quite different to the phenotypes we have described here . To confirm this , we have examined the expression of hmx3a , fgf3 , fgf8a and pax5 in mib1ta52b/ta52b mutants ( S8 Fig ) . We found no evidence for any expansion of hmx3a or pax5 expression , posterior duplication of fgf3 , or posterior upregulation of fgf8a expression . By contrast , anteroposterior asymmetry in otic expression was retained for all four genes , but all were in fact down-regulated at the anterior pole of the otic vesicle . Expression of fgf3 in pharyngeal endoderm beneath the ear was also unaffected in mib1ta52b/ta52b mutants ( S8F’ Fig ) , in contrast to the effects of fgf3 mis-expression ( Fig 4B’ ) or Hh inhibition ( Fig 4H’ ) . Given the early anterior-specific otic expression of hmx3a [2] , the dependence of this expression on Fgf signalling [6 , 26 , 27] , and the rapid change in otic hmx3a expression after mis-expression of fgf3 or Hh inhibition ( this work ) , we hypothesised that hmx3a is required for normal otic anterior development . A previous study using morpholino-mediated knockdown suggested a requirement for both hmx3a and hmx2 in acquisition of anterior otic identity and expression of pax5 [2] . However , the effects of individual gene knockdown or mutation were not reported . To test the individual requirement for hmx3a function in the acquisition of otic anterior identity , we examined the ear phenotype in homozygous mutants for a recessive truncating allele lacking the homeodomain , hmx3aSU3 , which we generated using CRISPR/Cas9 technology ( Fig 5A; Materials and Methods ) . In homozygous hmx3aSU3/SU3 mutants , the otoliths were positioned close together at 33 hpf , were side by side at 48 hpf , had started to fuse at 66 hpf and had fully fused by 4 dpf ( Fig 5B and 5C ) . This phenotype appeared to be fully penetrant ( 38/143 embryos from a cross between heterozygous parents; 26 . 6% ) . Semicircular canal pillars and the dorsolateral septum were present in the ears of mutant embryos , although formation of the ventral pillar was delayed . Overall , the ear shape appeared more symmetrical than in wild-type siblings ( Fig 5B and 5C ) . We imaged ears from three mutant embryos at 3 dpf to analyse sensory patch formation ( Fig 5D–5E’; S9 Fig ) . In all three ears imaged , the two maculae appeared fused or closely juxtaposed . Although the anterior and posterior elements of the fused macula were not obviously distinct , the overall shape retained some anteroposterior asymmetry . Hair cells of the anterior ( utricular ) macula were displaced medially , and in one of three ears imaged , were reduced in number . In the two other examples , however , normal numbers of hair cells were present ( S9 Fig ) . The posterior macula was misshapen , and lacked the anterior extension present in the wild type . All three cristae were present ( n = 3 ears; Fig 5D’ and 5E’; S9 Fig ) . To understand the basis of the hmx3aSU3/SU3 mutant otic phenotype at 3–4 dpf , we examined expression of markers at earlier ( otic vesicle ) stages . At 24 hpf , expression of both hmx3a and hmx2 was reduced in intensity within the otic epithelium . On the medial side of the ear , the spatial extent of hmx3a and hmx2 expression was unchanged ( Fig 5F–5I , black arrowheads ) , but levels were reduced ( white arrowheads ) ; anteroventrally , there was a reduction in hmx3a expression in presumed neuroblasts ( Fig 5F’ and 5G’ , blue and light blue arrowheads ) , and a mild posterior expansion of the spatial extent of expression for both genes in ventral otic epithelium ( Fig 5G’ and 5I’ , red arrowheads ) . Expression of the anterior markers pax5 and fgf3 was drastically reduced within anterior otic epithelium in hmx3aSU3/SU3 mutants at 24 hpf ( Fig 5J–5M’ , arrowheads ) , although expression of fgf3 in pharyngeal pouch endoderm ventral to the ear was unaffected ( Fig 5M and 5M’ , double-headed white arrows ) . Expression of the same markers in hmx3aSU3/SU3 mutants at 27 hpf was similar , but otic expression of hmx2 was more strongly reduced than that of hmx3a , especially in the anterior pole in the area corresponding to the normal expression domain of fgf3 and pax5 ( S10 Fig ) . Expression of the posterior marker fsta at 30 hpf did not reveal any significant duplication of expression in anterior otic epithelium in hmx3aSU3/SU3 mutant ears ( S10 Fig ) . Taken together , the results suggest that a loss of hmx3a function results in a similar phenotype to that of fgf3-/- ( liat21142/t21142 ) mutants [3 , 6 , 28] . Although some anteroposterior asymmetry has been lost , the phenotype is not as strong as the double-posterior duplications that result from inhibition of all Fgf signalling or over-activity of Hh signalling , which show a complete loss of the anterior macula and lateral crista , and duplication of elements of the posterior macula [6 , 12] . We conclude that hmx3a function is required for normal anterior otic expression of pax5 and fgf3 . However , loss of hmx3a function is not sufficient to result in a complete loss of anterior character and duplication of posterior structures at the anterior of the ear . As otic expression of the anterior markers pax5 and fgf3 is strongly reduced in hmx3aSU3/SU3 single mutants , and because expression of hmx3a is an early transcriptional response to manipulations of both Fgf and Hh signalling , we hypothesised that mis-expression of hmx3a alone would be sufficient to drive the expression of pax5 and fgf3 in the posterior of the otic placode and to give rise to a double-anterior duplication , bypassing the requirement for Fgf or Hh pathway manipulation . To test this idea , we created a transgenic line driving expression of the hmx3a coding sequence under the control of the hsp70 heat-shock promoter . A 60-minute heat shock of Tg ( hsp70:hmx3a ) embryos at 12 hpf resulted in a robust and widespread expression of the hmx3a transgene two hours later ( Fig 6A–6D ) . To avoid any disruption of otic placode induction , and to be comparable to the fgf3 heat-shock experiments , we heat-shocked Tg ( hsp70:hmx3a ) embryos at 14 hpf to induce systemic mis-expression of hmx3a . After 30 minutes at 39°C , heat-shocked embryos were incubated at 33°C for 30 minutes before being returned to 28 . 5°C and incubated until 22 . 5 hpf , when they were fixed for processing and analysis , or until 3 dpf for assessment of any ear duplication ( Fig 6E–6Q’ ) . Despite robust expression of the hmx3a transgene , the ears of Tg ( hsp70:hmx3a ) embryos heat-shocked for 30 minutes at 14 hpf did not recapitulate the duplicated double-anterior otic phenotype seen in Tg ( hsp70:fgf3 ) embryos . Position and number of the otoliths , morphology of the semicircular canal pillars and position of the sensory patches in heat-shocked embryos were normal at 3 dpf ( Fig 6E–6H’; compare with Fig 1B ) . At 22 . 5 hpf , otic vesicles were slightly smaller and rounder in heat-shocked transgenic embryos than those in heat-shocked non-transgenic siblings , but markers were expressed normally in most cases ( Fig 6I–6Q’ ) . Otic expression of hmx2 was mildly up-regulated in a graded fashion ( higher at the anterior ) in 5/16 transgenic embryos at 22 . 5 hpf ( Fig 6I–6K’ ) , similar to the de-repression of hmx expression seen after Hh inhibition in wild-type embryos . There was also a mild up-regulation of pax5 expression in posterior otic epithelium at 22 . 5 hpf in 3/20 embryos ( Fig 6L–6N’ ) , but pax5 was never expressed in a broad zone as in the heat-shocked Tg ( hsp70:fgf3 ) embryos . A weak patch of fgf3 expression appeared in posterior otic epithelium at 22 . 5 hpf , similar to the duplicated zone of endogenous fgf3 expression in heat-shocked Tg ( hsp70:fgf3 ) embryos , but in only 2/15 embryos ( Fig 6O–6Q’ ) . To check that the hmx3a transgene was functional , we sequenced cDNA from transgenic embryos , amplified with transgene-specific primers , which indicated that the open reading frame was intact ( S11 Fig ) . We also examined the phenotype of transgenic embryos after an even earlier heat shock , during otic placode induction ( 8–9 hpf and 10–11 hpf ) . Here , we saw a range of otic abnormalities in 80% of transgenic embryos ( n = 106 ) , including missing otoliths , but some embryos also had small heads and eyes ( S12 Fig ) . Ear patterning appeared normal in about 20% of transgenic embryos heat-shocked at these earlier stages . Longer ( 1- or 2-hour ) heat shocks at 14–15 hpf also resulted in normal otic patterning ( n = 49; S12 Fig ) . We conclude that the hmx3a transgene is likely to be functional , but that its mis-expression alone during otic placode stages ( 14–15 hpf , which should result in strong systemic expression until at least 17 hpf ) cannot substitute for Fgf mis-expression or Hh inhibition in the generation of a double-anterior duplicated ear . Up-regulation of hmx3a in the ear at later stages , beyond 18 hpf , was not sufficient either , as our late fgf3 heat shock experiments demonstrated ( S2 Fig ) . Taken together , our data and those from previously-published studies suggest a temporal hierarchy of events for otic anteroposterior patterning dependent on extrinsic sources of Fgf and Hh signalling ( Fig 7 ) . To assess whether this network of inferred genetic regulatory interactions can account for the dynamic expression patterns we observe , we developed a mathematical model of otic anteroposterior patterning in the wild-type ear and following manipulation of the Fgf and Hh signalling pathways . The model is based on a set of differential equations describing the genetic interactions in the otic epithelium outlined in Fig 7A . In addition , patterning in the model is dependent on the existence of two sources of spatial information . First , we assume that otic competence to express fgf genes in response to Fgf and Hmx3a protein is localised to the two poles of the developing otic vesicle . This is necessary in the model to ensure that induced endogenous fgf mRNA expression in the otic epithelium ( intrinsic fgf ( fgfi ) ) is restricted to the poles , even when fgf3 is expressed uniformly throughout the tissue following heat shock . Second , we represent the effect of fgf mRNA expression in rhombomere 4 as an anterior-to-posterior gradient of extrinsic Fgf ( Fgfe ) protein , present at high levels up to 30% of the otic vesicle length ( corresponding to the position of the rhombomere 4/5 boundary ) , and forming a decreasing spatial gradient across the remainder of the otic axis ( Fig 7A; S1 Model ) . Although we do not have a measure of actual Fgf protein concentration , our assumption is supported by measurements of fluorescence across the otic anteroposterior axis in the Tg ( dusp6:d2EGFP ) reporter line , which expresses a destabilised GFP variant as an indirect readout of Fgf activity [29] ( S13 Fig ) . We assume a spatially uniform level of Hh signalling throughout the otic epithelium ( see discussion in [11] ) , and that Hh signalling antagonises the effects of Fgf signalling on otic anterior marker genes ( fgfi , hmx3a and pax5 ) by increasing their response threshold for Fgf-induced expression . This functional attenuation is unlikely to be at the level of an immediate target of Fgf signalling such as etv4 , as Hh inhibition did not result in major changes to etv4 expression ( [6]; this work ) . One possibility is that it could reflect integration of activity of the two signalling pathways at the level of binding sites in the promoters of the otic anterior genes . In addition , we propose that Hmx3a , together with Fgf and Hh , regulates its own expression and that of other genes in the network . Currently , our data do not distinguish whether these regulatory relationships are direct or indirect . The dynamic behaviour of the model is presented in Fig 8 ( for full details , see S1 Model ) . In wild-type embryos ( Fig 8 , left-hand column ) , expression of hmx3a and pax5 is triggered in anterior otic tissue . The extent of expression is determined by the spatial reach of the extrinsic Fgf protein ( Fgfe ) gradient from rhombomere 4 . Although all cells in the model are competent to express the anterior markers hmx3a and pax5 , they do not receive sufficient Fgfe to do so at the posterior otic pole in a wild-type embryo . After transient heat shock-induced systemic mis-expression of fgf3 at 14 hpf ( Fig 8 , middle column ) , expression of both hmx3a and pax5 is induced across the entire anteroposterior axis . However , the ability of heat shock-induced fgf3 mis-expression to trigger endogenous intrinsic fgf ( fgfi ) expression requires the coincidence of both Fgf protein and competence to express fgfi at the poles , and so fgfi is not induced in the middle of the otic axis . After decay of heat shock-induced Fgf protein , expression of pax5 is lost from central regions , but is maintained at the poles by Fgf signalling from fgfi expression . By contrast , expression of hmx3a is maintained in central regions due to its autoregulation . De-repression of anterior markers after Hh pathway inhibition ( Fig 8 , right-hand column ) results from a lowering of the threshold for response to Fgf signalling , establishing duplicate expression domains of pax5 and fgf3 at the posterior pole . Thus , although both heat shock-driven mis-expression of fgf3 and inhibition of Hh signalling result in anterior duplications ( Fig 8; compare the patterns at the 36 hpf time point ) , the transient dynamics exhibited by the model at earlier time points are distinct .
The zebrafish otic placode is a convenient system in which to understand the gene network dynamics that lead to asymmetries along the axis of a developing organ . Asymmetries in gene expression are evident from early ( otic placode ) stages , but the system is clearly equipotential , since either a gain of Fgf signalling or a loss of Hh pathway activity at otic placode stages can produce remarkably similar double-anterior zebrafish ears at 3 dpf [6 , 11] . Interestingly , this final duplicated pattern arises via very different intermediate states in terms of gene expression patterns , depending on the signalling pathway that has been disrupted , as we have shown here . Mis-expression of fgf3 at 14 hpf leads to a rapid loss of asymmetry , with broad expansion of anterior otic markers across the entire anteroposterior axis of the ear within a few hours of heat shock-driven mis-expression . Expression of pax5 , which is required for normal development of the anterior ( utricular ) macula [3] , later resolves into two discrete domains . By contrast , initial asymmetries in gene expression persist for several hours after inhibition of Hh pathway activity , with new duplicate expression domains of anterior markers ( pax5 , fgf3 and fgf8a ) only appearing nearly a day later at the posterior otic pole . We have identified hmx3a as an early otic transcriptional response to manipulations of both signalling pathways . However , although a loss of hmx3a demonstrates its requirement for normal otic patterning , this does not result in a complete double-posterior duplication , and mis-expression of hmx3a does not appear to be sufficient to drive the formation of a double-anterior ear . Our data and mathematical model suggest that the Fgf/Hh system is sufficient to pattern the anteroposterior axis of the ear . In our scheme , there is only one input ( extrinsic Fgf activity ) that has a graded distribution across the otic anteroposterior axis . Notably , there is no need to infer an opposing graded input of extrinsic signalling activity that is high at the posterior of the ear . Although Retinoic Acid ( RA ) is thought to form such a gradient , and contributes to anteroposterior patterning in both the chick and zebrafish ear [30 , 31] , its activity can clearly be over-ridden by manipulations of Fgf or Hh signalling in generating either double-anterior or double-posterior zebrafish ears . Our model therefore differs from other models of axial patterning , for example in generation of dorsoventral pattern in the vertebrate neural tube . Here , information from two anti-parallel noisy gradients is integrated and refined by cross-repressing interactions between target genes , providing precise positional information along the axis [32 , 33] . However , the sufficiency of our network and model does not necessarily rule out a contribution from the RA gradient in generating correct anteroposterior patterning in the wild-type ear . At present , we do not have a full mechanistic explanation for the differences in response dynamics after manipulations of the Fgf and Hh signalling pathways . Although hmx3a responds rapidly to manipulation of Fgf signalling , its regulation may well be indirect; a recent study identified only one gene , Etv5 , as a direct up-regulated target of Fgf signalling during induction of otic-epibranchial precursor cells in the chick [34] . In zebrafish , transcription of etv4 and spry4 is known to be an early response to Fgf signalling [35–37] , with spry4 expression appearing within one hour of implantation of a bead coated with Fgf8 protein during epiboly stages [37] . Our work here shows that robust , systemic expression of etv4 occurs within two hours of the onset of heat shock in Tg ( hsp70:fgf3 ) embryos; we had previously shown strong expression of etv4 in the otic placode four hours after heat shock [6] . Thus , Etv4 is a good candidate for an immediate early transcriptional effector of Fgf signalling in our proposed genetic network . However , as etv4 mRNA expression is not strongly perturbed by Hh pathway inhibition ( [6]; this work ) , effects of Hh and Fgf on hmx3a expression are likely to be integrated further downstream , for example at the level of the hmx3a promoter . The slower response of hmx3a transcription to Hh inhibition might reflect the persistence of Hh pathway effectors , such as Gli activator proteins , which must be degraded before the effect of inhibiting Smoothened with cyclopamine can take effect . As the otic vesicle develops , additional levels of regulation are likely to contribute to the regulation of hmx3a and other genes in the network . For example , in the chick , regulation of Hmx3 expression in the dorsolateral otocyst has recently been shown to be influenced by both Shh and non-canonical BMP signalling through PKA and GLI3R [38] . In addition , negative feedback on otic fgf expression via sprouty genes [4] is likely to help to restrict gene expression to the poles and sharpen expression domain boundaries within the otic epithelium . One of the intriguing features of the double-anterior ears is that systemic mis-expression of an anteriorising factor ( fgf3 ) gives rise to two defined and separate anterior maculae with mirror-image symmetry , rather than establishing uniform anterior identity across the entire medial otic domain . The final duplicate pattern develops despite the initial broad expression of anterior markers after heat shock , which demonstrates that the entire medial side of the otic placode and vesicle is competent to express hmx genes and pax5 in response to Fgf signalling . However , a day after heat shock , expression of pax5 is lost from the centre of this domain and only maintained at the anterior and posterior poles of the otic vesicle , suggesting either that expression is subsequently repressed in the central domain , or that an intrinsic factor or factors is required to maintain expression at the poles . Attractive candidates for the latter role include atoh1a , which is expressed in discrete domains at the anterior and posterior otic poles at 14 hpf [9] . Atoh1a is thought to act in a positive feedback loop together with Fgf signalling in the zebrafish ear [9 , 39] . Fgf pathway activity is also observed at both poles of the otic vesicle at 24 hpf ( this work ) , 28 hpf and 50 hpf using a destabilised fluorescent transgenic reporter , Tg ( dusp6:d2EGFP ) [29] . We hypothesise that a positive feedback loop involving a pole-specific factor and all three fgf genes contributes to the maintenance of anterior-specific gene expression and generation of the double-anterior pattern . This builds on previous feedback models for anterior otic patterning and the regulation of otic pax5 expression [2 , 3] . A similar broad medial expansion of hmx3a and pax5 has been recently reported to result from systemic mis-expression of sox2 or sox3 at 12 . 5 hpf [40] . However , this early mis-expression results in a smaller and mis-shapen otic vesicle ( most likely due to a disruption of otic induction ) , and phenotypes were not followed beyond 30 hpf . It will be interesting to see whether a duplicated anterior pattern results from these manipulations . The otic phenotype of hmx3aSU3/SU3 single mutants closely resembles that of hmx3a/hmx2 double morphants [2] , and of fgf3t21142/t21142 mutants [3 , 6 , 28] . The similarity of the fgf3 and hmx3a otic mutant phenotypes suggests that a major role for the extrinsic Fgf3 signal is to activate hmx3a expression in anterior otic epithelium . Note that pharmacological inhibition of all Fgf signalling [6] or over-activity of the Hh pathway [12] both result in a stronger otic phenotype than in hmx3aSU3/SU3 mutants . The retention of some anteroposterior asymmetries in gene expression and the fused sensory macula in hmx3aSU3/SU3 mutants , together with the presence of the lateral crista , suggest that the hmx3aSU3/SU3 otic phenotype , like that of fgf3t21142/t21142 mutants , does not represent a complete double-posterior duplication . We also failed to see strong ectopic expression of the posterior marker fsta at the anterior of the ear in hmx3aSU3/SU3 mutants , although this is a less reliable indicator of posterior duplication; it is expressed at both poles of the ear following strong Fgf inhibition [6] , but lost altogether in the extreme double-posterior ears that can result from elevated Hh signalling [12] . Despite the similarities between the loss-of-function phenotypes for fgf3 and hmx3a in the zebrafish ear , the gain-of-function effects for each of the two genes are strikingly different . Whereas mis-expression of fgf3 at 14 hpf reliably generates a complete double-anterior ear , mis-expression of hmx3a at the same time point had very little effect on otic development . It is remarkable just how robust the embryo is to this kind of perturbation , considering that the systemic high levels of transgene expression must be energetically expensive to support . Indeed , there is usually some transient developmental delay after heat shock , but gross patterning of the ear at 3 dpf appeared normal in Tg ( hsp70:hmx3a ) heat-shocked embryos . Why , then , is hmx3a ineffective in establishing duplicate anterior development when mis-expressed ? It is possible that it needs to be delivered together with hmx2; the two genes are tightly linked on zebrafish chromosome 17 [41] , spatially co-expressed in the zebrafish otic vesicle ( although with different temporal onset ) [2] , and are known to have partially overlapping roles in the mouse ear [42] . A predicted hmx3b gene ( RefSeq XM_017358610 . 2 ) is also present in the zebrafish genome on chromosome 12 . If a second Hmx family protein or other binding partner was limiting , this might explain the lack of activity of the mis-expressed hmx3a transcript . Alternatively , Hmx3a could act as a competence factor , only functioning in the context of high Fgf or low Hh signalling to initiate duplicate anterior otic development . It is also possible that Fgf signalling abrogates an unidentified negative regulator of the otic anterior gene network at the same time as activating the expression of hmx3a . In the presence of such an inhibitor , systemic over-expression of hmx3a would be ineffective at activating the expression of genes such as hmx2 , pax5 and fgf3 in posterior otic domains . Anterior-specific otic expression of Hmx3 and Hmx2 , including their temporal order of expression onset in the ear , is conserved between zebrafish , mouse and chick [2 , 43–45] . Loss of Hmx3 function in the mouse causes a range of reported otic defects with variable penetrance and expressivity , which depend on the nature of the targeted mutant allele . A homozygous targeted deletion of exon 1 and part of exon 2 of Hmx3 resulted in a variable disruption of the lateral ( horizontal ) and posterior semicircular canal ducts , and loss of the lateral ( horizontal ) crista [46] . A weaker phenotype was seen after disruption of the homeodomain in exon 3 of Hmx3; in these mutants , all three semicircular canal ducts were present , but the lateral ( horizontal ) ampulla and crista were missing . The utricular and saccular maculae were juxtaposed in a common utriculosaccular chamber [42 , 44] , as we found in the zebrafish hmx3aSU3/SU3 mutant . A notable difference between the mouse and zebrafish mutants is the presence of all three cristae , including the lateral crista , in the zebrafish hmx3aSU3/SU3 mutants . Formation of the ventral pillar for the lateral canal was also present , although delayed . It will be interesting to see whether mutations in hmx2 ( not currently available ) affect morphogenesis of the zebrafish semicircular canal system; in the mouse , targeted disruption of Hmx2 results in a loss of all three semicircular canal ducts , with partial or complete loss of some ampullae and cristae , in addition to a fused utriculosaccular chamber [47] . In humans , HMX3 and HMX2 are located together , close to FGFR4 , on chromosome 10; hemizygous microdeletions that remove all three genes are thought to be causative for syndromes characterised by inner ear morphological anomalies , vestibular dysfunction and sensorineural hearing loss [48 , 49] . In conclusion , our study demonstrates that although Fgf gain-of-signalling and Hh loss-of-signalling produce similar morphological duplications of the zebrafish ear , they do so via distinct dynamical patterns of gene expression , providing valuable insights into normal anterior otic development . In addition , we determine that hmx3a , a gene expressed as an early transcriptional response to both Fgf and Hh manipulation , has a conserved role in correct separation of the sensory maculae within the otic vesicle , and is required—but not sufficient—for normal anterior otic development . We have also shown that our proposed genetic network for zebrafish otic anterior development can be recapitulated with a mathematical model that assumes interactions between a graded extrinsic source of Fgf , a uniform inhibitory influence of Hh , and equipotential competence to adopt an anterior identity at the otic poles . Interactions between these inputs and their downstream targets within the otic tissue ( hmx3a , hmx2 , pax5 and fgf genes ) lead to correct anteroposterior patterning in the developing zebrafish ear . The model will be a useful framework for further elucidation and functional validation of the proposed gene regulatory network required for the acquisition of anterior otic identity in the zebrafish .
All animal work in the Whitfield lab was covered by licencing from the UK Home Office ( PPL 40/3655 , P66302E4E ) . All zebrafish experiments conducted in the Lewis lab were approved by the Syracuse University Institutional Animal Care and Use Committee ( IACUC ) . Adult zebrafish ( Danio rerio ) were kept in circulating water at 28 . 5°C with a 14-hour light/10-hour dark cycle . The wild-type line used was AB; mutant alleles were hmx3aSU3 ( this work; see below for details ) , mibta52b [23–25] , and smohi1640Tg [19]; transgenic lines were Tg ( dusp6:d2EGFP ) [29] , Tg ( hsp70:fgf3 ) [13] , Tg ( hsp70:fgf8a ) ×17 [14] and Tg ( hsp70:hmx3a ) ( this work; see below for details ) . The Tg ( hsp70:fgf3 ) line was maintained on a mitfaw2/w2 background to reduce pigmentation . Embryos were staged as described [15] and incubated at 28 . 5°C in E3 ( 5 mM NaCl , 0 . 17 mM KCl , 0 . 33 mM CaCl2 , 0 . 33 mM MgSO4 , 0 . 0001% methylene blue ) , unless otherwise indicated . Embryos were cultured in E3 at 28 . 5°C prior to heat shock . For heat shock , embryos from either a cross between two hemizygous transgenic carriers , or an outcross between a transgenic carrier and a wild-type , were transferred to 25 ml of preheated E3 in a Falcon tube and incubated at 39°C for 30 minutes , unless otherwise indicated . Embryos were then returned to their original plates of E3 , which had been preheated to 33°C during the heat shock , and incubated for a further 30 minutes at 33°C . Plates were then returned to 28 . 5°C and incubated until embryos reached the desired stage for fixation . In heat-shock experiments with mixed batches of transgenic and non-transgenic embryos , a transgenic genotype was confirmed by expression of tdTomato in Tg ( hsp70:hmx3a ) embryos or abnormal shape of the yolk extension in Tg ( hsp70:fgf3 ) embryos , in addition to analysis of the phenotypes described in the text . Embryos were treated in 12-well plates ( 3 ml total volume; ≤30 embryos per well ) at 28 . 5°C with InSolution Cyclopamine , V . californicum ( Calbiochem ) . Chorions were punctured with a sterile hypodermic needle prior to treatment to improve compound penetration . After treatment , embryos were washed twice in E3 before either being fixed or incubated in E3 before fixation later . Vehicle-only controls consisted of an equivalent volume of the solvent ( ethanol ) to that used in the highest experimental treatment concentration . Embryos from the same batch ( siblings ) were randomly allocated into control and treatment groups . Embryos were dechorionated and fixed in 4% paraformaldehyde overnight at 4°C . In situ hybridisation was carried out as described [50] . For most experiments , at least 25 embryos ( biological replicates ) were stained in any given batch . Where relevant , numbers of embryos with the phenotype of interest and total number in the batch ( e . g . 29/30 ) are shown directly on the figure panels ( see figure legends for details ) . Analysis of gene expression via in situ hybridisation is not quantitative , but we have chosen markers that give a clear and robust qualitative response to changes in signalling pathway activity . We have used information from these spatial expression patterns to infer parameters for the mathematical model ( see below ) . Where appropriate , we have measured the spatial extent of expression along the medial side of the otic vesicle in a dorsal view using ImageJ . The 3’ UTR of fgf3 was amplified from wild-type ( AB strain ) genomic DNA in a nested PCR , incorporating the T7 promoter , using the following primers: F1 TCTCTTGACACAGATGGAGATCC , R1 AATATACAAAGTACTCCTGATTGCA; F2 AAGGCCACTGAGAGTCCAAAA , T7-R2 TAATACGACTCACTATAGGGCAGTAGCCTATCACATGTACGT . Each PCR was run for 30 cycles with an annealing temperature of 53°C . The single guide RNA ( sgRNA ) targeting hmx3a was designed using CHOPCHOP [51 , 52] . The sgRNA DNA template was generated using the cloning-free method of Gagnon and colleagues [53] . The template was transcribed and purified using the standard protocols of the MEGAshortscript T7 kit ( AM1354 , Thermo Fisher Scientific ) . sgRNA was resuspended in 40 μl of sterile water and the concentration and purity measured using spectrophotometry , before aliquoting for storage at -80°C . To make Cas9 mRNA , pCS2-nls-zCas9-nls plasmid DNA [54] was digested with NotI and purified by phenol:chloroform extraction , before being transcribed and purified using standard protocols of the mMESSAGE mMACHINE SP6 kit ( AM1340 , Thermo Fisher Scientific ) . The resultant mRNA was resuspended , assayed and stored as for the sgRNA . The single cell of one-cell stage AB wild-type embryos was injected with 2 nl of a mixture of 200 ng/μl sgRNA + 600 ng/μl nls-ZCas9-nls mRNA . Founders were identified by high resolution melt analysis , using the following primers: PMA F: CGAATGCTAATTTGGCCTCTATTACT and PMA R: TTTTGTTGTCGTCTTCATCGTCC , and Precision Melt Supermix for High Resolution Melt ( HRM ) Analysis ( 172–5112 , Bio-Rad ) , performed on a CFX96 Touch System ( 1855195 , Bio-Rad ) , equipped with Precision Melt Analysis Software ( 1845025 , Bio-Rad ) . Amplification data were generated using the following program: 95 . 0°C for 3 minutes , followed by 45 cycles of 95 . 0°C for 15 seconds , 60 . 0°C for 20 seconds and 70 . 0°C for 20 seconds . Melt data were generated using the following program: 65 . 0°C for 30 seconds , 65 . 0°C–95 . 0°C at an incremental rate change of 0 . 2°C , held for 5 seconds each step , 95 . 0°C for 15 seconds . Stable F1 heterozygous fish were confirmed by sequencing . All subsequent genotyping was performed by PCR , using the primers F: TGGCAAAGTGACACGACCAG and R: GAGAACACCGTGCGAGTTTTC , Taq DNA Polymerase ( M0320S , NEB ) and the PCR program: ( 94 . 0°C for 2 minutes , 35 cycles of: 94 . 0°C for 30 seconds , 64 . 9°C for 30 seconds and 72 . 0°C for 30 seconds , followed by a final extension at 72 . 0°C for 2 minutes ) . The hmx3aSU3 allele is a 69 bp insertion , flanked on either side by 2-base mismatches . The insertion introduces a premature stop codon at nucleotides 352–354 of the edited coding sequence . The insertion in the mutant allele can be distinguished by performing gel electrophoresis on a 2% TBE agarose gel ( 100V for 40 minutes ) . The wild-type allele generates a 331 bp product , compared with the 400 bp mutant allele product . The zebrafish hmx3a cDNA sequence ( RefSeq NM_131634 . 2 ) , including the complete open reading frame , endogenous Kozak sequence and 15 bp of 3’ UTR , was cloned into a Tol2-containing ubi:tdTomato destination vector , flanked by a 5’ hsp70 promoter and a 3’ SV40 late polyadenylation signal sequence , using the Tol2kit [55] ( Invitrogen ) . 50 ng of this construct were injected into one-cell stage embryos together with 50 ng of in vitro-transcribed transposase RNA . Injected embryos ( G0 ) were raised to adulthood , and their progeny ( F1 ) screened for expression of the tdTomato marker . F1 embryos with positive expression were raised to adulthood to generate a stable Tg ( hsp70:hmx3a ) transgenic line . Progeny were tested by in situ hybridisation after heat shock to check for misexpression of the hmx3a transgene . To check that the hmx3a coding sequence was intact , cDNA was generated from heat-shocked embryos from an in-cross of Tg ( hsp70:hmx3a ) fish by RT-PCR and sequenced . PCR primers ( F: TACGACTCACTATAGGGCGAATTG; R: GCAATTAACCCTCACTAAAGGGAA ) were designed to be transgene-specific , binding to residual multiple cloning site sequences within the integrated construct , with a predicted amplicon size of 1053 bp . Sequences were aligned and displayed using Ensembl MUSCLE and ExPasy BOXSHADE . Embryos were fixed in 4% PFA overnight , washed in PBS ( 3×10 minutes ) and permeabilised in 2% Triton-X100 ( Sigma ) for 3–4 days at 4°C . Following further washes in PBS ( 3×5 minutes ) , embryos were stained with FITC-phalloidin ( 1:20; Sigma ) or Alexa Fluor 647-phalloidin ( 1:100; Thermo ) in PBS ( overnight , 4°C ) . Embryos were washed in PBS ( 3×60 minutes ) , dissected in PBS and mounted in Vectashield ( Vectorlabs ) prior to confocal imaging . Live and fixed embryos were imaged on either an Olympus BX51 or a Zeiss Axio Imager M1 compound microscope using brightfield , DIC and epifluorescence optics as appropriate , and CellB or Axiovision image acquisition software , respectively . For confocal imaging , either a Nikon A1 or a Zeiss LSM 710 confocal microscope was used . For fluorescent imaging requiring large fields of view , a Zeiss Axio Zoom . V16 stereomicroscope with Zen acquisition software was used . The S1 Movie and associated still images shown in S13 Fig were acquired with a Zeiss Z . 1 light-sheet microscope . Sample drift was corrected using the Manual Drift Correction plugin within FIJI ( Fiji Is Just ImageJ ) [56] . FIJI was used for all image processing . Figure panels were assembled using Adobe Photoshop 2015 . 5 . 0 . All dorsal views ( except in S13 Fig ) are shown with anterior to the top; lateral views show anterior to the left . Statistical analyses were performed using GraphPad Prism version 7 . 0c for Mac OSX ( GraphPad software , La Jolla California USA , www . graphpad . com ) . See figure legends and S1 Data for details . Full information for generation of the mathematical model , including a list of parameters used , is given in S1 Model .
|
Understanding how signalling molecules impart information to developing organ systems , and how this is interpreted through networks of gene activity , is a key goal of developmental genetic analysis . In the developing zebrafish inner ear , differences in gene expression arise between the anterior and posterior poles of the ear placode , ensuring that sensory structures in the ear develop in their correct positions . If signalling pathways are disrupted , a mirror-image ear can result , developing with two anterior poles . We have used genetic , pharmacological and mathematical modelling approaches to decipher the pathway of gene action required to specify anterior structures in the zebrafish ear . Patterns of gene expression are dynamic and plastic , with two different routes leading to the formation of duplicate anterior structures . Expression of the hmx3a gene is an early response to the anterior signalling molecule Fgf3 , but is not sufficient to drive the formation of ectopic anterior structures at the posterior of the ear . The hmx3a gene codes for a protein that regulates other genes , and in humans , mutation of HMX genes results in diseases affecting inner ear function . Our work provides a framework for understanding the dynamics of early patterning events in the developing inner ear .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"fish",
"otolith",
"vesicles",
"ears",
"vertebrates",
"animals",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"inner",
"ear",
"model",
"organisms",
"experimental",
"organism",
"systems",
"embryos",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"embryology",
"animal",
"studies",
"gene",
"expression",
"biological",
"tissue",
"head",
"hedgehog",
"signaling",
"signal",
"transduction",
"zebrafish",
"eukaryota",
"anatomy",
"cell",
"biology",
"genetics",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cell",
"signaling",
"organisms"
] |
2019
|
Anteroposterior patterning of the zebrafish ear through Fgf- and Hh-dependent regulation of hmx3a expression
|
St . Louis encephalitis virus is a complex zoonoses . In 2005 , 47 laboratory-confirmed and probable clinical cases of SLEV infection were reported in Córdoba , Argentina . Although the causes of 2005 outbreak remain unknown , they might be related not only to virological factors , but also to ecological and environmental conditions . We hypothesized that one of the factors for SLE reemergence in Córdoba , Argentina , was the introduction of a new SLEV genotype ( SLEV genotype III ) , with no previous activity in the area . In order to evaluate this hypothesis we carried out a molecular characterization of SLEV detections from mosquitoes collected between 2001 and 2004 in Córdoba city . A total of 315 mosquito pools ( 11 , 002 individuals ) including 12 mosquitoes species were analyzed . Overall , 20 pools ( 8 mosquitoes species ) were positive for SLEV . During this study , genotypes II , V and VII were detected . No mosquito pool infected with genotype III was detected before the 2005 outbreak . Genotype V was found every year and in the 8 sampled sites . Genotypes II and VII showed limited temporal and spatial activities . We cannot dismiss the association of genotype II and V as etiological agents during the outbreak . However , the silent circulation of other SLEV strains in Córdoba city before the 2005 outbreak suggests that the introduction of genotype III was an important factor associated to this event . Not mutually exclusive , other factors such as changes in avian hosts and mosquitoes vectors communities , driven by climatic and environmental modifications , should also be taken into consideration in further studies .
The St . Louis encephalitis ( SLE ) , caused by the homonymous virus ( SLEV , genus Flavivirus , family Flaviviridae ) , is a complex zoonoses in the New World [1] . In South America , SLE is an emerging arbovirosis , with febrile illness and encephalitis cases reported in Argentina and Brazil [2] , [3] . SLEV reemerged in the central region of Argentina during 2002 [2] and , since then , outbreaks have been reported in Córdoba ( 2005 ) [4] , Entre Rios ( 2006 ) , Buenos Aires ( 2010 ) and San Juan provinces ( 2011 ) [5] . SLEV shows biological and molecular variability among strains isolated throughout its geographic distribution [6]–[8] . Based on complete Envelope gene sequencing , SLEV strains can be classified into 8 lineages or genotypes ( I–VIII ) [9] . According to a phylogeographic analyses , genotypes I and II are prevalent in the USA , while the others were found only in countries from Central and South America [9]–[11] . Exceptionally , genotype V strains were recently isolated in Florida [12] . SLEV strains circulating in Argentina were clustered with genotype III ( 79V-2533 –year 1978- , CbaAr-4005 , CbaAr-4006 – year 2005- ) , V ( 78V-6507 – year 1978- ) , both isolated from Culex spp . Mosquitoes , and VII ( CorAn-9124 , CorAn-9275 – year 1966- ) , isolated from rodents [9] , [13] . In USA , the transmission cycles of SLEV are maintained by Culex mosquito vector species ( Culex quinquefasciatus , Cx . tarsalis and Cx . nigripalpus ) and Columbiformes ( Mourning doves - Zenaida macroura ) and Passeriformes ( House finches - Carpodacus mexicanus , House sparrows - Passer domesticus ) bird host species . Humans and mammals represent a dead-end host for the virus . Although it is widely distributed in the American continent , its ecology is scarcely known outside the USA [1] . Available data in Argentina suggests that Cx . quinquefasciatus mosquito would act as a main vector , while Picui ground dove ( Columbina picui ) and Eared dove ( Zenaida auriculata ) would be important hosts in urban and rural environments [13]–[16] . However , an alternative transmitting rodent-mosquito cycle was postulated in Argentina for genotype VII SLEV strains ( CorAn-9124 , CorAn-9275 ) , isolated from rodents [14] . SLEV reemerged in the central region of Argentina in 2002 [2] . In 2005 , 47 laboratory-confirmed and probable clinical cases of SLEV infection , including nine fatalities , were reported in the central Córdoba Province [4] . During this outbreak two SLEV genotype III strains were isolated [13] . This was the first SLEV-induced encephalitis outbreak reported in South America . Although the causes of 2005 outbreak remain unknown , they might be related not only to virological factors , but also to changes in the structure and dynamics of vectors and/or avian amplifying hosts populations , and environmental conditions . We hypothesized that one of the factors for SLE reemergence in Córdoba , Argentina was the introduction of a new SLEV genotype , with no previous activity in the area . In order to evaluate this hypothesis we carried out a molecular characterization of SLEV detections from mosquitoes collected between 2001 and 2004 in Córdoba city .
Mosquito collections were carried out during 2001 to 2004 in the city of Córdoba ( 31°24′30″S , 64°11′02″O ) ( Córdoba Province , Argentina ) , just before the 2005 SLE local outbreak . The city is situated at 450 m over sea level and its extension covers 576 km2 , of which 37 . 2% is urbanized . The population is 1 , 330 , 023 inhabitants ( http://200 . 51 . 91 . 231/censo2010/ ) . The area belongs to the phytogeographic region of the Espinal , Chaqueño Domain , intensively modified by human activities ( urbanization , agriculture , cattle , and industry ) . The city is surrounded by land crops ( soy , fruit tree ) , industries and autochthonous shrubs patches isolated . The climate is mild without warm winters and with water deficit , in spite of its relatively high precipitation levels ( 750 and 800 mm ) , due to high evapotranspiration . A total of 8 collection sites were selected based on accessibility , owners' authorizations and previous evidence of mosquitoes abundance; most of them are located in peripheral areas of the city of Córdoba ( Figure 1 ) . Mosquitoes were collected using CDC light traps ( supplemented with dry ice ) and chicken and rabbit baited can traps . Three light traps and 2 baited traps were set up and maintained in each site during 2 nights per season . Traps remained active during 18:00 until 09:00 . The trapping schedule is detailed in Table 1 . Collected mosquitoes were transported alive in refrigerated containers to the laboratory . Individuals were identified on a chill table and sorted by species , sex , collection date and site , and non-engorged and engorged status . Mosquitoe pools were homogenized using pestles and mortars in minimum essential medium ( MEM ) supplemented with 10% fetal bovine sera ( FBS ) , 1% gentamicine and 1% Fungizone . Pools containing 1–25 mosquitoes were homogenized in 1 ml of MEM and those containing 25–50 mosquitoes were added 2 ml of MEM . Homogenates were centrifuged at 11 , 400 g during 30 min at 4°C for decontamination . Supernatants were stored in tubes at −70°C until utilization . RT PCR SLEV positive mosquito pools were subjected to viral isolation . One hundred µl of mosquito homogenate were inoculated onto 24 hs VERO cells monolayers , incubated for 60 min at 37°C , and observed daily for cythopatic effect . After 7 day post inoculation , blind passages were realized .
A total of 315 mosquito pools ( 11 , 002 individuals ) including 12 mosquitoes species were analyzed . Overall , 20 pools ( 8 mosquitoes species ) were positive for SLEV ( Table 2 ) . The BLASTn search and the phylogenetic analyses were performed using a 212 bp fragment of the Envelope protein . Viral isolation attempts were unsuccessful due to probable multiple freeze/thaws cycles during the process of mosquito homogenates . Based on the classification proposed by Kramer and Chandler [9] , three different SLEV genotypes ( II , V , and VII ) were detected in Córdoba city before the encephalitis human outbreak associated to SLEV genotype III ( Figure 2 ) . Genotype V was detected in 5 of the 8 collection sites during the 4 years ( Figure 3 ) and it was found in the 8 infected mosquito species collected in our study ( Aedes aegypti , Anopheles albitarsis , Ae . albifasciatus , Ae . scapularis , Culex apicinus , Cx . interfor , Cx . quinquefasciatus , Psoropohora ferox ) ( Table 2 ) . Sequences belonging to genotype V share a 100% of homology with SLEV strain 78V6507 ( AF205481 ) . This genotype was previously isolated from Cx . quinquefasciatus mosquitoes in Santa Fe province in 1978 [20] . Genotypes II and VII were detected only sporadically during the study . However , this is the first report of SLEV genotype II in Argentina . The nucleotide sequence corresponding to this genotype share 100% homology with SLEV strains SPAn-11916 ( EF117302 ) , was isolated in 1968 from rodents collected in Sao Paulo ( Brazil ) , and Parton-MSI-7 ( EF158070 ) was isolated from ill humans in Missouri ( USA ) in 1933 . Only 2 positive pools ( Ae . albifasciatus and Cx . quinquefasciatus mosquitoes ) collected in Bajo Grande and Guiñazú in 2003 proved to be infected with this genotype in our study ( Figure 3 ) . Sequences clustering with Genotype VII have a 100% of homology with SLEV strains CorAn-9124 ( EF158063 ) and CorAn-9275 ( EF158068 ) isolated in 1966 . Although genotype VII has been isolated from small rodents ( Calomys musculinus , Mus musculus ) in the province of Córdoba [14] , there was no previous report of its activity neither in mosquitoes nor in the capital city . ; it was detected in 3 sites ( Guiñazú , Bajo Grande , San Carlos ) during 2002 and 2003 ( Table 2 , Figure 3 ) . Mosquitoes infected with genotype VII were identified as Ae . albifasciatus , Ae . scapularis and Cx . interfor .
Three SLEV genotypes circulated simultaneously in Córdoba city between 2001 and 2004 . None of these genotypes are related to the SLEV Genotype III CbaAr-4005 strain isolated during the encephalitis human outbreak that occurred later in 2005 [13] . The most prevalent variant was genotype V , genetically related to SLEV 78V-6507 strain isolated in Santa Fe province in 1978 [20] . Our results confirm the endemicity of SLEV in Córdoba city during 2001–2005 . The sustained activity of genotype V in Bajo Grande , Botanical Garden , Guiñazú , Pediatric Hospital , and San Carlos between 2001 and 2004 ( Figure 1 ) indicates local genotype persistence and probable overwintering of SLEV activity in this temperate area . Culex mosquitoe abundance decreases drastically during winter , so vector transmission to vertebrates would not be maintained during this season . Supporting endemicity of SLEV , Flores et al . [21] detected vertical transmission of SLEV genotype V in local Culex mosquito populations under laboratory conditions . We believe that this mechanism would satisfactory explain the maintenance of some viral variants until the next favorable season for mosquito vector proliferation . The presence of a predominant genotype could be evidence for a higher viremogenic capacity of this strain in avian hosts . Higher viremias would enhance the transmission by mosquitoes , increasing viral circulation and expanding the distribution of this variant . In fact , SLEV genotype V ( strain 78V-6507 ) developed higher viremias in birds than genotype III and VII strains [7] . SLEV not only requires the ability to infect avian hosts with highly enough viremias to be deemed infectious to mosquitoes , but it also needs to be able to disseminate the mosquito mid-gut and enter the salivary glands . This ecological and evolutionary feature should not be unattended , indeed . In a different way , genotypes II and VII showed limited temporal and spatial activities ( Figure 1 ) . Powers et al . [22] pointed out that arboviruses maintained by a rodents-mosquitoes cycle show limited and constrained geographic distributions , which agree with the limited dispersion capacity observed in rodents . In fact , Genotype VII SLEV strains ( CorAn-9124 , CorAn-9275 ) were isolated only from small rodents in Córdoba province [14] , which suggests that mammals are the actual hosts for these strains . Moreover , the 3 mosquito species ( Ae . albifasciatus , Ae . scapularis and Cx . interfor ) infected with Genotype VII frequently feed on mammals hosts [20] , [23]–[25] . Our data support the hypothesis that some SLEV strains are being maintained through alternative rodents-mosquitoes transmission cycles [14] . Analyzing temporal and geographic patterns of SLEV in Texas by molecular techniques , Chandler et al . [26] detected the presence of dominant and limited strains fluctuating in space and time . This activity pattern could be the result of multiple intermittent virus introductions by birds from neighboring regions [27] . Other flavivirus such as Japanese encephalitis ( JEV ) and West Nile virus ( WNV ) showed similar dynamics [28] , [29] . These dynamics are characterized by the introduction/persistence of certain genotypes and the presence/absence of clinical cases in the study area . Another possibility is that SLEV genotype III had been silently present in Córdoba prior to 2001 . Although a Dengue virus has different ecological requirements compared with JEV , SLEV and WNV , it has been demonstrated that some serotypes can remain silent for many years causing periodic epidemics [30] . During 2005 , a human encephalitis outbreak was caused by SLEV in Córdoba city [4] , and two genotype III strains were isolated [13] . Molecular characterization determined that both variants are closely related to SLEV strain 79V-2533 , isolated 27 years ago in Santa Fe province [13] . This evidence supports the hypothesis that the introduction of a more virulent genotype could have caused the mentioned outbreak . In a recent study , Diaz et al . [31] compared epidemic ( CbaAr-4005 ) and non-epidemic ( 79V-2533 ) genotype III SLEV strains . The epidemic variant produced higher viremias in House sparrows ( Passer domesticus ) than the non-epidemic strain . According to this , the epidemic strain ( CbaAr-4005 ) appears to broaden the number of avian species that are likely to be competent amplifying hosts relative to the non-epidemic 79V-2533 strain [31] . The introduction of new strains and the extension of their geographic distributions are factors that can cause the emergence and reemergence of flaviviruses in different regions [32] . For example , the introduction , spread and establishment of WNV in America and Japanese Encephalitis Virus in Australasia and the annual introduction of SLEV strains in the USA ( California state ) [27] . During our four-year study a non-genotype III SLEV strain was detected . One year prior to the outbreak a total of 2 , 093 mosquitoes were analyzed and only one mosquito pool was positive ( genotype V ) ( Table 1 ) . The results here exposed suggest that genotype III was introduced in Córdoba city a few months before the outbreak and it could be one of the factors contributing to the outbreak . SLEV strains show biological and molecular variability [6]–[9] . Strains belonging to genotypes I , II , III , and V showed pathogenicity in mice and Rhesus monkeys [6] . During SLEV human encephalitis outbreaks in the USA , genotypes I and II were frequently isolated [9] . Spinsanti et al . [2] confirmed one human encephalitis case in Córdoba concomitant with our detection of Genotype V in the same area during the same year ( Pediatric Hospital , 2002 – Figure 1 ) . Although genotype III was detected in mosquitoes collected around the encephalitis human cases during the 2005 outbreak , we cannot dismiss the association of genotype II and V as etiological agents during the outbreak . However , the silent circulation of other SLEV strains in Córdoba city before the 2005 outbreak suggests that the introduction of genotype III was an important factor associated to this event . Not mutually exclusive , other factors such as changes in avian hosts and mosquitoes vectors communities , driven by climatic and environmental modifications , should also be taken into consideration in further studies .
|
The St . Louis encephalitis is a complex zoonoses in the New World . In South America ( Argentina and Brazil ) , SLE is an emerging arbovirosis . SLEV reemerged in Argentina during 2002 and , since then , outbreaks have been reported in 2005 , 2006 , 2010 and 2011 . During the 2005 outbreak two SLEV genotype III strains were isolated . Although the causes of the 2005 outbreak remain unknown , they might be related not only to virological factors , but also to changes in the structure and dynamics of vectors and/or avian amplifying hosts' populations and environmental conditions . We hypothesized that one of the factors for SLE reemergence in Córdoba , Argentina , was the introduction of a new SLEV genotype , with no previous activity in the area . No mosquitoes were detected infected with genotype III during this four-year study , even 10 months before the outbreak . The silent circulation of other SLEV strains in Córdoba city before the 2005 outbreak suggests that the introduction of genotype III was an important factor associated to this event . Not mutually exclusive , other factors such as changes in avian hosts and mosquitoes vectors communities , driven by climatic and environmental modifications , should also be taken into consideration in further studies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"virology",
"biology",
"microbiology"
] |
2012
|
Silent Circulation of St. Louis Encephalitis Virus Prior to an Encephalitis Outbreak in Cordoba, Argentina (2005)
|
In response to environmental and physiological changes , the synapse manifests plasticity while simultaneously maintains homeostasis . Here , we analyzed mutant synapses of henji , also known as dbo , at the Drosophila neuromuscular junction ( NMJ ) . In henji mutants , NMJ growth is defective with appearance of satellite boutons . Transmission electron microscopy analysis indicates that the synaptic membrane region is expanded . The postsynaptic density ( PSD ) houses glutamate receptors GluRIIA and GluRIIB , which have distinct transmission properties . In henji mutants , GluRIIA abundance is upregulated but that of GluRIIB is not . Electrophysiological results also support a GluR compositional shift towards a higher IIA/IIB ratio at henji NMJs . Strikingly , dPAK , a positive regulator for GluRIIA synaptic localization , accumulates at the henji PSD . Reducing the dpak gene dosage suppresses satellite boutons and GluRIIA accumulation at henji NMJs . In addition , dPAK associated with Henji through the Kelch repeats which is the domain essential for Henji localization and function at postsynapses . We propose that Henji acts at postsynapses to restrict both presynaptic bouton growth and postsynaptic GluRIIA abundance by modulating dPAK .
Coordinated action and communication between pre- and postsynapses are essential in maintaining synaptic strength and plasticity . Presynaptic strength or release probability of synaptic vesicles involves layers of regulation including vesicle docking , fusion , and recycling , as well as endocytosis and exocytosis . Also , how postsynapses interpret the signal strength from presynapses depends largely on the abundance of neurotransmitter receptors at the synaptic membrane [1 , 2] . During long-term potentiation , lateral diffusion of extrasynaptic α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor ( AMPAR ) to synaptic sites is accelerated [3 , 4] and the exocytosis of AMPAR is enhanced near the postsynaptic density ( PSD ) , causing an accumulation of synaptic receptors [5 , 6] . In contrast , under the long-term depression condition , synaptic AMPAR is reduced by hastened endocytosis [7 , 8] . While molecular mechanisms are proposed to play roles in regulating and fine-tuning postsynaptic glutamate receptor ( GluR ) abundance in plasticity models , the developmental regulation of GluR abundance at the synaptic surface still needs to be elucidated . Synapses at the Drosophila neuromuscular junction ( NMJ ) use glutamate as the neurotransmitter , and have properties reminiscent of mammalian central excitatory synapses [9 , 10] . Homologous to vertebrate AMPAR and kainate receptors , Drosophila GluR subunits assemble as tetramers to gate ion influx [11] . Each functional receptor contains essential subunits ( GluRIIC , GluRIID and GluRIIE ) and either GluRIIA or GluRIIB; therefore , synaptic GluRs can be classified according to their subunit compositions as either A- or B-type receptors [12–17] . These two types of receptors exhibit distinct developmental and functional properties . Newly-formed PSDs tend to accumulate more GluRIIA channels , while the IIA/IIB ratio becomes more balanced when PSDs mature [18] . In addition , GluRIIB channels have much faster desensitization kinetics , which results in smaller quantal size than GluRIIA channels [12] . Therefore , the synaptic composition of these two types of GluRs greatly influences the postsynaptic interpretation of neuronal activities . The Drosophila homolog of p21-activated kinase ( dPAK ) regulates GluRIIA abundance at the PSD; GluRIIA receptor clusters at the postsynaptic membrane are strongly reduced in dpak mutants [19] . However , overexpression of dPAK in postsynapses is not sufficient to increase GluRIIA cluster size , suggesting that dPAK activity in regulating GluRIIA abundance is tightly controlled . Ubiquitination and deubiquitination play critical roles in regulating synaptic functions [20–24] . In loss-of-function mutants for highwire , a gene encoding a conserved E3 ubiquitin ligase , NMJs overgrow , producing supernumerary synaptic boutons [25 , 26] . This phenotype is duplicated by overexpression of the deubiquitinating enzyme Fat facets ( Faf ) in presynapses [27] . These studies underline the importance of balanced ubiquitination in synapse formation and function . Cullin-RING ubiquitin ligases ( CRLs ) are large protein complexes that confer substrate ubiquitination [28 , 29] . Importantly , CRLs promote ubiquitination through substrate receptors that provide specific recognition of substrates for ubiquitination . The BTB-Kelch proteins are suggested to be the substrate receptors for Cul3-scaffolded CRLs [30–32] . In this study , we identified a BTB-Kelch-containing protein , Henji , also known as Dbo [33] , which regulates NMJ growth and synaptic activity by restricting the clustering of GluRIIA . Synaptic size of henji mutants was significantly expanded , as viewed under transmission electron microscopy ( TEM ) . Immunostaining for dPAK and GluRIIA also suggests larger areas of PSDs in the absence of Henji , and the intensity of each fluorescent punctum becomes stronger , indicating abnormal accumulation of these PSD proteins . By genetically reducing one gene dosage of dpak in henji mutants , GluRIIA accumulation and abnormal bouton morphology was suppressed . In contrast , reducing the gluriia gene dosage in henji mutants restored bouton morphology but failed to suppress dPAK accumulation . Thus , Henji regulates bouton morphology and GluRIIA clustering levels likely through a control of dPAK . Interestingly , while overexpression of dPAK , either constitutively active or dominantly negative , had no effects on GluRIIA clustering , overexpression of these dPAK forms in henji mutants modulated GluRIIA levels , indicating that Henji limits the action of dPAK to regulate GluRIIA synaptic abundance . Henji localized to the subsynaptic reticulum ( SSR ) surrounding synaptic sites , consistent with the idea that Henji functions as a gatekeeper for synaptic GluRIIA abundance .
We are interested in how ubiquitination might regulate synaptic function through controlling specific synaptic proteins . As putative substrate receptors of Cul3-based E3 ubiquitin ligases , each BTB-Kelch protein could recognize one or multiple synaptic proteins to regulate their abundance and thus their synaptic functions [31] . The Drosophila genome encodes 16 BTB-Kelch proteins ( S1 Fig ) and those with available RNAi or P-element insertion lines were examined for NMJ morphological abnormality . By immunostaining for horseradish peroxidase ( HRP ) and Synapsin to reveal NMJ morphology , we found the P-element insertion ( PBac{PB}dboc04604 ) in the CG6224 locus induced satellite boutons ( see below ) . Whereas CG6224 is known to encode Dbo [34] , given the supernumerary bouton morphology , we named this mutant “henji” , meaning “very crowded” in Mandarin , and this name is used in this study . To study Henji function in vivo , we generated mutants by P-element excision in the annotated CG6224/dbo locus ( S2A Fig ) . Two excision mutants with truncation of the shared promoter region between CG6224 and CG6169 were lethal . Addition of the CG6169 genomic transgene rescued the lethality of these deletions , which are named henji1 and henji8 with disruption specifically in henji expression ( S2A Fig ) . Indeed , the henji mRNA expressions analyzed by reverse transcription PCR ( RT-PCR ) showed lower levels in all three henji mutants used in this study , with a medium level in the PBac{PB}dboc04604 insertion line ( named as henjiP in this study ) , a low level in henji1 , and an almost undetectable level in henji8 ( S2B Fig ) . As the protein translation start site was deleted in henji8 , we conclude that henji8 is a null allele , henji1 a strong loss-of-function allele , and henjiP a hypomorphic allele . At wild-type ( WT ) NMJs , each bouton is connected to adjacent boutons through linear or bifurcated branches . However , at henji NMJs , multiple smaller boutons emerged from large parental boutons ( Fig 1B , insets ) . These smaller boutons , defined as satellite boutons , usually resulted from more than three buds emanating from a parental bouton [35 , 36] . Satellite boutons were rarely found at WT NMJs , but were a prominent feature in all henji mutants we had examined , including henji1/1 and henji1/8 ( Fig 1B ) . The number of boutons with normal size , however , was either slightly reduced in henji1/8 or remained normal in henji1/1 ( Figs 1C and S2C ) , suggesting that the formation of satellite boutons is not at the expense of normal boutons . A bouton houses tens of synaptic sites where the presynaptic active zones ( AZs ) opposes the postsynaptic PSDs . We first examined PSD structure and PSD-localized GluR clusters in henji mutants by co-immunostaining with antibodies against PSD-specific dPAK and the GluRIIA subunit ( Fig 1D ) . In WT , dPAK localized as well-separated puncta . In the henji1/8 mutant , the area of individual puncta was expanded and the immunofluorescent intensity was enhanced ( Fig 1D ) . When normalized to co-stained HRP , the dPAK immunofluorescent intensity and punctum size were increased while the density of dPAK puncta was normal , as compared to WT control ( Fig 1E , upper bar graphs ) . As dPAK is required for PSD formation and regulates GluRIIA cluster formation [19] , the increase in dPAK levels and patch size suggests a possible enlargement of the PSD that houses GluR clusters . Consistently , GluRIIA immunopositive puncta increased in both intensity and size in the henji1/8 mutant ( Fig 1D and 1E ) . The increases in dPAK and GluRIIA immunointensities were also detected in henji1/1 ( Fig 2A and 2B ) . GluRIIB immunostaining signals in henji1/8 , however , showed no significant difference in the intensity to WT control ( S2D Fig ) . These results suggest that Henji regulates the dPAK level at PSDs and specifically confines GluRIIA cluster size . At NMJs lacking henji , the increase in GluRIIA but not GluRIIB abundance leads to a shift in the synaptic GluRIIA/GluRIIB ratio . To examine if henji is responsible for the defects observed in henji mutants , we generated a genomic rescue construct in which GFP was fused to Henji at the N-terminus , and the GFP-henji transgene is driven by the endogenous henji promoter . When introduced into henji1/8 , GFP-henji restored dPAK and GluRIIA to near WT levels ( Fig 1D ) . The intensities of dPAK and GluRIIA immunofluorescent signals showed no significant difference to WT controls ( Fig 1E ) . Thus , the lack of henji is the cause for the augmented dPAK and GluRIIA levels at NMJs . The increased PSD size in henji mutants prompted us to examine the opposing AZ in presynapses . Bruchpilot ( Brp ) , an essential component of the T-bar structure within AZs [37 , 38] , was expressed in a normal pattern and intensity at the henji1/8 NMJ ( S3A Fig ) . Each presynaptic Brp punctum matched an enlarged dPAK patch in postsynapses , showing a characteristic pattern between pre- and postsynapses . Compared to control , Brp punctum from the henji mutant was unaltered in the intensity , density , and size ( S3A Fig , bar graphs ) . Also , the levels and patterns of the SSR protein Discs large ( Dlg ) , the cell adhesion molecule Fasciculin II ( FasII ) , and the microtubule-associated protein Futsch at henji NMJs were indistinguishable to WT controls ( S3B Fig ) . The specific alterations in dPAK and GluRIIA expressions at henji mutant NMJs suggest that Henji functions in postsynapses . To determine the functional site of Henji , we performed a rescue experiment with tissue-specific GAL4 drivers to induce UAS-henji expression . As shown , henji1/1 also displayed higher levels of dPAK and GluRIIA in postsynapses ( Fig 2A and 2B ) . When expressed in the henji1/1 postsynapse by muscular C57-GAL4 , the intensities of both dPAK and GluRIIA puncta at NMJs were suppressed to WT levels ( Fig 2A and 2B ) . In addition , supernumerary satellite boutons in henji mutants were also suppressed by postsynaptic expression of henji ( Fig 2B , bottom panel ) . In contrast , presynaptic expression of UAS-henji using neuronal elav-GAL4 failed to suppress any of these phenotypes ( Fig 2A and 2B ) . Thus , henji is required in postsynapses to regulate postsynaptic dPAK and GluRIIA abundance and presynaptic bouton growth . With the requirement for henji in postsynapses , we examined Henji localization at NMJs . We raised antibodies against Henji , which failed to reveal any specific signal in immunostaining . Also , the GFP-henji transgene that was tagged with GFP failed to show any detectable expression level . These results suggest that Henji might be expressed at very low levels . We took advantage of the GFP-henji transgene that also includes a UAS for GAL4-induced expression . When muscular C57-GAL4 was added , GFP-Henji showed localization near the synaptic region ( Fig 2C , third row ) . The postsynaptic-enriched pattern of Henji is specific , as the expression of cytosolic GFP showed diffuse staining in muscle cells without any particular pattern ( Fig 2C , second row ) . Presynaptically expressed GFP-Henji by neuronal GAL4 drivers also displayed diffuse signals in terminal boutons and axonal tracts ( Fig 2C , bottom two rows , respectively ) . To further examine the postsynaptic-enriched expression , co-immunostaining with Dlg was performed . GFP-Henji localized to the SSR and extended slightly outward as compared to Dlg immunostaining ( Fig 2D ) . The postsynaptic localization of Henji suggests a direct mechanism for Henji to regulate the abundance of dPAK and GluRIIA . Considering the evident GluR compositional shift towards elevated GluRIIA levels in the henji mutants , we addressed whether synaptic transmission is also affected by performing electrophysiological recordings . The amplitude of evoked junctional potential ( EJP ) did not show any defect ( Fig 3A ) . However , the postsynaptic response to spontaneous neurotransmitter release ( quantal size ) , as assessed by measuring the miniature EJP ( mEJP ) amplitude , was strongly elevated in the henji1/1 mutant ( Fig 3B ) , consistent with an increase in the GluRIIA/GluRIIB ratio . The frequency of mEJP remained normal ( Fig 3C ) . The quantal content , representing the number of effective synaptic vesicles released upon a nerve stimulus , was calculated as the ratio of EJP/mEJP amplitudes . We found that quantal content values decreased significantly in the henji1/1 mutants as compared to WT control ( Fig 3D ) . Similarly , we also detected similar elevation of mEJP and normal EJP in henji1/P , leading to a reduction of the quantal content , as compared to the henji1/+ heterozygous control ( Fig 3F–3H , compare first two bars ) . To further confirm the reduction of the quantal content , failure analysis was performed , which showed decreased release probability in the henji1/1 mutant ( Fig 3E ) . These data suggest that GluRIIA accumulates in the henji mutant , causing an elevation in postsynaptic responses . However , homeostatic mechanisms might tune down presynaptic release to reduce the quantal content , thereby maintaining a normal EJP output . We then tested whether the enhanced mEJP amplitude in the henji mutant is caused by the absence of Henji in postsynapses . Muscle expression of henji suppressed the mEJP amplitude , both in the henji1/P mutant and WT background , whereas neuronal expression did not ( Fig 3G ) . As Henji plays a role in suppressing the GluRIIA level in postsynapses ( Fig 2A and 2B ) , the elevation of the GluRIIA level is consistent with the enhancement of mEJP in the henji mutant . The increase of dPAK and GluRIIA patches may be associated with an expansion of the synaptic size in henji mutants . To examine this possibility , ultrastructures of boutons were analyzed by TEM . Cross-sections of boutons showed electron-dense membrane regions , representing the matching sites between presynaptic AZ and postsynaptic PSD ( Fig 4A , within two arrows ) . In presynapses , synaptic vesicle-docked T-bars located within AZs , while in postsynapses , membranous SSR enwraps the bouton . We found that the electron-dense membrane region was expanded in henji mutants ( enlarged panels in Fig 4A ) . Quantification indicated that the length of the electron-dense membrane region significantly increased in the henji1/1 and henjiP/P mutants ( Fig 4B ) . Moreover , as the bouton perimeter did not differ significantly between henji mutants and WT control ( S1 Table ) , the synaptic membrane accounted for a larger proportion of the total membrane region in the lack of henji ( Fig 4B , right panel ) . These analyses indicate that the synaptic membrane region is expanded in the henji mutants . Given the elevation of synaptic dPAK levels in the henji mutant , we tested whether reducing the dpak gene dosage could have an effect on henji mutant phenotypes . When the dpak6 null allele was introduced into the henji1/8 mutant background , GluRIIA abundance was suppressed ( Fig 5A ) . Similarly , both kinase-dead dpak3 and Dock-interaction-disrupted dpak4 alleles [39] also suppressed GluRIIA abundance in the henji1/8 mutant , suggesting that these functional domains are critical for dPAK to regulate GluRIIA abundance ( Fig 5A ) . Satellite boutons in the henji1/8 mutant were also suppressed by dpak6 and , to a lesser extent , dpak3 and dpak4 ( Fig 5B ) . In removing one copy of the dpak6 null allele in the henji1/8 mutant , dPAK was indeed reduced to near the WT level ( S4 Fig ) , consistent with the idea that the reduction of the dPAK level is able to suppress henji phenotypes . These genetic suppressions suggest that the upregulated dPAK level contributes to GluRIIA accumulation and abnormal bouton morphology in the henji mutant . It has been shown that dPAK regulates synaptic GluRIIA abundance; in the dpak mutant , the GluRIIA level was reduced at the PSD [19] . We then examined the epistatic relationship between henji and dpak mutants . In dapk3/6 larvae that survived to late larval stages , the GluRIIA level was greatly reduced ( Fig 5C ) , consistent with the previous report [19] . In contrast , the GluRIIA level was enhanced in the henji1/1 mutant ( Fig 2A ) . In the henji1/1 dapk3/6 double mutant , GluRIIA was reduced to the level similar to that of the dapk3/6 single mutant ( Fig 5C ) . Thus , in the absence of dpak , the GluRIIA level fails to be upregulated in the henji mutant , suggesting that dpak functions downstream of or in parallel to henji . With the upregulated GluRIIA level in the henji mutant , we also examined any suppression effect by reducing GluRIIA in a henji mutant background . Introducing one copy of a gluriia deletion allele in the henji mutant strongly suppressed the satellite bouton phenotype ( Fig 5D , arrowheads ) . However , dPAK abundance was not suppressed by reducing a gluriia gene dosage ( Fig 5E ) . Taken together , these data support a model whereby Henji restricts postsynaptic GluRIIA abundance by downregulating the dPAK levels . In henji mutants , accumulated GluRIIA induces abnormal bouton growth in the retrograde direction , resulting in satellite bouton morphology . To further understand the role of Henji on postsynaptic regulation , we generated N-terminal GFP-tagged deletion constructs ΔBTB , ΔBACK and ΔKelch that truncate one of the three conserved domains of Henji ( Fig 1A ) . These Henji truncations and full-length control were expressed in postsynapses for rescuing henji1/8 mutant phenotypes . As expected , full-length Henji when expressed in postsynapses suppressed the elevated dPAK and GluRIIA in henji1/8 while ΔKelch failed to do so , suggesting that the Kelch repeats region is essential for Henji function in postsynapses ( Fig 6A ) . Surprisingly , ΔBTB and ΔBACK significantly suppressed elevated intensities of both dPAK and GluRIIA in the henji1/8 mutant . These results suggest that both BTB and BACK domains are dispensable for Henji to function in postsynapses . To further address functional domains of Henji in postsynapses , full-length and truncations of Henji were overexpressed in postsynapses , and synaptic abundances of dPAK and GluRIIA were assessed . Significantly , full-length Henji when overexpressed in postsynapses caused reductions in dPAK and GluRIIA levels at NMJs , suggesting that Henji is sufficient to promote downregulation of dPAK and GluRIIA levels ( Fig 6B ) . Unlike the full-length Henji , truncating any of the three domains failed to downregulate dPAK and GluRIIA levels . Instead , ΔBTB produced a dominant-negative effect by inducing dPAK and GluRIIA accumulations while ΔBACK and ΔKelch had no effects ( Fig 6B ) . Therefore , the Kelch repeats seems to be the most critical domain of Henji to regulate the postsynaptic abundance of dPAK and GluRIIA . We then examined the domain requirement for Henji in postsynaptic localization . Full-length and truncated Henji transgenes were expressed in muscles of the henji1/8 mutant , and the protein localization was detected by GFP immunostaining . By co-staining with Dlg , full-length Henji localized to postsynaptic SSR ( Fig 7A ) . Similarly , absence of the BACK domain ( ΔBACK ) still retained proper synaptic localization of Henji . Lacking the BTB domain also retained some localization signals at the postsynapses . Finally , lacking the Kelch repeats ( ΔKelch ) completely abolished Henji postsynaptic localization ( Fig 7A ) . These analyses suggest that the Kelch repeats region are essential for proper postsynaptic localization of Henji . Considering the role of Kelch repeats in Henji postsynaptic function and localization , we tested whether Kelch repeats and dPAK interact physically . Indeed , both Flag-tagged full-length and Kelch repeats of Henji co-immunoprecipitated with Myc-tagged dPAK , providing evidence of physical interactions between Henji and dPAK ( Fig 7B , indicated by arrows ) . Taken together , these results indicate that the Kelch repeats that bind to dPAK are required for Henji localization and function to control dPAK and GluRIIA postsynaptic abundances . While dPAK is required for postsynaptic localization of GluRIIA , how dPAK functions to regulate GluRIIA abundance still remains unknown [19] . A constitutive-active form of dPAK that is membrane-tethered failed to increase GluRIIA abundance at PSDs , hinting another layer of regulation on dPAK . We tested whether Henji plays the critical role on limiting dPAK activity in GluRIIA regulation . To investigate this possibility , we generated Myc-tagged dPAK of WT , constitutive-active ( CA ) , or dominant-negative ( DN ) . The CA form contained a phospho-mimic T583E point mutation at the first autophosphorylation residue [40–42] . The DN form contained three point mutations , with H91L and H94L disrupting binding and activation by Cdc42 and Rac , and K459R eliminating kinase activity . Western blot analysis showed that , when driven by Tub-GAL4 , WT , CA , and DN were expressed at similar levels ( S5 Fig ) . All three transgenes were overexpressed postsynaptically and GluRIIA abundance was quantified . We found that , as previously reported [19] , overexpression of dPAK failed to alter GluRIIA abundance regardless of the activation status ( Fig 8A ) . To examine whether Henji confers the extra layer of regulation on dPAK activation , dPAK transgenes were overexpressed postsynaptically in the henji1/1 mutant background . Overexpression of dPAK WT did not alter the already enhanced GluRIIA level in the henji1/1 mutant , which may reflect a constraint in dPAK activation such as the requirement of CDC42/Rac1 in dPAK activation [40–42] . Interestingly , overexpression of the CA form further enhanced GluRIIA synaptic intensity in the henji1/1 mutants ( Fig 8B ) , a phenotype that was not detected when dPAK CA was overexpressed in the henji+/+ background ( Fig 8A ) . Also , dPAK DN suppressed GluRIIA intensity in the henji1/1 mutant , a phenotype that was not detected in the henji+/+ background , either ( Fig 8A and 8B ) . Taken together , we propose that Henji limits the action of dPAK to regulate GluRIIA abundance by adding another layer of regulation to conventional phosphorylation-mediated dPAK activation . Finally , we examined the localization of dPAK by Myc immunostaining . Overexpressed dPAK proteins of WT , CA and DN forms showed dispersed weak puncta in muscle cells without forming specific patterns ( Fig 8C ) . When overexpressed in the absence of henji activity , Myc-positive puncta became brighter and accumulated around the synaptic region . Some of the puncta were co-labeled with GluRIIB , representing specific accumulation at PSDs ( Fig 8D ) . Taken together , these analyses suggest that Henji functions to limit dPAK from localization at postsynaptic sites , which is important for modulating GluRIIA abundance .
PAK proteins transduce various signaling activities to impinge on cytoskeleton dynamics . Through kinase activity-dependent and -independent mechanisms , PAK regulates not only actin- and microtubule-based cytoskeletal rearrangement but also the activity of motors acting on these cytoskeletal tracks [43–46] . In mammalian systems , PAKs participate in many synaptic events including dendrite morphogenesis [47 , 48] , neurotransmitter receptor trafficking [49 , 50] , synaptic strength modulation [51] , and activity-dependent plasticity [52] . Pathologically , PAK dysregulation also contributes to serious neurodegenerative diseases [53] , such as Alzheimer disease [54 , 55] , Huntington’s disease [56 , 57] and X-linked mental retardation [58–60] . At Drosophila NMJs , dPAK has divergent functions; loss of dpak causes a dramatic reduction in both Dlg and GluRIIA synaptic abundance [19] , but the underlying molecular mechanisms have not been revealed . Our data show that Henji functions to restrict GluRIIA clustering but has no effect on Dlg levels ( S3B Fig ) , suggesting that Henji regulates one aspect of dPAK activities , probably via the SH2/SH3 adaptor protein Dock [19] . Alternatively , Henji may function to limit dPAK protein levels locally near the postsynaptic region , rendering its influence on GluRIIA clustering , while dPAK that regulates Dlg may localize outside of the Henji-enriched region . Supporting this idea , Henji is specifically enriched around the SSR region instead of dispersed throughout the muscle cytosol ( Fig 2D ) . Moreover , ectopic Myc-dPAK localized at the postsynapse only when henji was mutated ( Fig 8C and 8D ) , indicating that Henji regulates dPAK postsynaptic localization . The interaction with Rac , Cdc42 , or both triggers autophosphorylation and subsequent conformational changes of PAK , resulting in kinase activation . The myristoylated dPAK that has been shown to be active in growth cones [39] failed to enhance GluRIIA abundance at the NMJ [19] . This result shows that dPAK is necessary to regulate GluRIIA synaptic abundance , but is itself tightly regulated at the synaptic protein level or the kinase activity . Indeed , we provide evidence to show specific negative regulation of dPAK by Henji; overexpression of dPAK CA that could not enhance GluRIIA abundance in WT larvae further increased the already enhanced GluRIIA levels in the henji mutant ( Fig 8A and 8B ) . Similar to the CA form , the DN form also showed no effect on GluRIIA when simply overexpressed in the WT background , but exhibited strong suppression of GluRIIA in the henji mutant background ( Fig 8A and 8B ) . Thus , regardless of the possible conformational differences between the CA and DN forms , Henji appears to confer a constitutive negative regulation of dPAK at postsynapses , suggesting a tight control that could be at subcellular localization . In contrast to CA and DN forms , activation of dPAK requires binding to Rac1 and Cdc42 , and subsequent protein phosphorylation . This additional layer of regulation may serve as a limiting factor rendering dPAK WT from recruiting GluRIIA to PSDs regardless in WT or henji mutant background . The structural feature suggests that Henji could function as a conventional substrate receptor of the Cul3-based E3 ligase complex . At Drosophila wing discs , Dbo functions as a Cul3-based E3 ligase to promote Dishevelled ( Dsh ) downregulation [61] . Similar to the henji alleles , we confirmed that the dbo [Δ25 . 1] allele and dbo RNAi were competent to induce dPAK and GluRIIA accumulation at the postsynapse ( S6A Fig ) . In our immunoprecipitation experiment , we detected Henji and dPAK in the same complex ( Fig 7B , lane 3 ) , and dPAK also forms a complex with the C-terminal substrate-binding Kelch-repeats region ( lane 4 ) . However , we did not detect any notable or consistent increase in Henji-dependent dPAK poly-ubiquitination in both S2 cells and larval extracts . Also , the Cul3-binding BTB domain of Henji seems dispensable in the suppression of dPAK levels in henji mutants ( Fig 6A ) . Importantly , Cul3 knockdown in muscle cells failed to cause any accumulation of GluRIIA and dPAK at the NMJ ( S6A Fig ) . Sensitive genetic interaction between henji and Cul3 failed to induced dPAK and GluRIIA accumulation ( S6B Fig ) . Dbo functions together with another BTB-Kelch protein Kelch ( Kel ) to downregulate Dsh [61] . However , Kel negatively regulates GluRIIA levels without affecting dPAK localization at the postsynaptic site ( S6C Fig ) . This data argues that Kel functions in a distinct pathway to Henji in postsynaptic regulation of GluRIIA . Taken together , we found no direct evidence to support that dPAK is downregulated by Henji through ubiquitination-dependent degradation . Alternately , Henji could bind dPAK near the postsynaptic region and this interaction may block the recruitment or localization of dPAK onto postsynaptic sites ( Fig 9 ) . Under this model , dPAK is less restricted and has a higher propensity to localize at postsynaptic sites in the absence of Henji , resulting in synaptic accumulation of dPAK and GluRIIA expansions . As many synaptic events require rapid responses , local regulation of protein levels becomes crucial in synapses . To achieve accurate modulation , certain synaptic proteins should be selectively controlled under different developmental or environmental contexts . Indeed , emerging evidence shows that various aspects of synapse formation and function are under the control of the ubiquitin proteasome system ( UPS ) , including synapse formation [62 , 63] , morphogenesis [64] , synaptic pruning and elimination [65 , 66] , neurotransmission [67–69] , and activity-dependent plasticity [21 , 70] . In particular , the membrane abundance of postsynaptic GluR that modulates synaptic function can be regulated by components of the UPS . When Apc2 , the gene encoding Drosophila APC/C E3 ligase , is mutated , GluRIIA shows excess accumulation but the molecular mechanism was not elucidated [71] . Similarly , loss of the substrate adaptor BTB-Kelch protein KEL-8 in C . elegans also results in the stabilization of GLR-1-ubiquitin conjugates [72] . However , no evidence shows direct ubiquitination and degradation of GLR-1 by KEL-8 . Also , absence of the LIN-23-APC/C complex in C . elegans affects GLR-1 abundance at postsynaptic sites without altering the level of ubiquitinated GLR-1 . Therefore , GLR-1 receptor endocytosis and recycling or ubiquitination and degradation of GLR-1-associated scaffold proteins are proposed to be the underlying mechanism for E3 ligase regulation [23 , 67] . In mammals , endocytosis of AMPAR can be influenced by poly-ubiquitination and degradation of the prominent postsynaptic scaffold protein PSD-95 [21 , 73] . In this study , we describe a novel regulation by the BTB-Kelch protein Henji on synaptic GluRIIA levels . By limiting GluRIIA synaptic levels , Henji modulates the postsynaptic output in response to presynaptic glutamate release . In the absence of Henji , quantal size is elevated ( Fig 3B and 3G ) , coinciding with an increase in the postsynaptic GluRIIA/GluRIIB ratio ( Figs 1D and 1E and S2D ) . In a previous study , increases in the GluRIIA/GluRIIB ratio by overexpressing a GluRIIA transgene in the muscle or by reducing the gene copy of gluriib promote NMJ growth , but co-expression of both GluRIIA and GluRIIB did not alter the bouton number [74] . Combined with our findings , those data provide a link between an increased GluRIIA-mediated postsynaptic response and bouton addition at NMJs . However , satellite boutons were not detected following GluRIIA overexpression [74] . One possibility is that satellite boutons are considered as immature boutons [36 , 75] and their appearance may indicate the tendency for NMJ expansion , as in the case of excess BMP signaling [75] . Failure to become mature boutons may be caused by the lack of cooperation with other factors such as components of the presynaptic endocytic pathway [76] , actin cytoskeleton rearrangement [77–79] or neuronal activity [36] . We found no significant alterations in endocytosis and the BMP pathway in the henji mutant ( S7 Fig ) . Nevertheless , we cannot rule out that Henji may modulate other presynaptic events that are defective in henji mutants to interfere with bouton maturation ( S1 Table ) .
w1118 was used as the WT control and to backcross all henji alleles described in this study . Flies of all genotypes were reared at 25°C for experiments . We performed P-element-mediated imprecise excision to generate henji1 and henji8 alleles ( S2A Fig ) . Plasmids of UAS-Flag-henji ( full-length ) , UAS-Flag-Kelch ( 533–623 a . a . ) , UAS-GFP-henji ( full-length ) , UAS-GFP-ΔBTB ( delete 1–167 a . a . ) , UAS-GFP-ΔBACK ( delete 168–276 a . a . ) , UAS-GFP-ΔKelch ( delete 305–623 a . a . ) , UAS-Myc-dpak WT ( full-length ) , UAS-Myc-dpak CA ( point mutation T583E ) and UAS-Myc-dpak DN ( triple mutations H91L , H94L and K459R ) were constructed using the Gateway System into the pUAST vector ( Invitrogen and Drosophila Genomics Resource Center , DGRC ) . The genomic rescue transgene GFP-henji was constructed by fusing GFP to the ATG codon of henji cDNA and driven by the putative promoter region containing the genomic sequence between CG6169 ATG and henji ATG . C57-GAL4 , elav-GAL4 , kelDE1 , and dpak6 were from Bloomington Drosophila Stock Center ( BDSC ) . RNAi lines for Cul3 ( 109415 ) and gbb ( 5562R ) [80] were from Vienna Drosophila RNAi Center ( VDRC ) and National Institute of Genetics ( NIG ) , respectively . Mutant strains that have been described are dpak3 and dpak4 [39] , dbo[Δ25 . 1] and dbo RNAi [61] , and GluRIIA-GFP rescuing gluriia and gluriib double mutants [81] . Primary antibodies used were: mouse anti-Dlg ( 4F3 , 1:100 , Developmental Studies Hybridoma Bank , DSHB ) , mouse anti-GluRIIA ( 1:100 , DSHB ) , rabbit anti-dPAK ( 1:1000 ) [82] , rabbit anti-GluRIIB ( 1:1000 ) [14] , mouse anti-Brp ( 1:100 , DHSB ) , mouse anti-FasII ( 1:100 , DSHB ) , mouse anti-Futsch ( 1:100 , DSHB ) , chicken anti-GFP ( 1:100; Abcam Co . ) , mouse anti-Myc ( 9E10 , 1:100 , Santa Cruz Co . ) , rabbit anti-pMAD ( 1:250 ) [83] and rabbit or goat anti-HRP conjugated FITC , TRITC and Cy5 ( Jackson ImmunoResearch Laboratories ) . Secondary antibodies used were anti-rabbit or -mouse Cy3 and Cy5 ( Jackson ImmunoResearch Laboratories ) . Muscles , though not shown in figures , were revealed by staining with FITC-conjugated phalloidin ( 1:1000; Sigma Co . ) . A2 to A6 segments of NMJ4s and A3 segments of NMJ6/7s of wandering third instar larvae were analyzed . Larvae were dissected in cold calcium-free HL3 saline ( 70 mM NaCl , 5 mM KCl , 20 mM MgCl2 , 10 mM NaHCO3 , 5 mM trehalose , 115 mM sucrose , and 5 mM HEPES , pH 7 . 2 ) and larval fillets were fixed in 4% paraformaldehyde for 20 min and washed in PBT ( 0 . 03% triton-X-100 ) for 10 min three times . For GluRIIA and GluRIIB staining , larval fillets were fixed in Bouin’s fixative ( Sigma Co . ) for 5 min . Fixed fillets were incubated with primary antibodies overnight at 4°C , washed in PBT three times , and incubated with secondary antibodies for 2 hr at room temperature . Larval fillets were mounted in solution containing PBS with 87 . 5% glycerol and 0 . 22 M 1 , 4-diaza-byciclo ( 2 . 2 . 2 ) octane ( Dabco , Sigma Co . ) . Images were acquired via LSM 510 confocal microscopy ( Carl Zeiss ) using 40x water and 100x oil objectives . Images were processed by LSM5 image examiner ( Carl Zeiss ) and Adobe Photoshop Creative Suite , and further quantified by Image J for immunofluorescence intensities , punctum densities and cluster sizes . For quantification , Z-section images were projected for further processing . Immuno-positive regions were defined by using Image J in which threshold setting was used to eliminate background noise . Intensity of each synaptic protein was normalized to corresponding HRP intensity . Satellite bouton numbers were normalized to corresponding muscle areas . For image presentation , immunostaining images presented in figures represent single sections except the Futsch images in S3B Fig were from projection of Z sections . Unpaired Student t-test is used in calculating statistical significance . Drosophila S2 cells were maintained in Schneider's medium ( Thermo Fisher Scientific ) at 25°C . S2 cells ( 5 x 106 cells in each 10 cm dish ) were transfected with 1 μg DNA of individual constructs using Cellfectin ( Invitrogen ) . S2 cells were collected and homogenized in RIPA lysis buffer ( 20 mM Tris-HCl , pH8 . 0 , 150 mM NaCl , 5 mM EDTA , 1% Triton-X-100 , 2 mM Na3VO4 , 50 mM NaF and 1 mM PMSF , supplemented with protease inhibitor cocktail ( Roche ) . Protein concentrations were calculated with the aid of protein assay ( Bio-Rad Laboratories ) . For immunoprecipitation , Myc-tagged dPAK was co-transfected with either Flag-tagged Henji or the Kelch repeats domain of Henji into S2 cells . Cell lysates were incubated with beads coated with anti-Myc ( 9E10 , Santa Cruz ) and the immunoprecipitates were blotted with mouse anti-Flag antibody ( 1:1000 , Sigma Co . ) . Antibodies used for immunoblotting were anti-dPAK ( 1:5000 ) , anti-Myc ( 1:1000 ) , and anti-α-tubulin ( 1:200000 , Sigma Co . ) . Larval fillets were dissected in cold calcium-free HL3 saline and subsequently fixed overnight in modified Trump’s universal fixative ( 4% paraformaldehyde , 1% glutaraldehyde in 0 . 2 M cacodylate buffer , pH 7 . 2 ) . The PELCO BioWave® laboratory microwave system was used for subsequent steps . Samples were post-fixed with 1% aqueous osmium tetroxide in 0 . 2 M cacodylate buffer ( pH 7 . 2 ) under 20 inHg vacuum . After stained with 2% uranyl acetate for 30 min at room temperature , samples were dehydrated in gradually-increasing ethanol concentrations ( 50% , 70% , 80% and 90% ) . Later , fillets were infiltrated in Spurr’s resin with gradual increases of concentrations ( 25% , 50% , 75% and 100% ) . Ultrathin sections , obtained by ultramicrotome ( Leica ) , were further stained with uranyl acetate and lead citrate . Images were viewed by Tecnai G2 Spirit TWIN ( FEI Company , Hillsboro , OR ) . Electron-dense region was determined by Gatan DigitalMicrograph . A line was drawn along the bouton membrane and spots with lowest intensity were picked and labeled as the boundary of electron-dense region . Bouton parameters were quantified by Image J . For sample preparation , larvae were dissected with the segmental nerves cut close to the ventral ganglion region in cold modified calcium-free HL3 . 1 saline ( 70 mM NaCl , 5 mM KCl , 10 mM MgCl2 , 10 mM NaHCO3 , 5 mM trehalose , 115 mM sucrose , 5 mM HEPES , pH 7 . 2 ) . Samples were then incubated in modified HL3 . 1 saline containing 0 . 8 mM CaCl2 for stimulation , and recordings were taken at room temperature . The two electrodes for voltage-clamping were filled with 3 M KCl and impaled in muscle 6 of the A3 segment . One microelectrode ( 15~20 MΩ ) monitored the muscle membrane potential while the other ( 5~8 MΩ ) delivered electric currents . 5-8V stimulation was given to stimulate the nerve . The muscle membrane potential was clamped at -60 mV . Without any stimulation on the segmental nerves , mEJPs within 100 sec were recorded . For evoking an EJP , the segmental nerve was stimulated by a suction electrode every 30 sec with pulse duration of 0 . 1 msec at the voltage two times that of the threshold . For failure analysis EJP is evoked in 0 . 2 mM [Ca2+] , the failure rate was calculated by ln ( n/N ) , with n the number of failure events , and N the total number of stimuli [16] . For high-frequency stimulation , the segmental nerve was stimulated at 13 . 3 Hz for 4 min in 2 mM [Ca2+] buffer . Data were digitized by a DigiData 1440 interface ( Molecular Devices ) at 50 kHz , and weak signals were filtered at 10 kHz , and analyzed by Clampfit10 ( Molecular Devices ) .
|
To meet various developmental or environmental needs , the communication between pre- and postsynapse can be modulated in different aspects . The release of presynaptic vesicles can be regulated at the steps of docking , membrane fusion and endocytosis . Upon receiving neurotransmitter stimuli from presynaptic terminals , postsynaptic cells tune their responses by controlling the abundance of different neurotransmitter receptors at the synaptic membrane . The Drosophila NMJ is a well-defined genetic system to study the function and physiology of synapses . Two types of glutamate receptors ( GluRs ) , IIA and IIB , present at the NMJ , exhibit distinct desensitization kinetics: GluRIIA desensitizes much slower than GluRIIB does , resulting in more ionic influx and larger postsynaptic responses . By altering the ratio of GluRIIA to GluRIIB , muscle cells modulate their responses to presynaptic release efficiently . However , how to regulate this intricate GluRIIA/GluRIIB ratio requires further study . Here , we describe a negative regulation for dPAK , a crucial regulator of GluRIIA localization at the PSD . Henji specifically binds to dPAK near the postsynaptic region and hinders dPAK localization from the PSD . By negatively controlling dPAK levels , synaptic GluRIIA abundance can be restrained within an appropriate range , protecting the synapse from unwanted fluctuations in synaptic strengths or the detriment of excitotoxicity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"nervous",
"system",
"cell",
"processes",
"electrophysiology",
"neuroscience",
"animals",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"molecular",
"biology",
"techniques",
"epigenetics",
"drosophila",
"research",
"and",
"analysis",
"methods",
"staining",
"specimen",
"preparation",
"and",
"treatment",
"genetic",
"interference",
"proteins",
"hyperexpression",
"techniques",
"gene",
"expression",
"ubiquitination",
"molecular",
"biology",
"insects",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"arthropoda",
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"rna",
"endocytosis",
"immunostaining",
"anatomy",
"synapses",
"nucleic",
"acids",
"phenotypes",
"cell",
"biology",
"physiology",
"secretory",
"pathway",
"genetics",
"post-translational",
"modification",
"biology",
"and",
"life",
"sciences",
"neurophysiology",
"organisms"
] |
2016
|
Dbo/Henji Modulates Synaptic dPAK to Gate Glutamate Receptor Abundance and Postsynaptic Response
|
Plants have a large panel of nucleotide-binding/leucine rich repeat ( NLR ) immune receptors which monitor host interference by diverse pathogen molecules ( effectors ) and trigger disease resistance pathways . NLR receptor systems are necessarily under tight control to mitigate the trade-off between induced defenses and growth . Hence , mis-regulated NLRs often cause autoimmunity associated with stunting and , in severe cases , necrosis . Nucleocytoplasmic ENHANCED DISEASE SUSCEPTIBILITY1 ( EDS1 ) is indispensable for effector-triggered and autoimmune responses governed by a family of Toll-Interleukin1-Receptor-related NLR receptors ( TNLs ) . EDS1 operates coincidently or immediately downstream of TNL activation to transcriptionally reprogram cells for defense . We show here that low levels of nuclear-enforced EDS1 are sufficient for pathogen resistance in Arabidopsis thaliana , without causing negative effects . Plants expressing higher nuclear EDS1 amounts have the genetic , phenotypic and transcriptional hallmarks of TNL autoimmunity . In a screen for genetic suppressors of nuclear EDS1 autoimmunity , we map multiple , independent mutations to one gene , DM2h , lying within the polymorphic DANGEROUS MIX2 cluster of TNL RPP1-like genes from A . thaliana accession Landsberg erecta ( Ler ) . The DM2 locus is a known hotspot for deleterious epistatic interactions leading to immune-related incompatibilities between A . thaliana natural accessions . We find that DM2hLer underlies two further genetic incompatibilities involving the RPP1-likeLer locus and EDS1 . We conclude that the DM2hLer TNL protein and nuclear EDS1 cooperate , directly or indirectly , to drive cells into an immune response at the expense of growth . A further conclusion is that regulating the available EDS1 nuclear pool is fundamental for maintaining homeostatic control of TNL immune pathways .
In plants , receptors that sense pathogen attack are central players in the biotic stress signaling network . Receptor activation triggers innate immunity pathways to protect cells and tissues from disease . In a first line of defense , surface pattern recognition receptors ( PRRs ) bind microbial molecules to activate disease resistance programs leading to pattern-triggered immunity ( PTI ) . A second critical immunity layer is mediated by intracellular nucleotide-binding/leucine-rich-repeat ( NLR ) receptors that recognize virulence factors ( called effectors ) which are delivered by pathogen strains to dampen PTI and promote disease [1] . Structural counterparts of plant NLRs called NOD-LRR ( nucleotide-binding/oligomerization-domain/leucine-rich-repeat ) receptors also sense pathogen interference in mammalian systems [2 , 3] . NLR and NOD-LRR proteins are ATP-driven molecular switches which become stimulated by direct binding of an effector molecule or effector modifications of an NLR-monitored host target [4 , 5] . In plants , NLR activation induces a robust resistance response called effector-triggered immunity ( ETI ) involving the amplification of PTI-related transcriptional programs and , often , host cell death at infection sites ( a hypersensitive response , HR ) [6] . NLRs are among the most rapidly evolving plant genes [7–9] , and expansion in NLR gene number and diversity , as paralogs within complex loci or allelic variants in different genotypes , is in part driven by pathogen effector pressure [10–13] . Receptor monitoring ( or guarding ) of important defense hubs that are targeted by multiple pathogen effectors probably further increases NLR recognition space [14–17] . Nevertheless , the rapid evolution of NLR genes creates potentially dangerous molecules if activated in the absence of a pathogen effector stimulus [4 , 18] . Loss of NLR homeostasis caused by mutation , mis-expression or disturbance of NLR-monitored co-factors leads to autoimmunity . Plant autoimmune backgrounds display constitutive defense gene expression and varying degrees of stunting , necrosis and reduced reproductive fitness [19] . As in ETI , NLR autoimmune phenotypes are often conditional on temperature with high temperatures ( 25–28°C ) suppressing disease resistance , transcriptional activation of defense pathways and HR-related cell death [19–21] . Temperature-conditioned autoimmunity can also arise in the progeny of inter- or intra-specific crosses between different genetic backgrounds to produce immune-related hybrid incompatibility ( HI ) ( known also as hybrid necrosis ) [19 , 22] . HI is caused by deleterious epistatic interactions between two or more loci that have diverged through genetic drift or selection in the different parental lineages [23–25] . Mapping of the causal interacting genes or allelic forms in several cases of temperature-conditioned HI shows that many are in NLR or immune-related loci [18 , 22 , 25–29] . Therefore , HI might expose altered NLR regulation and/or associations with monitored co-factors as immunity systems evolve . Effector-activated NLR receptors connect to a conserved basal resistance network to mobilize ETI defense pathways [6] . Although the downstream events are not well understood , signals in ETI ultimately converge on the nuclear transcription machinery to boost PTI-related defense programs [6] . A major NLR subclass in dicotyledenous species has an N-terminal Toll-Interleukin1-receptor ( TIR ) domain ( referred to as TNLs or TIR-NB-LRRs ) [9 , 30] and requires the nucleocytoplasmic , lipase-like protein ENHANCED DISEASE SUSCEPTIBILITY1 ( EDS1 ) for all measured ETI and autoimmunity outputs [21 , 31–34] . Interactions between EDS1 and TNL proteins suggested that EDS1 provides an immediate link between TNLs and downstream resistance pathways [35–37] . Importantly , EDS1 nuclear accumulation was found to be necessary for A . thaliana basal immunity against virulent pathogen strains and TNL-triggered ETI , consistent with a central EDS1 role in transcriptional reprogramming of cells for defense [21 , 32 , 38] . Analysis of A . thaliana transgenic plants in which EDS1 was mis-localized to the cytoplasm or its nucleocytoplasmic trafficking disturbed , suggested also that the EDS1 cytoplasmic pool contributes to resistance [38 , 39] . Unlike many mis-regulated NLRs , over-accumulation of functional , nucleocytoplasmic A . thaliana EDS1 does not cause autoimmunity [38 , 40] . Here , we investigated the consequences of restricting A . thaliana EDS1 to the nuclear compartment . Our analysis shows that a low-level EDS1 nuclear pool , operating with signaling partners , is sufficient for mediating A . thaliana basal and TNL immunity without deleterious consequences for the plant . However , raising nuclear EDS1 amounts above a certain threshold leads to autoimmunity with many features of a deregulated TNL immune response . In a screen for genetic suppressors , we discover that the nuclear EDS1 autoimmune phenotype depends on presence of the ‘DANGEROUS MIX2’ ( DM2 ) RPP1-like TNL gene cluster . The DM2 locus is a hotspot for genes underlying immune-related HI . In our case , a cluster of eight RPP1-like TNL genes linked to an eds1 deletion mutation had been co-introgressed from A . thaliana accession Landsberg erecta into accession Columbia ( Col ) . We identify one gene , DM2h , within the DM2 RPP1-likeLer locus as necessary for nuclear EDS1 autoimmunity . We propose that a weak DM2hLer autoactivity which is normally constrained is exposed by nuclear EDS1 , producing EDS1-dependent defense expression and autoimmunity . A corollary of this damaging co-action between a TNL and nuclear EDS1 is that in wild-type plants , regulating the nuclear EDS1 pool likely helps to maintain TNL immune pathway homeostasis and growth .
We tested whether increased targeting of EDS1 to nuclei affects its disease resistance activity . For this , A . thaliana stable transgenic lines expressing genomic EDS1 under control of its native promoter and fused to a C-terminal yellow fluorescent protein ( YFP ) tag and SV40 nuclear localization signal ( NLS ) were generated in an eds1-2 deletion mutant in accession Col-0 ( Col ) ( Fig 1A ) . The eds1-2 mutation had been introgressed originally from accession Landsberg erecta ( Ler ) over eight backcrosses because Col contains a tandem duplication of two functional EDS1 genes [41] . Three independent EDS1-YFPNLS lines ( #A3 , #A5 and #B2 ) were taken to homozygosity and tested alongside a previously characterized Col eds1-2 transgenic line expressing functional , genomic EDS1-YFP [38] . EDS1-YFP protein accumulation in leaf extracts of the different transgenic lines was compared to that of native EDS1 in Col by immunoblotting with anti-EDS1 antibodies . The EDS1-YFPNLS protein levels ranged from lower than wild-type EDS1 ( in EDS1-YFPNLS line #B2 ) to higher than wild-type EDS1 ( EDS1-YFPNLS line #A5 ) , with highest accumulation in EDS1-YFPNLS line #A3 ( Fig 1B ) . Accumulation of EDS1-YFP ( without an additional NLS ) was intermediate between that of EDS1-YFPNLS lines #A5 and #A3 ( Fig 1B ) . Confocal laser scanning microscopy of leaf epidermal cells showed that EDS1-YFP distributed in the cytoplasm and nucleus , as expected [38] , whereas EDS1-YFPNLS was detected only in nuclei in lines #B2 , #A5 and #A3 ( Fig 1C ) . Biochemical purification of nuclei from leaf tissues showed that there was strong nuclear enrichment of EDS1 protein in the EDS1-YFPNLS line #A5 compared to EDS1-YFP ( Fig 1D ) . Growth of EDS1-YFP and EDS1-YFPNLS #B2 , #A5 and #A3 plants in soil under short-day conditions ( 10 h light period at 22°C ) was monitored over several weeks . EDS1-YFP and the EDS1-YFPNLS low expressor line #B2 were undistinguishable from wild-type Col or Col eds1-2 ( Figs 1E and S1A ) . By contrast , EDS1-YFPNLS #A3 seedlings became stunted and chlorotic after the first true leaves emerged at ~ 2 weeks and were dead at 4 weeks ( Figs 1E and S1A ) . EDS1-YFPNLS #A5 plants displayed stunting , curling of leaves and chlorosis from 4–5 weeks but remained viable and partially fertile ( Figs 1E and S1A ) . The developmental defects of EDS1-YFPNLS lines #A3 and #A5 co-segregated with the T-DNA selection marker . Also , the T-DNA insertion in EDS1-YFPNLS #A3 mapped to the first exon of At4g28490 , in which an insertion mutation ( in the haesa ( hae ) single mutant ) does not have a visible phenotype [42] . These results suggest that increased EDS1 nuclear localization or an imbalance in EDS1 nucleocytoplasmic partitioning , rather than EDS1 over expression , leads to EDS1 dose-dependent growth defects . We also generated Col eds1-2 transgenic lines expressing EDS1-YFP fused to a mutated , inactive NLS ( Figs 1A and S1 ) [43] . No line was found that expressed EDS1-YFPnls protein as highly as EDS1-YFPNLS in line #A5 . Two EDS1-YFPnls lines ( nls*#α5 and nls*#β5 ) were selected that had moderately high EDS1-YFP expression ( S1B Fig ) . These showed a nucleocytoplasmic distribution of EDS1-YFP ( S1C Fig ) and grew normally at 22°C ( S1D Fig ) . Because the developmental phenotypes in EDS1-YFPNLS lines #A3 and #A5 resemble A . thaliana autoimmunity backgrounds we measured expression of the EDS1-dependent defense marker genes PATHOGENESIS RELATED1 ( PR1 ) and AvrPphB SUSCEPTIBLE3 ( PBS3 ) in EDS1-YFPNLS transgenic and control lines . PR1 and PBS3 expression remained low in Col , Col eds1-2 and the phenotypically normal EDS1-YFP or EDS1-YFPNLS #B2 lines over a 3–6 week growth period ( Fig 2A ) . From ~ 4 weeks on , PR1 and PBS3 expression increased in EDS1-YFPNLS line #A5 ( Figs 2A and S1E ) , consistent with the appearance of macroscopic growth defects . High PR1 and PBS3 expression was also detected in 3-week-old dying EDS1-YFPNLS #A3 plants ( Fig 2A ) . By contrast , the EDS1-YFP-nls lines *#α5 and *#β5 did not have elevated PR1 expression ( S1F Fig ) . A gradual increase in EDS1 total protein accumulation over 3–6 weeks development was detected in both the EDS1-YFP and EDS1-YFPNLS lines ( Fig 2B ) , suggesting that there is a general rise in EDS1 steady state levels as plants age , regardless of EDS1 nucleocytoplasmic or nuclear distribution . Total and free SA levels were unchanged in 5-week-old EDS1-YFPNLS #B2 , Col and Col eds1-2 plants , but were high in line EDS1-YFPNLS #A5 ( Fig 2C ) . Hence , during development , accumulation of nuclear EDS1 in EDS1-YFPNLS lines #A3 and #A5 appears to reach a threshold for causing defense gene activation and disturbed growth . These results show that nuclear EDS1-YFPNLS in line #A5 , and more acutely in #A3 , has the capacity to transcriptionally activate defense pathways in the absence of a pathogen stimulus . We tested whether EDS1 targeted to nuclei is sufficient to confer basal disease resistance by spray-infecting leaves with the virulent bacterial pathogen Pseudomonas syringae pv . tomato strain DC3000 ( Pst DC3000 ) . As expected , Pst DC3000 growth was higher at 3 d post-infection ( 3 dpi ) in Col eds1-2 than in wild-type Col leaves , indicative of a loss of basal resistance in Col eds1-2 ( Fig 2D ) . The eds1-2 defect was fully complemented in EDS1-YFPNLS #B2 expressing low levels of EDS1-YFPNLS ( Figs 1B and 2D ) . Pst DC3000 growth was marginally reduced on EDS1-YFPNLS #A5 compared to wild-type Col plants ( Fig 2D ) . Similar resistance trends were observed in these transgenic lines in response to infection by a virulent oomycete pathogen , Hyaloperonospora arabidopsidis ( Hpa , isolate Noco2 ) ( Fig 2E ) . We then tested whether nuclear-enriched EDS1 functions in ETI by inoculating plants with Pst DC3000 delivering the Type-III secreted effector AvrRps4 ( Pst AvrRps4 ) , or with an incompatible Hpa isolate , Emwa1 . In accession Col , AvrRps4 is recognized by the nuclear TNL receptor pair RRS1/RPS4 [32 , 44–46] and Hpa Emwa1 by the TNL receptor RPP4 [47] , in EDS1-dependent ETI . Accordingly , Pst AvrRps4 growth at 3 dpi was restricted in wild-type Col in an EDS1-dependent manner ( Fig 2F ) . RRS1/RPS4 ETI against Pst AvrRps4 was also fully restored in EDS1-YFPNLS lines #A5 and #B2 ( Fig 2F ) , as well as in EDS1-YFPnls lines *#α5 and *#β5 ( S1G Fig ) . EDS1-YFPNLS #A5 and #B2 restricted Hpa Emwa1 growth as efficiently as wild-type Col , with all lines exhibiting a host hypersensitive response ( HR ) at attempted Hpa infection sites , as measured by Trypan Blue ( TB ) -staining of infected leaves ( Fig 2G ) . As expected , Col eds1-2 plants were fully susceptible to Hpa Emwa1 infection ( Fig 2G ) . No HR lesioning was observed in mock-inoculated EDS1-YFPNLS lines #A5 or #B2 , indicating that the host HR is pathogen-triggered ( Fig 2G ) . We concluded that even low levels of nuclear-targeted EDS1 , as in EDS1-YFPNLS #B2 , are sufficient for Arabidopsis basal and TNL-conditioned immunity . Because many Arabidopsis effector-triggered TNL and autoimmunity phenotypes are attenuated at elevated temperatures , we tested whether high temperature alters EDS1-YFP nuclear accumulation . At 28°C , accumulation of the nucleocytoplasmic TNL proteins tobacco N , Arabidopsis RPS4 and SNC1 ( SUPPRESSOR OF npr1-1 CONSTITUTIVE1 ) inside nuclei and EDS1-dependent transcriptional reprogramming are reduced [21 , 48 , 49] . Macroscopic growth defects and enhanced PR1 expression in EDS1-YFPNLS lines #A3 and #A5 at 22°C were also suppressed when plants were propagated at 28°C ( S1A and S1E Fig ) . Confocal laser scanning microscopy of leaves taken directly from plants grown at 22°C or 28°C showed that the distribution of nucleocytoplasmic EDS1-YFP or nuclear EDS1-YFPNLS fluorescence signals did not change substantially between the two temperature regimes ( Fig 3A ) . Therefore , high temperature suppression of EDS1-YFPNLS autoimmunity in line #A5 is not due to a failure in EDS1 nuclear import . However , steady state levels of EDS1-YFPNLS were lower in plants grown at 28°C compared to 22°C , as monitored on immunoblots with anti-EDS1 antibodies ( Fig 3B ) . A decrease in native EDS1 protein accumulation was also detected in wild-type Col grown at 28°C ( Fig 3B ) . Therefore , growth at 28°C leads to reduced EDS1 protein accumulation regardless of whether EDS1 is confined to the nucleus or free to shuttle between the nucleus and cytoplasm [38] . This is in line with a reported lowering of EDS1 transcript levels under high temperature conditions [50] . We concluded that suppression of autoimmunity in EDS1-YFPNLS #A5 , and probably also #A3 at 28°C ( S1A Fig ) , is caused by reduction of nuclear EDS1 to below a threshold needed to elicit autoimmunity . A . thaliana EDS1 forms resistance signaling complexes with either one of two sequence-related partners , PHYTOALEXIN DEFICIENT4 ( PAD4 ) and SENESCENCE ASSOCIATED GENE101 ( SAG101 ) [31 , 40 , 51 , 52] . Whereas PAD4 compensates genetically for a loss-of-function sag101 mutation , SAG101 only partially compensates for loss of PAD4 in basal resistance against virulent pathogens and in TNL mediated ETI [40 , 51 , 53] . The enhanced disease susceptibility phenotype of a pad4 sag101 double mutant is as penetrant as an eds1 loss-of-function mutation and is not alleviated by over-expressing functional EDS1-HA [40 , 51] . Thus , EDS1 requires PAD4 and , in the absence of PAD4 , SAG101 for disease resistance signaling in basal immunity and ETI . We tested the genetic dependence of EDS1-YFPNLS #A5 autoimmunity on PAD4 and SAG101 by crossing EDS1-YFPNLS #A5 with Col pad4-1 and sag101-1 single null mutants or a Col pad4-1 sag101-1 double mutant and selecting lines that were homozygous for the EDS1-YFPNLS transgene and eds1-2 in the respective homozygous mutant backgrounds . Developmental ( Fig 4A ) and PR1 expression ( Fig 4B ) autoimmune phenotypes of EDS1-YFPNLS #A5 were fully rescued by pad4-1 and pad4-1 sag101-1 but not by sag101-1 . This indicates that autoimmunity caused by nuclear-enriched EDS1 has the same genetic requirements for PAD4 and SAG101 as EDS1-mediated basal immunity and ETI in wild-type plants . EDS1-YFPNLS protein abundance was substantially lower in pad4-1 and pad4-1 sag101-1 mutant backgrounds , and similar to levels of native EDS1 in Col wild-type ( Fig 4C ) . Reduced EDS1 accumulation can be largely attributed to reduced EDS1 expression as measured by qRT-PCR in the same plants ( Fig 4D ) . EDS1 is also directly stabilized by PAD4 or SAG101 [51 , 52] . A . thaliana RRS1/RPS4 TNL resistance in EDS1-YFPNLS #A5 against Pst AvrRps4 displayed the same genetic dependence on PAD4 and SAG101 as wild-type EDS1 in Col ( Fig 4E ) . We conclude that the defense-promoting actions of PAD4 or SAG101 predominantly target the EDS1 nuclear pool in plant immunity . We next tested whether EDS1-YFPNLS #A5 autoimmunity requires signaling by the defense hormone salicylic acid ( SA ) because EDS1-PAD4 promote SA-dependent and SA-independent defense pathways [41 , 54–56] . Also , SA feeds-forward to induce PAD4 expression [53] . For this , loss-of-function mutations in the SA biosynthetic enzyme gene ISOCHORISMATE SYNTHESIS1 ( ICS1 , Col sid2-1 ) or the SA-response regulator gene NON-EXPRESSOR OF PR GENES1 ( NPR1 , Col npr1-1 ) were introduced into the EDS1-YFPNLS #A5 background . High accumulation of SA in 5-week-old EDS1-YFPNLS #A5 was abolished in EDS1-YFPNLS #A5/sid2-1 plants ( S2A Fig ) , confirming that SA in this line is produced mainly by ICS1 [57] . SA levels were not lower in EDS1-YFPNLS #A5 /npr1-1 , consistent with NPR1 operating downstream of SA accumulation [58] . Both sid2-1 and npr1-1 abolished enhanced expression of the SA-dependent PR1 marker gene in EDS1-YFPNLS #A5 ( S2B Fig ) , but only slightly compromised accumulation of EDS-YFPNLS protein ( S2C Fig ) . Strikingly , neither sid2-1 nor npr1-1 suppressed EDS1-YFPNLS #A5 stunting ( S2D Fig ) . We concluded that EDS1-YFPNLS #A5 immune-related growth defects are SA-independent or have a lower SA threshold . Altogether , the genetic epistasis data suggest that EDS1-YFPNLS autoimmunity operates by similar mechanisms as pathogen-elicited basal resistance or ETI , with EDS1-PAD4 controlled pathways branching into SA-dependent and SA-independent signaling sectors . Previously , we found that shifting plants from high ( 28°C , permissive ) to moderate ( 19°C , restrictive ) temperature can be used to trigger EDS1-dependent autoimmunity in a transgenic A . thaliana RPS4 over-expression line ( 35S:RPS4-HS ) [21] . Analysis of global gene expression changes in 35S:RPS4-HS and 35S:RPS4-HS eds1-2 leaf tissues over a 24 h time course showed that temperature-conditioned RPS4 autoimmunity at 8 h and 24 h post temperature shift ( pts ) largely mirrors EDS1-dependent transcriptional reprogramming in RRS1/RPS4 ( TNL ) ETI against Pst AvrRps4 [21] . Moreover , a set of EDS1-dependent induced or repressed marker genes from Pst AvrRps4-triggered tissues at 6 h post infection ( hpi ) displayed the same EDS1-dependent trends in 35S:RPS4-HS leaves at 8 h pts [21] . We performed Affymetrix ATH1 GeneChip analysis of 4-week-old untreated EDS1-YFPNLS line #A5 and wild-type Col plants grown at 22°C to measure the extent to which EDS1-YFPNLS #A5 autoimmunity resembles pathogen-elicited or temperature-induced A . thaliana immune responses . More than 2000 genes were significantly up- or down-regulated ( p-value < 0 . 01 , > 2-fold change ) in EDS1-YFPNLS line #A5 compared to Col at 22°C . Genes exhibiting at least 4-fold transcriptional differences in EDS1-YFPNLS #A5 compared to Col were then used for hierarchical clustering with transcriptome data sets from bacterial NLR-conditioned PTI or ETI , as well as 35S:RPS4-HS temperature-conditioned autoimmunity ( Fig 5 and S1 Table ) . This analysis established that the EDS1-YFPNLS #A5 transcriptome was most similar to 35S:RPS4-HS gene expression changes at 8 h and 24 h pts and to those of ETI interactions ( Pst AvrRps4 , Pst AvrRpm1 6 h; Fig 5 ) . The EDS1-YFPNLS #5 transcriptome was most different to those of Pst AvrRps4-elicited or temperature-shift induced eds1 mutant backgrounds ( Fig 5 ) . Notably , EDS1-dependent induced and repressed genes in the EDS1-YFPNLS #A5 transcriptome overlapped substantially with EDS1-dependent induced and repressed gene sets in RRS1/RPS4-mediated ETI or 35S:RPS4-HS autoimmunity ( Fig 5 ) . Two clusters of induced and repressed genes were unique to EDS1-YFPNLS #A5 ( a and b in Fig 5 , S2 Table ) and might correspond to adaptation to prolonged defense activation in the EDS1-YFPNLS #A5 line . The above results suggest that EDS1-YFPNLS transgenic line #A5 behaves much like a TNL autoimmune background . Therefore , expressing high levels of nuclear targeted EDS1 is sufficient to induce transcriptional defense reprogramming without pathogen activation of a TNL receptor . We performed a genetic suppressor screen of the EDS1-YFPNLS #A3 seedling lethality to identify components contributing to nuclear EDS1 autoimmunity . As shown above , high levels of EDS1-YFPNLS expression in EDS1-YFPNLS line #A3 caused rapid decline and eventual death of 3- to 4-week-old plants at moderate temperature ( 22°C ) ( Figs 1 and S1 ) . The lethality phenotype was fully penetrant at 22°C and stable after three generations of propagating EDS1-YFPNLS #A3 at 28°C . Seeds of EDS1-YFPNLS line #A3 were mutagenized with ethyl methane sulfonate ( EMS ) . This led to the isolation of mutants we have named ‘near death experience’ ( nde ) , which exhibited restored seedling viability and vigor to varying extents at 22°C . Seven putative dominant mutations ( nde1 to nde7 ) were identified by screening EMS mutagenized seedlings directly in the M1 generation ( Fig 6A ) . A further 175 M2 pools ( nde8–175; each derived from ~ 100 M1 plants propagated at 28°C ) were screened at 22°C and ~ 50 additional nde mutants isolated from independent M2 pools ( Fig 6A ) . Here , we describe analysis of a single nde complementation group containing alleles isolated in both the M1 ( nde1-1 , nde1-3 ) and M2 ( nde1-13 , nde1-150 and nde1-175 ) suppressor screens . nde1-1 and nde1-3 were initially scored as dominant suppressor mutations . When grown at 22°C , homozygous nde1-1 and nde1-3 M3 generation seedlings were indistinguishable from wild-type Col , whereas the parental EDS1-YFPNLS line #A3 was severely stunted ( Fig 6B ) . Further lowering of the growth temperature to 16°C did not produce nde1-1 and nde1-3 stunting or chlorosis . Homozygous nde1-1 and nde1-3 plants were backcrossed to the parental EDS1-YFPNLS #A3 line and segregation of the seedling lethality phenotype at 22°C recorded in the F2 generation ( BC1-F2 ) . In both mutants , fully rescued nde , intermediate , and seedling lethal phenotypes segregated in a 1:2:1 ratio ( nde1-1: 79:150:59 , Chi2 = 3 . 28 ) . This mode of inheritance suggests that nde1-1 and nde1-3 are loss-of-function alleles at single semi-dominant loci . EDS1-YFPNLS localization remained entirely nuclear in nde1-1 and nde1-3 leaves , although YFP fluorescence intensity in the mutant lines was reduced compared to EDS1-YFPNLS line #A3 , assessed by confocal laser-scanning microscopy ( Fig 6C ) . Therefore , we reasoned that phenotypic rescue was not due to interference with EDS1-YFP nuclear import but more likely reduced EDS1-YFP nuclear accumulation in nde1-1 and nde1-3 . The SA-response marker gene PR1 was strongly induced in 3-week-old EDS1-YFPNLS line #A3 seedlings shifted to 18°C for 24h , but not in nde1-1 and nde1-3 ( Fig 6D ) . EDS1 displayed a similar expression pattern to PR1 in these seedlings ( Fig 6D ) . Therefore , mutations in nde1-1 and nde1-3 attenuate EDS1 mRNA accumulation under conditions inducing autoimmunity in the parental NLS#A3 line . Accumulation of EDS1-YFPNLS protein was monitored in the same plants . EDS1 levels in nde1-1 and nde1-3 were lower than in the parental NLS#A3 line and comparable to those in line NLS#A5 showing autoimmunity under the same conditions ( Fig 6E ) . Thus , suppression of autoimmunity in nde1-1 and nde1-3 is not solely caused by a reduction of EDS1-YFPNLS levels . The similarity of nde1-1 and nde1-3 phenotypes ( Fig 6B ) prompted us to perform an allelism test . nde1-1 x nde1-3 F1 plants grew normally at 22°C ( S3A Fig ) . Approximately 400 F2 plants originating from four individual nde1-1 x nde1-3 F1 plants also showed no signs of stunting or chlorosis at 22°C ( S3A Fig ) . Therefore , the possibility of F1 phenotypic rescue through actions of independent semi-dominant alleles ( non-allelic non-complementation ) was excluded , unless the independent non-allelic variants are closely linked . Segregation of a specific PCR marker for the nde1-1 mutation generated after A . thaliana whole genome sequencing ( see below ) confirmed that nde1-1 x nde1-3 F1 plants were derived from true crosses ( S3B Fig ) . The nde1-13 , nde1-150 and nde1-175 mutations obtained in screens of EMS-mutagenized M2 plants fully rescued viability of EDS1-YFPNLS #A3 at 22°C and were inherited in a semi-dominant manner . Also , nde1-13 , nde1-150 and nde1-175 were found to be allelic with nde1-1 after crossing and growing PCR-validated seedlings in the F2 generation . We concluded that nde1-1 , nde1-3 , nde1-13 , nde1-150 and nde1-175 form a single complementation group of semi-dominant suppressors of nuclear EDS1 autoimmunity . We performed mapping-by-sequencing of the nde1-1 and nde1-3 mutations ( see Materials and Methods ) [59–61] . A . thaliana Col x Ler SNPs were used to delineate the introgressed Ler portion of DNA containing the eds1-2 mutation [41] to an approximately 6 Mb region in the parental EDS1-YFPNLS#A3 line ( S4 Fig ) . Few polymorphisms with the Col reference sequence were detected in the remainder of the genome . Using SHOREmap [62] , nde1-1 and nde1-3 were mapped to an approximately 5 Mb candidate region on the lower arm of chromosome 3 , coinciding with the parental Ler introgression ( Figs 7A and S5 ) . However , no locus containing a mutation in both nde1-1 and nde1-3 bulk sequences , expected for allelic mutations , was identified . We considered that NDE1 might be a Ler-specific gene or structural variant that is not present in the Col reference genome . Genetic crosses of EDS1-YFPNLS #A3 and nde1-1 to Col and Col eds1-2 , respectively , confirmed that NDE1 encodes a Ler-specific autonecrosis-inducing factor which is lacking in Col ( S3 Table ) . NDE1 was fine-mapped to a 90 kb interval in the Col reference genome by recombination mapping , and a physical contig of this region , which in accession Ler spans 134 kb , assembled using a previous construction of the same locus in Ler [27] ( see Materials and Methods ) . Notably , the NDE1 mapping interval contained QTL3Ler , a polymorphic region covering two TNL RPP1-like paralogs in Col [27] . The RPP1-like nomenclature derives from its close relatedness to a cluster of TNL RPP1 genes in A . thaliana accession Ws-2 whose different paralogs confer isolate-specific Hpa ( formally Peronospora parasitica ) resistance [27 , 63 , 64] . In accession Ler , the QTL3 region has expanded to contain seven complete and one truncated RPP1-like genes ( denoted R1-R8 , Fig 7B ) [27] and the RPP1-likeLer cluster was found to be the causal locus in a recessive deleterious epistatic interaction with Strubbelig-Receptor Family 3 ( SRF3 ) allelic forms from A . thaliana accessions Kashmir ( Kas-2 ) and Kondara ( Kond-0 ) , producing immune-related HI [25] . RPP1-likeLer R1-R8 correspond to DM2a-h paralogs of the DANGEROUS MIX2 locus which underlies multiple negative epistatic interactions among A . thaliana genetic accessions leading to HI [18 , 22] . For simplicity , we now refer to the RPP1-likeLer R1-R8 genes as RPP1-likeLer DM2a-h ( Fig 7B ) . We reasoned that the nde1 mutations might affect one or more of the RPP1-likeLer DM2a-h ( R1-R8 ) genes . Illumina reads from mapping-by-sequencing were re-analyzed against a reference genome containing the Ler NDE1 mapping interval . No canonical EMS changes were identified within the NDE1 mapping interval but manual inspection revealed a prominent drop in read coverage along the RPP1-likeLer cluster in nde1-3 , extending from the DM2c-d ( R3-R4 ) intergenic region to the DM2h ( R8 ) 5’ region ( Fig 7B ) . This was consistent with a large deletion or structural rearrangement in this line , which was confirmed by diagnostic PCR ( S6A Fig ) . Similarly , a 14 bp deletion leading to a premature STOP was detected in the fifth exon of DM2h in nde1-1 ( Figs 7C and S6B ) . No additional SNPs were detected within the mapping interval in nde1-1 bulk sequencing data , indicating that NDE1 is DM2h . The DM2h coding region from nde1-13 , nde1-150 and nde1-175 was therefore obtained by Sanger-sequencing . From this , EMS mutations leading to a premature stop in nde1-13 ( W1129Stop ) or amino acid exchanges R1069C and C945Y , respectively in nde1-150 and nde1-175 , were detected ( Fig 7C ) . Also , EDS1-YFPNLS #A3 necrosis was restored in T2 progeny of the nde1-1 mutant transformed with a RPP1-likeLer genomic DM2h construct ( S7 Fig ) . These results show that DM2h ( R8 ) within the RPP1-likeLer TNL gene cluster interacts genetically with EDS1-YFPNLS resulting in autoimmunity . Having identified RPP1-likeLer DM2h as causal in nuclear EDS1 autoimmunity , we tested whether the EDS1-YFPNLS #A3 or #A5 autoimmune response is accompanied by induced DM2h expression . DM2h expression was significantly reduced in nde1 alleles compared to autoimmune lines EDS1-YFPNLS #A5 and #A3 , but there was only a two-fold increase in DM2h expression in the autoimmune lines ( S8A Fig ) , although these had induced PR1 expression ( Fig 6D ) . This suggests that the DM2h gene itself is not strongly responsive to autoimmunity , in agreement with Alcazar et al ( 2014 ) . A previous screen for senescence-associated mutants in A . thaliana accession Ler identified an EMS-induced mutation , onset of leaf death 3–1 ( old3-1 ) in the cysteine metabolic enzyme-coding locus O-acetylserine ( thiol ) lyase A1 , which also displays negative epistasis with the RPP1-likeLer gene cluster [65 , 66] . Notably , old3-1 caused autonecrosis in Ler , but not Col , and was suppressed by amiRNA silencing of the RPP1-likeLer cluster [65] . More specifically , silencing of DM2g ( R7 ) most closely correlated with the suppression of old3-1 dwarfism [65] . Here , we tested whether the RPP1-likeLer DM2h gene contributes to autonecrosis induced by old3-1 . From a Col x Col eds1-2 cross , we selected two independent near isogenic lines ( NILs ) containing the RPP1-likeLer locus and wild-type EDS1 from Col ( Col-RPP1-likeLer ) . Similarly , we selected two independent NILs containing the RPP1-likende1-1 locus and wild-type EDS1 , but not the EDS1-YFPNLS #A3 transgene from a Col x nde1-1 cross ( Col-RPP1-likende1-1 ) . Hence , the NILs differ mainly in the presence of a 14bp deletion in DM2h ( R8 ) in Col RPP1-likende1-1 but not Col-RPP1-likeLer . We used these NILs first to test whether DM2h ( R8 ) contributes to other resistance responses not related to autoimmunity . NILs were infected with virulent ( Pst DC3000 , Hpa Noco2 ) and avirulent ( Pst AvrRps4 , Hpa Cala2 ) pathogen isolates ( S9 Fig ) . There were no measurable differences in resistance between the NILs , suggesting that DM2h does not act as a helper NLR or generally lower NLR resistance thresholds . The NILs developed normally and were crossed with Ler old3-1 . F2 plants homozygous for old3-1 and either RPP1-likeLer or RPP1-likende1-1 were selected and symptoms of autonecrosis monitored in F3 progeny . old3-1 plants grown at 28°C were not autonecrotic ( Fig 8A ) [65] . At 18°C , Col and Ler were healthy but old3-1 plants became necrotic ( Fig 8A ) . Col/Ler hybrids containing old3-1 and RPP1-likeLer , but not hybrids containing old3-1 and RPP1-likende1-1 ( lacking functional DM2h ) , also became necrotic ( Fig 8A ) . Similarly , PR1 and EDS1 expression was upregulated in Ler old3-1 and Col/Ler RPP1-likeLer old3-1 plants , but not Col/Ler RPP1-likende1-1 old3-1 , old3-1 grown at 28°C , or wild-type plants ( Fig 8B ) . Induction of EDS1 in autoimmune lines was also detectable on western blots ( S10A Fig ) . Dampening of old3-1-induced autonecrosis by the nde1-1 mutation in RPP1-likeLer DM2h was observed when isogenic F1 plants heterozygous for old3-1/OLD3 and either homozygous for RPP1-likeLer or heterozygous for RPP1-likeLer/RPP1-likende1-1 were scored for necrosis ( S10 Fig ) . We concluded from these genetic data that functional RPP1-likeLer DM2h is essential for old3-1 autoimmunity . In a previous study , the RPP1-likeLer DM2c ( R3 ) gene was induced and contributed to negative epistasis between the RPP1-likeLer cluster and SRF3Kas/Kond in temperature-conditioned HI [67] . DM2c was also upregulated in the EDS1-YFPNLS #A3 and #A5 autoimmune lines compared to EDS1-YFP ( S8B Fig ) . Genetic analysis suggested that natively expressed RPP1-likeLer DM2c was necessary but not sufficient for Ler x Kas-2 autoimmunity [67] . We tested for a genetic contribution of RPP1-likeLer DM2h to SRF3Kas/Kond autoimmunity by analyzing F2 progeny from crosses of the Col-RPP1-likeLer or Col-RPP1-likende1-1 NILs with Kas-2 or Kond and scoring for incompatible hybrids ( Fig 8C ) . In accordance with the previously described recessive genetic interaction between the RPP1-likeLer locus and SRF3Kas/Kond [25 , 27] , HI segregated at a 1/16 ratio in F2 progeny from crosses of RPP1-likeLer introgression lines with Kas-2 or Kond ( Fig 8C ) . By contrast , no incompatible hybrids emerged from crosses of RPP1-likende1-1 with Kas-2 or Kond ( Fig 8C ) . We concluded that RPP1-likeLer DM2h is required for conditioning SRF3Kas/Kond HI . Together , the data show that three different deleterious genetic interactions involve the DM2h gene of the RPP1-likeLer TNL complex locus in A . thaliana .
In plants , TNL receptors recognizing different pathogen effectors converge on the nucleocytoplasmic regulator EDS1 to transcriptionally reprogram cells for ETI . Here , we find that low levels of A . thaliana EDS1 enriched in the nuclear compartment ( in EDS1-YFPNLS line #B2 ) are sufficient to confer basal immunity and TNL-triggered ETI against oomycete ( Hpa ) or bacterial ( Pst DC3000 ) pathogen strains ( Fig 2 ) . Therefore , a small nuclear EDS1 pool appears to be competent in disease resistance signaling . In an earlier study we proposed a positive cytoplasmic role for EDS1 in A . thaliana immunity , based on intermediate resistance phenotypes of lines in which EDS1-YFP was mis-localized to the cytoplasm [38] . A reinterpretation of those data is that residual low amounts of nuclear EDS1 after subcellular mis-localization can confer resistance , at least with respect to the pathogen strains tested here . Nuclear-targeted EDS1 has identical genetic requirements for its signaling partner genes PAD4 and SAG101 as that of native , nucleocytoplasmic EDS1 in A . thaliana wild-type basal and TNL immunity ( Fig 4 ) . This highlights nuclear actions of PAD4 ( or SAG101 when PAD4 is not present ) in promoting EDS1 resistance , probably as EDS1 heteromeric complexes [40 , 52] . We find that increased levels of A . thaliana nuclear-enriched EDS1 lead to autoimmunity exhibiting characteristic temperature-conditioned defense gene expression , accumulation of SA , and stunting of plant growth ( in EDS1-YFPNLS line #A5 ) or lethality ( in EDS1-YFPNLS line #A3 ) ( Figs 1 , S1 , 2 and 5 ) . Thus , above a certain threshold , nuclear EDS1 produces many of the hallmarks of TNL autoimmunity . In an extensive genetic screen for suppressors of the autoimmune response in EDS1-YFPNLS line #3 , we identify as causal four independent ( nde1 ) mutations in one gene , DM2h ( R8 ) within the RPP1-likeLer TNL complex locus ( Figs 6 and 7 ) [18 , 22 , 27] . Further genetic analysis shows that RPP1-likeLer DM2h underlies two additional cases of A . thaliana autoimmunity , one with a mutated form of a cysteine metabolic enzyme ( O-acetylserine ( thiol ) lyase A1 ) gene in old3-1 within Ler [65 , 66] , the other causing HI with allelic forms of a receptor-like kinase gene , SRF3 , present in A . thaliana Kas-2 and Kond and other Central Asian accessions ( Fig 8 ) [25 , 27] . These data show that three different deleterious genetic interactions involving the RPP1-likeLer gene cluster converge on DM2h . Because RPP1-likeLer DM2h ( R8 ) is necessary for nuclear EDS1-PAD4 autoimmunity and defense gene expression ( Figs 6 , 7 and 8 ) , we conclude that the DM2h protein directly or indirectly drives EDS1-PAD4 defense amplification . In one model , DM2h behaves as a weakly autoactive TNL protein which , in its native Ler background or in the NILs expressing wild-type EDS1 ( Figs 8 and S9 ) , is effectively constrained . In other genetic backgrounds , DM2h weak autoactivity can be exposed in a temperature-dependent manner , as in incompatible hybrids ( Fig 8 ) . In this model , DM2h initiates EDS1/PAD4 signaling and DM2h autoactivity becomes deleterious when EDS1 nuclear accumulation rises above a threshold , producing autoimmunity . An alternative explanation for dependence of EDS1-YFPNLS line #A3 autoimmunity on RPP1-likeLer DM2h is that EDS1 nuclear over-accumulation causes transcriptional mis-regulation of the DM2h gene as part of a feed-forward expression loop . However , DM2h expression was not significantly induced in EDS1-YFPNLS lines #A3 or #A5 compared to non-autoimmune EDS1-YFP containing RPP1Ler . Similarly , DM2c was only mildly up-regulated in autoimmune EDS1-YFPNLS lines ( S8 Fig ) . In a previous analysis to identify genes within the RPP1-likeLer locus underlying HI with SRF3Kas/Kond , strong up-regulation of DM2cLer ( R3 ) but not DM2hLer ( R8 ) correlated with autoimmunity [67] . In that study , DM2hLer was excluded as a causal gene based on the autoimmune phenotypes of selective amiRNA knock-down lines . Nevertheless , no single gene within the RPP1-likeLer locus was able to reconstitute temperature-conditioned HI with SRF3Kas/Kond , leading the authors to propose co-actions of DM2cLer ( R3 ) with one or more RPP1-likeLer genes in the genetic incompatibility [67] . Our mapping of four independent nde1 mutant alleles ( nde1-1 , nde1-13 , nde1-150 , nde1-175 ) to the DM2hLer gene , and establishing that nde1-1 also suppresses HI in crosses with A . thaliana strains Kas-2 or Kond , provides genetic proof that DM2hLer ( R8 ) is a key factor in nuclear EDS1 autoimmunity and Ler x Kas-2/Kond immune-related HI . Interestingly , DM2h ( R8 ) contains a predicted N-myristoylation motif and a bipartite NLS ( S11 Fig ) . The autoimmunity-inducing genetic interactors SRF3Kas/Kond and EDS1-YFPNLS localize to the plasma membrane [25] and nucleus ( Fig 1C ) , respectively . Further analysis is needed to determine whether DM2h activity involves its membrane-tethering and/or nuclear localization . It is significant that high levels of nucleocytoplasmic EDS1 expressed under its native or a constitutive ( Cauliflower Mosaic virus 35S ) promoter do not lead to autoimmunity [this study , 38 , 40] , unlike nuclear-targeted EDS1 . We deduce from this that EDS1 nucleocytoplasmic trafficking through nuclear pore complexes [38] limits potentially hazardous actions of EDS1 in nuclear resistance signaling . This is likely to be important for maintaining cellular homeostasis and the trade-off between defense and fitness , especially under conditions when the plant is not being attacked by pathogens . Fusing EDS1 to a strong NLS might prolong transcriptional reprogramming activity of a nuclear EDS1 pool or draw other components into the nucleus . Because the EDS1-YFPNLS #A5 plants do not show macroscopic defects until 4–5 weeks after planting at 22°C , EDS1-YFPNLS is unlikely to cause severe clogging of the NPC import/export machinery which would have immediate effects on physiology and development . Artificially raising the EDS1 nuclear pool likely exposes ‘dangerous’ TNL alleles such as DM2h and drive cells and tissues into an activated immune response without a pathogen trigger . Here , DM2h immune reactivity does not appear to enhance ETI conditioned by other TNL genes ( S9 Fig ) , suggesting a degree of specificity in DM2h co-action with DM2c and EDS1 . Further analysis is required to establish whether DM2h and DM2c interact genetically or molecularly with each other , as found for a number of functional NLR and NOD-LRR receptor pairs [3 , 5] , or indeed with nuclear EDS1 , in the different autoimmunity backgrounds . Whatever the mechanism of resistance deregulation , DM2 locus steering of EDS1-YFPNLS #A5 plants towards defense at the expense of growth involves SA-dependent and SA-independent signaling sectors ( Figs 5 and S2 ) , broadly resembling defense pathway bifurcations in pathogen-triggered EDS1-PAD4 basal and TNL immune responses [41 , 54 , 55] . Notably , stunting of EDS1-YFPNLS line #A5 at moderate temperature was not alleviated by mutations in SA biosynthesis ( sid2-1 ) or SA signaling ( npr1-1 ) genes . By contrast , growth defects and necrosis , respectively , in a moderately incompatible Ler x Kas-2 recombinant inbred line ( RIL ) and a severely dwarf RPP1-likeLer x Kas-2 NIL , were fully suppressed by sid2-1 [27] . The varying penetrance of SA pathway mutants in these two EDS1-dependent autoimmune backgrounds suggests that the consequences of EDS1 over-accumulation in EDS1-YFPNLS line #A5 versus RPP1-likeLer x Kas-2 HI are not identical , possibly due to different genetic modifiers or pathway fine-tuning between nuclear and nucleocytoplasmic EDS1 . Immune-related incompatibilities in plants between natural genetic variants ( HI ) often involve highly variable NLR gene clusters [18] . A body of evidence suggests that HI can expose divergent evolutionary trajectories of immune receptor genes through genetic drift , coevolution or local adaptation [18 , 23 , 67] . The occurrence of HI in crosses between genetic backgrounds might also shape which immune receptor or receptor cofactor genes can be assembled in any one genome [18 , 19 , 23] . The A . thaliana polymorphic RPP1-like DM2 locus is especially remarkable in that genes within it underlie multiple , independent epistatic interactions causing autoimmunity . Whereas the DM2 region in A . thaliana Col reference strain and related species Arabidopsis lyrata consists of just two RPP1-like genes [18 , 67] , the locus has expanded to contain seven complete RPP1-like genes in A . thaliana Ler ( Fig 7 ) and eight in accessions Bla-1 and Uk-1 [18] . These three accessions produce different DM2-based incompatibilities suggestive of a locus undergoing rapid evolutionary change [9 , 12 , 18] . The signatures of host-pathogen co-evolutionary conflict are especially evident at A . thaliana polymorphic RPP gene clusters or allelic variants recognizing different isolates of the adapted downy mildew pathogen Hpa [10 , 68 , 69] . The originally mapped RPP1 resistance locus in accessions Ws-2 and Nd-0 encodes TNL variants ( namely RPP1-WsB and RPP1-NdA ) that confer allele-specific recognition of Hpa-derived Atr1 effector proteins by direct effector binding at the TNL receptor C-terminal LRR domain , cooperating with the central NB-ARC activation domain [63 , 64 , 70 , 71] . Although it is not known whether genes within the different DM2 haplotypes recognize specific Hpa or other pathogen strains , the DM2h LRR domain has a signature of diversifying selection among A . thaliana accessions , suggestive of variation in pathogen effector recognition surfaces [18] . In our study , the DM2hLer nde1-175 mutation causes a non-synonymous C945Y exchange in an LRR consensus sequence residue of LRR9 ( Figs 7C and S11 ) . The C/N residues of the LxxLxLxxN/CxxL consensus form hydrogen bonds with backbone carbonyl groups throughout the entire LRR solenoid fold [72 , 73] . The nde1-175 mutation might thus perturb the overall shape of the DM2h-LRR domain or the local arrangement of LRR9 and neighboring LRRs . An R1069C exchange in DM2hLer nde1-150 lies within the C-terminal LRR-flanking region . A W1129SToP mutation in nde1-13 causing a truncation encompassing only 41 amino acids ( S11 Fig ) , points to functional importance of this extreme C-terminal region . Unexpectedly , deletion mutants at the RPP1-likeLer DM2 locus in nde1-1 ( 14 bp ) and nde1-3 ( ~ 50 kb ) ( Fig 7 ) were obtained in the M1 generation screen of EMS-mutagenized EDS1-YFPNLS line #A3 . When these M1 plants were initially grown under restrictive temperature conditions , healthy young leaves emerged from necrotic rosettes suggesting that the lesions in nde1-1 and nde1-3 might not have originated in embryonic cells targeted by EMS treatment but from recombination events later in seedling development . A spontaneous recombination event within the Col RPP5 locus was reported for the bal variant in which there is 55 kb duplication encompassing the SNC1 gene [74] . In nde1-1 , a 4 bp ( ATTG ) micro-syntenic sequence flanking the 14 bp deletion might have directed somatic homologous recombination ( SHR ) at this position ( Fig 7C ) . Although the origin of these genetic lesions remains speculative , unequal crossing-over and illegitimate recombination events are known to create sequence and locus-size variation in Resistance gene clusters [8 , 75] . Also , gene recombination rates were reported to increase with biotic stress [76–78] . It is conceivable that the nde1-1 and nde1-3 alleles represent snapshots in the evolution of a plant NLR gene locus . The origin of the RPP1-likeLer DM2a-h haplotype was recently traced to a natural A . thaliana population in Gorzów Wielopolski , Poland [67] . Genetic analysis of plants within this wild population showed that the RPP1-likeLer DM2a-h locus has been maintained in genetically different individuals over many generations [67] . Further study of the Gorzów population will allow an exploration of the genetic and ecological forces shaping the evolution of this interesting TNL complex locus [67] .
Wild-type Arabidopsis thaliana accessions used were Col-0 ( Col ) and Landsberg erecta ( Ler ) . Col eds1-2 [41] , pad4-1 [79] , sag101-1 , pad4-1 sag101-1 [51] , sid2-1 [57] , npr1-1 [80] , Ler old3-1 [66] mutant lines and the EDS1-YFP transgenic line [38] are published . An SV40 NLS was introduced at the 3’ end of the mYFP open reading frame by PCR , and a binary vector containing a BASTAR plant-selectable marker and a pEDS1:gEDS1-YFPNLS expression cassette within the T-DNA borders was generated as described [38] . pad4-1 , sag101-1 , sid2-1 and npr1-1 mutations were introduced into EDS1-YFPNLS line #A5 by crossing and selecting homozygous backgrounds using PCR-based gene-specific markers ( S3 Table ) . Plants were grown in soil at a 20°C: 22°C night: day cycle ( 200μE m2 s-1 ) and 60% relative humidity . For suppression of autoimmunity , plants were germinated at 22°C for 7d , and then shifted to 26°C/28°C ( night/day ) with 10h illumination . For homogeneous and stringent autoimmunity induction , plants were grown at either 20°/22°C or 26°C/28°C , and were then shifted to 18°C . Spray inoculation of 4- to 6-week-old plants with Pseudomonas syringae pv tomato ( Pst ) strain DC3000 or Pst DC3000 expressing the effector AvrRps4 ( Pst AvrRps4 ) was performed with bacterial suspensions of 1x107 colony forming units ml-1 as described [38] . Bacterial entry was routinely checked by determining in planta bacterial titers at 3 hpi , and was similar between all genotypes used in this study . Conidiospore suspensions of 4x104 spores ml-1 were used for Hyaloperonospora arabidopsidis ( Hpa ) infections . Lactophenol trypan blue ( TB ) staining of Hpa- and mock-infected leaves and pathogen spore counts were as described previously [81] . Disease resistance assays were repeated independently at least three times with similar results . Total RNA of three independent biological replicates from 4-week-old Col and EDS1-YFPNLS #A5 leaf tissues was isolated with an RNeasy Plant Mini kit supplied with RNase-Free DNase set ( Qiagen ) according to the manufacturer´s instructions . RNA quality was assessed on a Bioanalyzer ( Agilent ) . Biotinylated cRNA was prepared according to manufacturer’s instructions from 1 μg total RNA ( MessageAmp II-Biotin Enhanced Kit; Ambion ) . After amplification and fragmentation , 12 . 5 μg of cRNA was hybridized for 16 h at 45°C to a GeneChip ATH1-121501 Genome Array . GeneChips were washed and stained with Fluidics Script FS450-004 in the Affymetrix Fluidics Station 450 and scanned using a GeneChip Scanner 3000 7G . The data were analyzed with Affymetrix GeneChip Operating Software version 1 . 4 using Affymetrix default analysis settings and global scaling as the normalization method . Probe signal values were subjected to the GeneChip robust multi-array average ( GC-RMA ) summarization algorithm [82] to obtain expression level values . The microarray data were submitted to Gene Expression Omnibus ( accession number GSE65415 ) . Results were analyzed by the following linear model using the lmFit function in the limma package in the R environment: log2 ( expression level value ) sample + replicate . The eBayes function in the limma package was used for variance shrinkage in calculation of p-values . The Storey’s q-values were calculated using the q-value function in the q-value package from the p-values [83] . 1045 genes with at least 4-fold changes and q-value < 0 . 01 in EDS1-YFPNLS #A5 compared to Col were selected for the clustering analysis . Expression values for these 1045 genes were extracted from publicly available data sets and were used for the clustering analysis . Hierarchical clustering analysis was performed using Cluster 3 . 0 software [84] with uncentered Pearson correlation and complete linkage , and visualized by Treeview software [84] . Total RNA was extracted using RNeasy Plant Mini Kit or TRI Reagent ( Ambion ) . Reverse transcriptase ( RT ) reactions were performed with 1–2 μg of total RNA using SuperScriptII™ ( Invitrogen ) or RevertAid ( Thermo Scientific ) . RT reactions were diluted 1:5 and 2 μl used for qPCR reactions on a Bio-Rad iQ5 or CFX Connect Real Time-PCR Detection System with EvaGreen ( Biotium ) or AbsoluteBlue ( Thermo Scientific ) dyes . UBQ10 ( At4g05320 ) transcript levels were used as an internal reference in all samples . Primer efficiencies were between 90–110% for all oligos , and data was analyzed using dCt . Gene expression was evaluated in at least three independent experiments with similar results . Total protein extracts were prepared by grinding leaf tissues in liquid nitrogen . Samples were resuspended in 2x Laemmli loading buffer ( 0 . 5 w/v ) , boiled for 10 min and centrifuged to remove cell debris . Proteins were separated by SDS-PAGE and electro-blotted to nitrocellulose membranes for protein gel blot analysis . Equal loading was monitored by staining membranes with Ponceau S ( Sigma-Aldrich ) . Anti-EDS1 [51] , anti-GFP ( Roche ) , anti-Histone H3 ( Agrisera ) and anti-cFBPase ( Agrisera ) antibodies and secondary antibodies coupled to AP or HRP ( Sigma , GE Healthcare ) were used for detection . For live cell imaging , Arabidopsis leaves were examined with a Zeiss LSM780 confocal laser-scanning microscope directly after removing the leaves from plants grown at the different temperature regimes . Quantification of free and total SA in leaf tissues was done as previously described [85] . Similar results were obtained in at least three independent experiments . For isogenic mapping , nde1-1 and nde1-3 mutant plants ( genotype Col eds1-2 pEDS1:EDS1-YFPNLS nde1-1/-3 ) were backcrossed to the parental transgenic line EDS1-YFPNLS#A3 and plants showing the nde phenotype were selected from segregating F2 populations . Leaf material from > 100 segregants was pooled and DNA extracted using a DNeasy Maxi DNA Kit ( Qiagen ) . DNA pools and DNA from the parental line EDS1-YFPNLS#A3 were used for library construction , and ~ 30M 100 bp single end reads per sample were produced on an Illumina HiSeq2500 at the Max-Planck Genome Centre , Cologne and analyzed with the short read analysis pipeline SHORE [86] . High quality ( SHORE score > 20 ) SNPs between nde1-1/-3 and the parental line were used in SHOREmap backcross . SNPs with allele frequency estimates ≥ 0 . 8 were considered as significant , defining mapping intervals from 10–23 Mb on chromosome 3 for both mutations . SNPs in this region were analyzed for their effect on gene coding sequences using TAIR 10 . nde1-1 was further fine-mapped using DNAs from ~ 400 single phenotyped plants from the same BC1-F2 population . SNPs detected by Illumina sequencing were converted to CAPS or dCAPS markers , and a final mapping interval supported by several recombinants on each side was defined by markers at 16 . 118 Mb and 16 . 428 Mb in the Col reference genome . No more additional SNPs for marker generation were available in this interval . For further mapping , transgenic line EDS1-YFPNLS#A3 was outcrossed to Col . F2 plants containing the EDS1-YFPNLS transgene were selected by BASTA resistance and individuals containing the compatible ColNDE1 region ( no signs of autonecrosis ) were phenotypically selected by shifting BASTA-resistant plants to 18°C for 10d . Under these conditions , the EDS1-YFPNLS transgene and NDE1 in the hemizygous states were sufficient to induce macroscopic growth phenotypes . The ColNDE1 region was mapped using ~ 400 plants with Ler/Col markers . A final mapping interval of ~ 90 kb in the Col reference was defined by markers JS676/677 and JS678/679 at 16 . 164 and 16 . 253 Mb on chromosome 3 , respectively . A physical contig from Ler containing both markers was constructed using FJ446580 . 1 [27] and scaffold 1526 from a reference-guided Ler assembly [87] . An updated reference containing this contig , but not the respective region from Col , was created . Illumina reads were mapped against this reference using a CLC Genomics Workbench to generate data for Fig 7B and 7C . Col NILs containing RPP1-likeLer or RPP1-likende1-1 were generated by crossing Col eds1-2 x Col or Col x nde1-1 , respectively . F2 individuals with a recombination event between Col EDS1 and the RPP1-likeLer cluster were selected using Ler/Col markers JS663/664 , JS655/656 and an eds1-2/EDS1 marker . Absence of the EDS1-YFPNLS insertion was tested by PCR using oligos JS661/328 and JS661/665 . Ler old3-1 plants were pollinated from four individual lines homozygous for RPP1-likeLer/nde1-1 and hetero- or homozygous for EDS1/eds1-2 . For Sanger sequencing of additional nde1 alleles , DM2h ( R8 ) from nde1-13 , -150 , -175 was PCR-amplified with JS724/725 , and the PCR product cloned into a compatible Golden Gate plasmid by BsaI cut/ligation [88] . Two independent clones per allele were sequenced , and each contained only a single mutation . The same DM2h amplicon was also cloned from Ler , the resulting construct transformed into Agrobacterium strain GV3101 pMP90 and nde1-1 mutant plants transformed by floral dip . Primary transformants were selected on media containing Kanamycin , transferred to soil and cultivated at 28°C . T2 seeds were first cultivated at 28°C and subsequently shifted to 18°C to monitor induction of autoimmunity . Oligonucleotides are listed in S4 Table . For total protein extracts , 100 mg of Arabidopsis leaves were homogenized in liquid nitrogen and 500 μl of SDS extraction buffer ( 4% w/v SDS , 100 mM Tris-HCl pH 7 . 6 , 100 mM DTT , protease and phosphatase inhibitor cocktails ) were added . Samples were boiled for 10 min , centrifuged at 12000 x g for 10 min , and the supernatant recovered . Nuclei were extracted from 2g fresh leaves using a previously described method [89] . Briefly , fresh leaves were chopped in 30 ml nuclear extraction buffer ( 2 . 0 M hexylene glycol ( 2-methyl-2 , 4-pentandiol ) , 20 mM PIPES/KOH pH7 . 0 , 10 mM MgCl2 and 5 mM 2-mercaptoethanol , protease and phosphatase inhibitor cocktails ) , filtered through 5 layers of cheesecloth , and subjected onto a 30% / 80% percoll density gradient . After centrifugation at 2000 x g for 30 min , the layer between 30–80% percoll was collected , loaded on 30% percoll , and re-centrifuged at 2000 x g for 30 min . The pellet was collected , mixed with 100 μl SDS-extraction buffer , and boiled for 10 minutes . Samples were centrifuged ( 10000 x g , 10 min ) , and the supernatant was collected as the nuclear protein fraction . Protein concentrations were determined using Pierce 660 nm absorbance assay in presence of IDCR reagent following the manufacturer’s protocol ( Thermo Fisher Scientific ) . Coomassie Brilliant Blue staining of membranes was used as loading control .
|
Plants tune their cellular and developmental programs to different environmental stimuli . Central players in the plant biotic stress response network are intracellular NLR receptors which intercept specific disease-inducing molecules ( effectors ) produced by pathogenic microbes . Variation in NLR gene repertoires between plant genetic lines is driven by pathogen selection pressure . One evolutionary question is how new , functional NLRs are assembled within a plant genome without mis-activating defense pathways , which can have strong negative effects on growth and fitness . This study focuses on a large , polymorphic sub-class of NLR receptors called TNLs present in dicotyledenous plant lineages . TNL receptors confer immunity to a broad range of pathogens . They also frequently underlie autoimmunity caused by their mis-regulation or deleterious allelic interactions with other genes in crosses between different genetic lines ( hybrid incompatibility , HI ) . TNL pathogen-triggered and autoimmune responses require the conserved nucleocytoplasmic protein EDS1 to transcriptionally reprogram cells for defense . We discover in Arabidopsis thaliana that high levels of nuclear-enriched EDS1 induce transcriptional activation of defenses and growth inhibition without a pathogen effector stimulus . In a mutational screen , we identify one rapidly evolving TNL gene , DM2hLer , as a driver of nuclear EDS1 autoimmunity . DM2hLer also contributes to two separate cases of EDS1-dependent autoimmunity . Genetic cooperation between DM2hLer and EDS1 suggests a functional relationship in the transcriptional feed-forward regulation of defense pathways .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"plant",
"anatomy",
"medicine",
"and",
"health",
"sciences",
"immunology",
"brassica",
"plant",
"biotechnology",
"plant",
"science",
"model",
"organisms",
"genetically",
"modified",
"plants",
"plants",
"extraction",
"techniques",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"genetically",
"modified",
"organisms",
"protein",
"extraction",
"gene",
"expression",
"leaves",
"genetic",
"loci",
"agriculture",
"plant",
"and",
"algal",
"models",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"agricultural",
"biotechnology",
"autoimmunity",
"organisms"
] |
2016
|
Arabidopsis thaliana DM2h (R8) within the Landsberg RPP1-like Resistance Locus Underlies Three Different Cases of EDS1-Conditioned Autoimmunity
|
The phenotypic outcome of a mutation cannot be simply mapped onto the underlying DNA variant . Instead , the phenotype is a function of the allele , the genetic background in which it occurs and the environment where the mutational effects are expressed . While the influence of genetic background on the expressivity of individual mutations is recognized , its consequences on the interactions between genes , or the genetic network they form , is largely unknown . The description of genetic networks is essential for much of biology; yet if , and how , the topologies of such networks are influenced by background is unknown . Furthermore , a comprehensive examination of the background dependent nature of genetic interactions may lead to identification of novel modifiers of biological processes . Previous work in Drosophila melanogaster demonstrated that wild-type genetic background influences the effects of an allele of scalloped ( sd ) , with respect to both its principal consequence on wing development and its interactions with a mutation in optomotor blind . In this study we address whether the background dependence of mutational interactions is a general property of genetic systems by performing a genome wide dominant modifier screen of the sdE3 allele in two wild-type genetic backgrounds using molecularly defined deletions . We demonstrate that ∼74% of all modifiers of the sdE3 phenotype are background-dependent due in part to differential sensitivity to genetic perturbation . These background dependent interactions include some with qualitative differences in the phenotypic outcome , as well as instances of sign epistasis . This suggests that genetic interactions are often contingent on genetic background , with flexibility in genetic networks due to segregating variation in populations . Such background dependent effects can substantially alter conclusions about how genes influence biological processes , the potential for genetic screens in alternative wild-type backgrounds identifying new loci that contribute to trait expression , and the inferences of the topology of genetic networks .
Fundamental to the logic of genetic analysis is that the observed variation in a phenotype for a genetically mediated trait is causally linked to one or more DNA lesions/variants . However , it is well known that the phenotypic effects of many individual mutant alleles are context dependent , with respect to environmental influences , as well as the “wild-type” genetic background in which the mutation is observed . Indeed , genetic background has long been known to influence observed phenotypic expression across traits , organisms , and a range of allelic effects , including hypomorphs , amorphs/nulls and neomorphs [1]–[9] . These results make it clear that the phenotypic effects of a mutation ( i . e . penetrance and expressivity ) are themselves “complex traits” , subject to environmental and polygenic influences [1] . Far beyond being a minor curiosity in genetics , the background dependent effects of a number of mutations have been at the heart of debates over the conclusions and the ability to replicate key findings from several studies , including the genetics of life span [10]–[14] , stress tolerance [15]–[17] and pigmentation [18]–[20] . Although the basic influence of genetic background on the expressivity of mutations is well documented , the wider consequences of such influences are poorly understood [21] . In particular , the extent to which wild-type background influences the magnitude and sign of genetic interactions remains unclear . Research to date addressing this question [4] , [22] , [23] , has largely focused on a small set of mutations , and defined genetic backgrounds . Recent work has demonstrated that the magnitude of genetic interactions can be influenced by environmental factors [24] , and even ploidy level [25] . Yet the generality of such findings remains uncertain . Thus this remains an essential , but poorly explored area of fundamental genetics , as our understanding of epistasis , and our inferences of the topology of genetic networks are often derived from studies of genetic interactions [26]–[33] . In addition , modifier screens have been extremely important , and have identified large numbers of genes that interact to influence the visible expression of the phenotype of the focal mutation , even when the modifier may not have a visible phenotype by itself [34] , [35] . We have previously shown that the phenotypic effects of an allele of the scalloped gene ( sdE3 ) in Drosophila melanogaster is profoundly influenced by wild-type genetic background ( Figure 1B ) , with effects extending to wing disc transcriptional profiles [36] . One gene that was transcriptionally regulated in a background-dependent matter , optomotor blind/bifid ( omb/bi ) , was then examined in a double mutant combination with sdE3 . We demonstrated that the phenotypic consequence of the interaction between these mutations was markedly influenced by wild-type genetic background . In one wild-type background the double mutant combination resembled the individual sdE3 phenotype , while in the other wild-type background , the omb mutation behaved as a strong synthetic enhancer of sd [36] . Our findings clearly demonstrate the influence of wild-type genetic background on this genetic interaction , but an important challenge is to determine whether such context dependent effects are widespread . To address this question we performed a genome wide-screen for dominant modifiers of sdE3 using two wild-type genetic backgrounds . Our results suggest that the majority ( ∼74% ) of all modifiers are background-dependent . The background-dependence of the modifier alleles are in part due to the wild-type strains differing in overall sensitivity to mutational perturbations . Using a subset of the deletions spanning the range of phenotypic effects of modifiers , we observed that the interaction effects were consistent using an additional allele , sdETX4 . Furthermore , we show that the deletion effects are a result of the interaction with mutations at the sd locus , and not a simple consequence of haplo-insufficiency in the genomic region of the deletion . We also demonstrate that the background-dependent interactions of modifiers with sdE3 are linked to the same genomic regions that contribute to the background-dependent effects of the allele itself . We argue that the phenotypic expressivity of mutations can be considered a quantitative trait , and a more comprehensive , context-dependent view of the effects of mutations needs to emerge .
Genetic modifier screens are powerful tools to both identify interacting factors that contribute to signaling networks , as well as to infer their topology . This approach has shaped our understanding of the genetic basis of many traits , across numerous organisms . However little is known about how wild-type genetic background influences genetic interactions . We previously demonstrated that the genetic interaction between mutations in two genes , sd and omb , is dependent on genetic background [36] . To determine if such an effect is a general phenomenon we performed an analysis of genome-wide genetic interactions between the sdE3 mutation and deletions generated in otherwise isogenic backgrounds spanning the autosomes of Drosophila . We first verified that deletions spanning a number of putative candidate genes ( Dll , wg , vg ) previously demonstrated to interact with sd modify the sdE3 phenotype . In each of these instances the deletions confirmed previous expectations for the interaction ( Figure S1B ) . We then screened the autosomes , with two independent sets of genomic deletions , DrosDel [37] and Exelixis/BSC [38] , [39] , each generated in an independent isogenic progenitor background ( Figure 1B ) . In total 723 deletion-bearing strains ( spanning ∼90% of the autosomal genome ) were crossed to sdE3 in each wild-type background . F1 males hemizygous for the sdE3 mutation and heterozygous for the deficiencies were scored . For the 198 deletion strains that consistently modified the sdE3 wing phenotype , ∼74% of the observed effects were dependent on wild-type ( Oregon-R vs . Samarkand ) genetic background ( Table 1 ) . Frequently , the background contingency was a result of severe effects in one wild-type genetic background , with modest or no effects in the other ( Figure 1A and 2 , Figure 3A ) . A complete list of modifier regions , and putative candidate genes can be found in Table S1 . An example of the physical location and contribution of these effects is illustrated using the left arm of chromosome 3 ( Figure 3 , Figure S4 ) , where background-independent and -dependent effects are illustrated , including some deletions with opposing effects in terms of modifying the sdE3 phenotype . We confirmed these results using a linear model ( ANOVA ) , by asking what proportion of all “significant” modifiers also had a “significant” interaction effect between genetic background and the deletion . Based upon these criteria ∼79% of modifiers demonstrated background dependence . While each cross was carried out independently , there were a large number of crosses performed , and each deletion bearing genotype was compared to a common set of controls from within each block of crosses ( see methods ) . Therefore we utilized several methods that adjust for multiple comparisons . While these methods will decrease the number of deletions deemed modifiers using standard comparisons ( i . e . α = 0 . 05 ) , we are primarily interested in the proportion of such modifiers that are due to background dependent effects . Using False Discovery Rate ( FDR ) we observed a similar frequency ( ∼78% ) as with unadjusted p-values , while with the sequential Bonferroni ( Holm ) it was ∼68% . Regardless of the exact approach used , it is clear that the vast majority of modifiers recovered are background dependent . We performed this screen using two different sets of deletions , each of which varied in the size of the deletion . We observed little association between deletion size and severity of phenotypic modification ( Samarkand: correlation-0 . 09 & -0 . 08 using Exelixis & DrosDel respectively; Oregon: −0 . 061 & −0 . 067 using Exelixis & DrosDel deletions respectively , Figure S5 ) . The lack of association between size of deletion and magnitude of effect suggests that it is unlikely that the observed effects are due to the number of genes perturbed in each deletion . These key results suggest that at least in sensitization screens , and possibly for many studies of genetic interaction , wild-type genetic background will have profound influences on the range of phenotypes observed and the modifiers that are identified , with only a subset of modifiers being background-independent . Using Flymine and Droid [40] , [41] as well as literature mining we examined all of the previously identified genes that act as genetic modifiers , protein-protein interacting partners , or are targets of transcriptional regulation by SD . From these sources we collated evidence for 19 genes that were covered by deletions in this screen ( i . e . excluding genes on the X ) , and all but one ( sens ) were recovered as genetically interacting with sdE3 ( Figure 3B ) . However , more than 50% of these specific loci demonstrated background-specific interactions with sdE3 , including vg , which is known to physically interact with SD to form a heterodimer , and is transcriptionally regulated by this complex . Several well-known genetically or physically interacting genes ( such as salm and yki ) showed surprisingly mild enhancement of the phenotype , which may be a result of the particular wild-type backgrounds used in this study . These findings suggest that even for well-characterized interacting genes , the influence of genetic background can be substantial , consistent with the flexible nature of genetic interactions . An important caveat to this interpretation is that many of these deletions may contain more than one gene . This could potentially mean that the interaction is due to both the deletion of the focal gene as well as other loci nearby . Yet , as described above , we observed no evidence for a relationship between deletion size and magnitude of effect , suggesting that this may be a minor contributing factor . To further validate , refine , and extend our analysis we quantified a subset of 44 of the Exelixis deletion lines that spanned the range of modifier phenotypes across both severity and background-dependence . Interestingly ( Figure 4 ) , the background-dependent interactions are clearly a result of both specific differences with respect to the nature of sensitizing mutational effects in each background , as well as to the degree of sensitivity to mutational perturbation . Indeed , the sdE3/Y; Deletion/+ combinations in the Oregon-R wild-type background demonstrated considerably more variation between deletion strains , compared to the same genotypes in Samarkand ( Figure 4 ) . Despite the fact that the sdE3 mutation in the Oregon-R background had more severe loss of wing tissue ( Figure 1 , Figure S1 ) , the range of both enhancement and suppression exceed that of the same mutation in the Samarkand background ( Figure 4 ) . The between deletion co-efficient of variation ( CV ) for wing size in the Oregon-R background is approximately double that ( 0 . 34 ) of the Samarkand background ( 0 . 15 ) . These results were confirmed using a Levene's test with a non-parametric bootstrap . Despite the differences in both degree and spectrum of sensitivity , there was still a moderate correlation of effects of the sdE3/Y; Deletion/+ combinations ( 0 . 66 , CI ( 0 . 46 , 0 . 8 ) ) across the two wild-type backgrounds . These data indicate many of the modifiers are acting in the same direction , although vary for magnitude of effect . Interestingly , even the non-genetic component of phenotypic variation observed for Oregon-R sdE3/Y; +/+ in crosses to the wild-type deletion progenitor shows considerably greater phenotypic variation for wing size compared to Samarkand ( Figure 4 ) , although it is unclear if this is related to the changes in within strain variation ( robustness ) . While the semi-quantitative measure of wing size used for the initial screen , and quantitative measure described above are highly correlated ( see methods ) , a few putative modifier regions failed to replicate in the tertiary validation cross with quantitative measures . Similarly a few deletion lines that were expected to not have an effect ( based on the initial screen ) , did have one with the quantitative measure . However these potential false positives and negatives are few , of similar numbers , and thus are not expected to influence the overall conclusions . One possible explanation for these results would be that the deletions influenced wing size , per se , and the results were not a specific consequence of the interaction between sd and the deletion . To investigate this we quantitatively examined females who were heterozygous for the sdE3 mutation and for the deletions ( i . e . sdE3/+ ; Deletion/+ ) across each genetic background . These females have qualitatively “wild-type” wings , and previous work did not observe an effect of sdE3 on wing size in females as heterozygotes [42] ( although it did influence wing shape ) . Therefore we quantified these females across the same set of deletions as described above . If the deletions were not generally acting as modifiers of the “sensitized” sd mutant phenotype in hemizygous males , but as general modulators of size , then we would expect a strong positive correlation between the effects on size in males and females ( sdE3/+ ; Deletion/+ vs . sdE3/Y ; Deletion/+ ) . The correlation between Samarkand and Oregon-R sdE3/+ ; Deletion/+ females was ∼0 . 8 , suggesting that the effects of the deletions on overall wing size is similar across backgrounds . However the correlations within each background ( i . e . sdE3/+; Deletion/+ vs . sdE3/Y ; Deletion/+ ) were 0 . 22 , ( CI −0 . 08 , 0 . 49 ) , and 0 . 21 , ( CI −0 . 08 , 0 . 48 ) respectively , and neither case was significantly different from 0 . The lack of a correlation indicates that the influence of the deletions in sdE3 hemizygous males is largely independent of any effects on overall wing size . More importantly the CV for wing size in females ( across deletions ) for both backgrounds was ∼0 . 03 , which is 5× and 10× less than that observed for sdE3 hemizygotes in Samarkand and Oregon-R respectively ( Figure S6 ) . This suggests that most of the phenotypic variation for wing size due to the deletion is observed when the backgrounds are “sensitized” with the sd mutation , while having relatively little influence on wild-type wing size . Are the loci influencing the background-specific genetic interactions the same as those that modulate phenotypic expressivity for wing size of the focal sdE3 mutation ? To address this question we generated a set of backcross lines between Oregon-R and Samarkand ( both fixed for sdE3 ) , where “long” wings were selected in the backcross to the Oregon-R background , and “short” wings in backcrosses to the Samarkand background ( Figure S3 ) . Using ∼30 SNPs polymorphic across backgrounds , we verified that these backcross lineages showed expected genotypes for more than 90% of markers ( i . e . phenotypically short wings but with Samarkand genotypes ) . Among the molecular markers that did introgress , include those tightly linked to the unknown causal loci on 2R near cytological band 48 and at the centromere of 3L [36] . If the loci modulating the magnitude of the genetic interactions were caused by genes other than those influencing the background-specific disruption of wing development , we would predict weak correlations between sdE3/Y; Deletion/+ in Oregon-R and the equivalent genotype from the “short” backcross ( with an otherwise Samarkand background ) . Similar logic prevails for the Samarkand and the “long” phenotype . However , even using semi-quantitative measures , it is clear that these are highly correlated; 0 . 82 ( CI 0 . 66–0 . 91 ) and 0 . 86 ( CI 0 . 73–0 . 93 ) respectively . These results are consistent with the loci influencing the background-dependent genetic interactions being the same as those influencing the background-dependent effects on the phenotypic expressivity of the focal sdE3 mutation . The results described above demonstrate that the loci that influence the background dependent nature are linked to those influencing phenotypic expressivity of the mutation itself . However , it was unclear if the observations were due to some particular properties of the sdE3 allele , or a more general function of perturbation at the sd locus . To address this , we retested a subset ( 29 ) of the deletions spanning the range of phenotypic effects with sdE3 , using an additional allele sdETX4 , across each genetic background . The phenotypic consequences of sdETX4 , while background-dependent , are somewhat weaker than sdE3 ( Figure S7A ) . Despite these phenotypic differences , there was a moderate to high correlation across the modifiers' effects on these two alleles . In the Oregon-R and Samarkand wild-type genetic backgrounds respectively , the correlation between the effects of the deletions on the phenotypes of the sdE3 and sdETX4 allele was 0 . 66 ( CI 0 . 38–0 . 82 ) , and 0 . 76 ( CI 0 . 55–0 . 88 ) . In addition the general pattern of greater sensitivity to mutational perturbation by modifiers of the sd phenotype appears to be generally maintained ( Figure S7B ) . These results demonstrate that even across multiple alleles , the background dependence of the modifiers is maintained . Although the primary goal of this study was to explore the flexibility in genetic interactions , not to identify candidate genes , for confirmatory purposes , we examined several genomic regions that demonstrated background-dependent or -independent modifiers ( Table S2 ) . Interestingly , one region , 49E1 , contained vg , which encodes a SD-regulated transcriptional factor that forms a heterodimer with SD . Fine mapping , followed by the use of candidate insertional mutants ( co-isogenic to the Exelixis deletions ) confirmed that the vgF02736 allele behaved as a background-dependent enhancer with strong enhancement in Samarkand , but very weak enhancement in Oregon-R . We followed this up by introgressing this allele into both the Samarkand and Oregon-R background . Again we observed background-specific enhancement of the sd phenotype . Other fine mapping regions suggest several candidate genes , although for at least one region , no obvious candidate gene could be determined ( Table S2 ) .
There are outstanding questions that our study is unable to address . The background dependent nature of the genetic interactions could be the result of a “third-order” effect between the sd mutation , the hemizygous allele uncovered over the deletion and other loci across each wild-type genetic background . An alternative , and perhaps simpler explanation would be of differential quantitative complementation uncovered by the deletion [46] . In such cases , the variation in the degree of the modification of the focal mutation ( sd ) is a direct result of the alleles that differ across backgrounds uncovered by the deletion . While we expect that our results are a combination of both explanations , it is likely that without very high resolution mapping of the genomic regions , or test of specific polymorphisms will we be able to determine the relative contribution of each type of interaction . However the previous work that motivated this current study , namely the background dependent interaction between sd and Omb was clearly due to a third order effect [36] . Understanding the degree to which increasingly higher order epistasis contributes to phenotypic variation is under-explored but of great importance [47] . One curious finding of our study was that the background ( Oregon-R ) that demonstrated the higher degree of phenotypic expressivity of the focal sd mutations , showed increased sensitivity to mutational perturbation ( both enhancers and suppressors ) as well as greater phenotypic variation within strain . Recent work has demonstrated that loci can influence trait variability ( “noise” ) directly [48]–[50] , including naturally occurring variants in the Hsp90 gene of Drosophila [51] . Indeed even cell-to-cell variation , and variation in penetrance appears to have a complex genetic architecture [48] influenced by variability in gene expression [52] . It is unclear whether the loci that contribute to increased phenotypic “noise” also contribute to the amplified sensitivity to mutational perturbation as seen in the Oregon-R vs . Samarkand wild-type backgrounds . In previous work Oregon-R does have higher levels of phenotypic variation in quantitative measures of wing shape , but no increased sensitivity to weak ( heterozygous ) mutational perturbation [42] . However the focal mutations used in the current study ( sdE3 and sdETX4 ) represented more severe perturbations to wing development , so this may not provide an adequate comparison . Regardless , this remains an unanswered question , and a potential link between so-called variance controlling genes and sensitivity to perturbation would have important implications for the genetic architecture of canalization and robustness [5] , [53] . One constraint of the current study is that we utilized a hypomorph of moderate phenotypic effect , as opposed to a null allele . While a formal definition of functional epistasis ( sensu [54] ) requires the use of null alleles , most interaction screens utilize alleles of comparable ( hypomorphic ) effect to allow the recovery of both enhancers and suppressors . Nevertheless , previous work has demonstrated that null alleles can also show background-dependence effects in the primary effect of the mutation , including on development , growth and viability [1] , [2] , and our results demonstrate that these conditional effects are likely to be reflected in the genetic interactions between mutations as well . In addition we demonstrated that the quantitative effects we observed with the interaction between sdE3 and segmental deletions in each wild-type genetic background were correlated when observed across another ( weaker ) allele , sdETX4 , suggesting that such effects are not due to a particular allele . We also demonstrated that the effects of these interactions are tightly linked to the same genomic regions that contribute to the primary background-dependent phenotypic effects of the mutations . Thus for our system , the genetic variants influencing the phenotypic expressivity of the focal mutation appear to be the same as those modulating both the magnitude , and potentially the sign of genetic interactions between mutations . While the positive and negative implications for modifier ( and other genomic ) screens is clear , the potential flexibility of genetic networks given segregating variation in a population needs to also be considered . In particular an allele entering a population ( either as a new mutation , or as a result of introgression from another population or species ) may not have a “fixed” effect on fitness; instead the genetically contingent effects of the allele result in a distribution of phenotypic effects , including a possible change in sign ( i . e . from deleterious to beneficial ) .
The Oregon-R strain was originally obtained from the Bloomington stock center , while Samarkand was obtained from the lab of Dr . Trudy Mackay . For both strains , we further inbred them to near isogenicity , and tested via a panel of 30 polymorphic markers to confirm there was no contamination or residual heterozygosity . A combination of sequencing and PCR-based genotyping suggests that these two strains have an approximately 2% divergence from one another , and that all sequenced regions examined to date are a subset of variation from natural populations . The X-linked sdE3 mutant allele ( obtained from the Drosophila stock center , Bloomington IN ) , used in this study is caused by a P{w[E] ry[1t7 . 2] = wE} transposon located in the third intron of the sd gene [55] . This mutant allele was introgressed into two lab wild-type strains , Oregon-R and Samarkand , both marked with white ( w ) , by repeated backcrosses involving homozygous mutant female and the wild type male for over 20 generations [36] . These lines have been subjected to extensive genotyping to verify the extent of the introgression , and to avoid contamination . The sdETX4 and vgF02736 alleles were also obtained from the Bloomington stock center , and were introgressed for 20 generations into each wild-type strain . To validate the primary findings of this study , we repeated crosses , and quantified wing size for a subset of 44 deletions , spanning the direction and magnitude of effects ( background dependent-independent , suppressor-enhancer , as well as negative controls ) observed in the genome-wide screen . A single wing from each of 5 male flies ( w sdE3/Y; Deletion/+ ) was dissected and mounted in glycerol , for both backgrounds . For the isogenic wild-type control strain , 30 individuals were used from each background-specific set of crosses to better ascertain the degree of variability . Images of the wings were captured using an Olympus DP30BW camera mounted on an Olympus BW51 microscope . Six landmarks ( Figure S2 ) were digitized using tpsDIG software [58] and centroid size was used as a measure of wing size . The landmarks were specifically chosen as they could be discerned on all wings ( Figure S2 ) . To quantitatively verify the background-dependent effects of a given deletion on wing size ( Figure 4 ) the following model was used:where Y is the Centroid Size , B is the background and D is the deletion . The analysis was performed using the lm function in R ( V 2 . 12 ) and 95% confidence intervals were constructed using confint . Significance was determined by non-overlapping confidence intervals with controls . The quantitative measure of wing size used for this analysis , correlates well with the semi-quantitative method and results used for the initial screening ( r = 0 . 82 , CI:0 . 69–0 . 9 in Oregon-R , r = 0 . 78 , CI:0 . 63–0 . 87 in Samarkand ) . This suggests high repeatability of the initial screen , as well as the semi-quantitative measure of wing size . To ascertain whether there was a commensurate effect of the genomic deletions in “wild-type” wings ( as opposed to the mutant phenotype caused by sd mutants ) , we quantified wing size in females heterozygous for the focal sdE3 mutation with each deletion ( w sdE3/w sd+; Deletion/+ ) digitizing the same 6 landmarks on the wing . Potentially the genomic regions ( from the wild-type strains ) that influence the genetic interaction between the deletions and sdE3 could be independent of those regions that influence the variation for phenotypic expressivity of the sdE3 mutation itself . To test this we generated lines that had “high expressivity” sdE3 phenotypes in an otherwise “low expressivity” background ( Figure S3 ) . A backcross-selection procedure was used to introgress the modifiers that contribute to the “large wing” phenotype from the Samarkand background into the “small wing” background of Oregon-R and vice-versa ( Figure S3 ) . Upon generation of these lines , we repeated the dominant modifier screen as described above using a subset of 32 of the 44 confirmed modifiers and negative controls . These lines were used in identical crosses to those outlined above , with sdE3/Y; Deletion/+ individuals examined . To narrow down several genomic regions to a set of a few candidate genes we utilized an additional set of overlapping deletions in DrosDel , Exelixis and BSC strains followed by use of P-element insertional mutations co-isogenic with the Exelixis panel of lines . We utilized this approach for four genomic regions ( 49E1 , 57B3-B5 , 63F2-F7 , and 86E13-E16 ) detailed in Table S2 .
|
Examining the consequences of how one mutation behaves when in the presence of a second mutation forms the basis of our understanding of genetic interactions , and is part of the fundamental toolbox of genetic analysis . Yet the logical interpretation of such mutational interactions depends on the generality of such findings . A small number of studies have demonstrated that factors such as the wild-type background in which the mutations are studied can have a profound impact on the observed phenotype of both specific effects of the mutation and the interactions between mutations . However , whether such findings are a common property of genetic interactions was unknown . We tested the generality of the background dependence of interactions between mutations and observed that the vast majority of the interactions were highly dependent on the wild-type background in which they are observed . We demonstrate that the same regions of the genome that contribute to the differences observed in the degree of severity of the mutational effect appear to also be responsible for the background dependence of the interaction .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"epistasis",
"mutation",
"genetic",
"mutation",
"mutation",
"types",
"phenotypes",
"genetic",
"screens",
"heredity",
"genetics",
"genetic",
"suppression",
"population",
"genetics",
"biology",
"complex",
"traits",
"gene",
"function"
] |
2013
|
The Conditional Nature of Genetic Interactions: The Consequences of Wild-Type Backgrounds on Mutational Interactions in a Genome-Wide Modifier Screen
|
The discovery of small molecules targeted to specific oncogenic pathways has revolutionized anti-cancer therapy . However , such therapy often fails due to the evolution of acquired resistance . One long-standing question in clinical cancer research is the identification of optimum therapeutic administration strategies so that the risk of resistance is minimized . In this paper , we investigate optimal drug dosing schedules to prevent , or at least delay , the emergence of resistance . We design and analyze a stochastic mathematical model describing the evolutionary dynamics of a tumor cell population during therapy . We consider drug resistance emerging due to a single ( epi ) genetic alteration and calculate the probability of resistance arising during specific dosing strategies . We then optimize treatment protocols such that the risk of resistance is minimal while considering drug toxicity and side effects as constraints . Our methodology can be used to identify optimum drug administration schedules to avoid resistance conferred by one ( epi ) genetic alteration for any cancer and treatment type .
Alteration of the normal regulation of cell-cycle progression , division and death lies at the heart of the processes driving tumorigenesis . A detailed molecular understanding of these processes provides an opportunity to design targeted anti-cancer agents . The term ‘targeted therapy’ refers to drugs with a focused mechanism that specifically act on well-defined protein targets or biological pathways that , when altered by therapy , impair the abnormal proliferation of cancer cells . Examples of this type of therapy include hormonal-based therapies in breast and prostate cancer; small-molecule inhibitors of the EGFR pathway in lung , breast , and colorectal cancers – such as erlotinib ( Tarceva ) , gefitinib ( Iressa ) , and cetuximab ( Erbitux ) ; inhibitors of the JAK2 , FLT3 and BCR-ABL tyrosine kinases in leukemias – such as imatinib ( Gleevec ) , dasatinib ( Sprycel ) , and nilotinib ( Tasigna ) ; blockers of invasion and metastasis; anti-angiogenesis agents like bevacizumab ( Avastin ) ; proapoptotic drugs; and proteasome inhibitors such as bortezomib ( Velcade ) [1] , [2] . The target-driven approach to drug development contrasts with the conventional , more empirical approach used to develop cytotoxic chemotherapeutics , and the successes of the past few years illustrate the power of this concept . The absence of prolonged clinical responses in many cases , however , stresses the importance of continued basic studies into the mechanisms of targeted drugs and their failure in the clinic . Acquired drug resistance is an important reason for the failure of targeted therapies . Resistance emerges due to drug metabolism , drug export , and alteration of the drug target by mutation , deletion , or overexpression . Depending on the cancer type and its stage , the therapy administered , and the genetic background of the patient , one or several ( epi ) genetic alterations may be necessary to confer drug resistance to cells . In this paper , we investigate drug resistance emerging due to a single alteration . For example , treatment of chronic myeloid leukemia ( CML ) with the targeted agent imatinib fails due to acquired point mutations in the BCR-ABL kinase domain [3] . To date , ninety different point mutations have been identified , each of which is sufficient to confer resistance to imatinib [4] . The second-generation BCR-ABL inhibitors , dasatinib and nilotinib , can circumvent most mutations that confer resistance to imatinib; the T315I mutation , however , causes resistance to all BCR-ABL kinase inhibitors developed so far . Similarly , the T790M point mutation in the epidermal growth factor receptor ( EGFR ) confers resistance to the EGFR tyrosine kinase inhibitors gefitinib and erlotinib [5] , which are used to treat non-small cell lung cancer . Other mechanisms of resistance include gene amplification or overexpression of the P-glycoprotein family of membrane transporters ( e . g . , MDR1 , MRP , LRP ) which decreases intracellular drug accumulation , changes in cellular proteins involved in detoxification or activation of the drug , changes in molecules involved in DNA repair , activation of oncogenes such as Her-2/Neu , c-Myc , and Ras as well as inactivation of tumor suppressor genes like p53 [6]–[11] . The design of optimal drug administration schedules to minimize the risk of resistance represents an important issue in clinical cancer research . Currently , many targeted drugs are administered continuously at sufficiently low doses so that no drug holidays are necessary to limit the side effects . Alternatively , the drug may be administered at higher doses in short pulses followed by rest periods to allow for recovery from toxicity . Clinical studies evaluating the advantages of different approaches have been ambivalent . Some investigations found that a low-dose continuous strategy is more effective [12] , while others have advocated more concentrated dosages [13] . The effectiveness of a low-dose continuous approach is often attributed to its targeted effect on tumor endothelial cells and the prevention of angiogenesis rather than low rates of resistance [14] . The continuous dosing strategy is often implemented as combination therapy , sometimes including a second drug administered at a higher dose in a pulsed fashion . A significant amount of research effort has been devoted to developing mathematical models of tumor growth and response to chemotherapy . In a seminal paper , Norton and Simon proposed a model of kinetic ( not mutation-driven ) resistance to cell-cycle specific therapy in which tumor growth followed a Gompertzian law [15] . The authors used a differential equation model in which the rate of cell kill was proportional to the rate of growth for an unperturbed tumor of a given size . They suggested that one way of combating the slowing rate of tumor regression was to increase the intensity of treatment as the tumor became smaller , thus increasing the chance of cure . The authors also published a review of clinical trials employing dosing schedules related to their proposed dose-intensification strategy , and concluded that the concept of intensification was clinically feasible , and possibly efficacious [16] . Later , predictions of an extension of this model were validated in a clinical trial evaluating the effects of a dose-dense strategy and a conventional regimen for chemotherapy [17] . Their model and its predictions have become known as the Norton-Simon hypothesis and have generated substantial interest in mathematical modeling of chemotherapy and kinetic resistance . For example , Dibrov and colleagues formulated a kinetic cell-cycle model to describe cell synchronization by cycle phase-specific blockers [18]; this model was then used for optimizing treatment schedules to increase the degree of synchronization and thus the effectiveness of a cycle-specific drug . Agur introduced another model describing cell-cycle dynamics of tumor and host cells to investigate the effect of drug scheduling on responsiveness to chemotherapy [19]; this model was then used to optimize scheduling of chemotherapeutics to maximize efficacy while controlling host toxicity . Other theoretical studies include a mathematical model of tumor recurrence and metastasis during periodically pulsed chemotherapy [20] , a control theoretic approach to optimal dosing strategies [21] , and an evaluation of chemotherapeutic strategies in light of their anti-angiogenic effects [14] . For a more comprehensive survey of kinetic models of tumor response to chemotherapy , we refer the reader to reviews [22]–[24] and references therein . There have also been substantial research efforts devoted to developing mathematical models of genetic resistance , i . e . resistance driven by genetic alterations in cancer cells . Since mutations conferring resistance can arise as random events during the DNA replication phase of cell division , the dynamics of resistant populations are well-suited to description with stochastic mathematical models . Coldman and co-authors pioneered this field by introducing stochastic models of resistance against chemotherapy to guide treatment schedules ( see e . g . , [25] , [26] and references therein ) . In 1986 , Coldman and Goldie studied the emergence of resistance to one or two functionally equivalent chemotherapeutic drugs using a branching process model of tumor growth with a differentiation hierarchy [25] . In this model , the birth and death rates of cells were time-independent constants and each sensitive cell division gave rise to a resistant cell with a certain probability . The effect of drug was modeled as an additional probabilistic cell kill law on the existing population , and the drug could be administered in a fixed dose at a series of fixed time points . The goal of the model was to schedule the sequential administration of both drugs in order to maximize the probability of cure . Later , the assumption of equivalence or ‘symmetry’ between the two drugs was relaxed [27] . These models were also extended to include the toxic effects of chemotherapy on normal tissue , and an optimal control problem was formulated to maximize the probability of tumor cure without toxicity [26] . More recently , Iwasa and colleagues used a multi-type birth-death process model to study the probability of resistance emerging due to one or multiple mutations in populations under selection pressure [28] . The authors considered pre-existing resistance mutations and determined the optimum intervention strategy utilizing multiple drugs . Multi-drug resistance was also investigated using a multi-type birth-death process model in work by Komarova and Wodarz [29] , [30] . In their models , the resistance to each drug was conferred by genetic alterations within a mutational network . The birth and death rates of each cell type were time-independent constants and cells had an additional drug-induced death rate if they were sensitive to one or more of the drugs . The authors studied the evolution of resistant cells both before and after the start of treatment , and calculated the probability of treatment success under continuous treatment scenarios with a variable number of drugs . Recently , the dynamics of resistance emerging due to one or two genetic alterations in a clonally expanding population of sensitive cells prior to the start of therapy were studied using a time-homogenous multi-type birth-death process [31] , [32] . One common feature of these models of genetic resistance is that the treatment effect is formulated as an additional probabilistic cell death rate on sensitive cells , separate from the underlying birth and death process model with constant birth and death rates . Under these model assumptions , the drug cannot alter the proliferation rate of either sensitive or resistant cells; however , a main effect of many targeted therapies ( e . g . imatinib , erlotinib , gefinitib ) is the inhibition of proliferation of cancer cells . Inhibited proliferation in turn leads to a reduced probability of resistance since resistant cells are generated during sensitive cell divisions . In this paper , we utilize a non-homogenous multi-type birth-death process model wherein the birth and death rates of both sensitive and resistant cells are dependent on a temporally varying drug concentration profile . This study represents a significant departure from existing models of resistance since we incorporate the effect of inhibition of sensitive cell proliferation as well as drug-induced death , obtaining a more accurate description of the evolutionary dynamics of the system . In addition , we generalize our model to incorporate partial resistance , so that the drug may also have an effect on the birth and death rates of resistant cells . The goals of our analysis also differ from those of previous work . Coldman and Murray were interested in finding the optimal administration strategy for multiple chemotherapeutic drugs in combination or sequential administration [26]; they aimed to maximize the probability of cure while limiting toxicity . Komarova was interested in studying the effect of multiple drugs administered continuously on the probability of eventual cure [30] . In contrast , in this paper we derive estimates for the expected size of the resistant cell population as well as the probability of resistance during a full spectrum of continuous and pulsed treatment schedules with one targeted drug . We then propose a methodology for selecting the optimal strategy from this spectrum to minimize the probability of resistance as well as to maximally delay the progression of disease by controlling the expected size of the resistant population , while incorporating toxicity constraints . In many clinical scenarios , the probability of resistance is high regardless of dosing strategy , and thus the maximal delay of disease progression is a more realistic objective than tumor cure . The methodology developed in this paper can be applied to study acquired resistance in any cancer and treatment type .
Let us now calculate the probability of resistance during a given dosing schedule . Under the assumption of complete resistance , the probability of extinction of a resistant cell clone starting from one resistant cell is , regardless of which dosing strategy is used [33] . The number of resistant cells produced from the sensitive cell population is on average proportional to the number of sensitive cell divisions; these two quantities are related through the mutation rate . As a preliminary calculation , consider the behavior of the sensitive cell population , , with constant growth rate and death rate . Under the assumption that the mutation rate is small enough such that the stochastic emergence of resistant cells from sensitive cells has negligible effects on the sensitive cell number , we approximate as a simple birth-death process . Recall that the initial size of the sensitive cell population is ; then the mean abundance of the sensitive cell population at time is approximated as . The number of sensitive cell divisions in the time interval is approximately given by ( 1 ) The number of surviving resistant cell clones arising from the sensitive cell population in the time interval is then binomially distributed as . Let us now study the probability of resistance under a general pulsed treatment regimen . Define as the expected number of sensitive cancer cells at the beginning of the treatment cycle , and as the expected number of sensitive cancer cells at the beginning of the treatment holiday . Then we have , , , etc . We obtain the general formulae for the number of sensitive cancer cells at the beginning and end of the treatment cycle as ( 2 ) where . The number of cycles before extinction of the sensitive cell population is approximated as . To estimate the total number of sensitive cancer cell births before extinction , we sum the number of births during the on- and off-treatment phases over all cycles . Let be the number of sensitive cancer cell births during the on-treatment phase and the number of births during the off-treatment phase . The expected number of births , , during the entire treatment regime is then approximated as ( 3 ) Here is the estimated number of sensitive cancer cell divisions in the treatment interval , evaluated as in equation ( 1 ) . Next , let us define the functions and . We can express each sum as the geometric series ( 4 ) Then we obtain the expected number of sensitive cancer cell births during the entire duration of therapy as . We can approximate with . Substituting in the correct expressions for and , we obtain a final estimate for the number of sensitive cell divisions before the extinction of sensitive cells as The number of surviving resistant cell clones produced from the sensitive cell population is a random variable with distribution . We can thus make a Poisson approximation to estimate the probability that at least one surviving resistant cell clone is produced before the extinction of sensitive cells as ( 5 ) In the special case of continuous dosing , , the number of sensitive cell divisions is approximated by ( 6 ) As a consistency check , this formula can also be arrived at via equation ( 1 ) with the initial size of the sensitive cancer cell population , , and the amount of time until the extinction of the sensitive cells , . Then the probability of resistance emerging during continuous dosing is again calculated using formula ( 5 ) with equation ( 6 ) . When the ( epi ) genetic alteration confers partial resistance to the cell ( i . e . when and/or ) , then the probability of resistance emerging during continuous dosing is given by ( 7 ) where is again calculated as in equation ( 6 ) . To accommodate this modification in pulsed schedules , we introduce ‘effective’ growth and death rates for resistant cells . The effective growth rate of resistant cells is given by , while the effective death rate of resistant cells is . For general pulsed schedules , the probability of resistance is then approximated by ( 8 ) where is calculated as in equation ( 5 ) . We next approximate the expected number of resistant cancer cells at time . To calculate this quantity , we estimate the number of surviving resistant cell clones produced during each small time interval and then calculate the growth of each resistant cell clone until time . More precisely , we take the convolution of the rate of production of resistant cell clones from the sensitive cell process with the average rate of clonal expansion of resistant cells . Let us first consider general pulsed treatment schedules . Using methods from the previous section , we find that the expected number of sensitive cancer cell divisions until time is given byHere denotes the fractional number of treatment cycles until time . After making the approximation , we have ( 9 ) Since the number of resistant cells produced directly from the sensitive cell population until time is binomially distributed , , the expected number of such cells is given by . Thus we estimate the average number of resistant cells at time as ( 10 ) In the special case of continuous dosing strategies , , the average number of resistant cells at time is given by ( 11 ) Once again we can check for consistency by deriving this formula via equation ( 1 ) . Recall that the expected number of sensitive cancer cell births starting from a population of size until time is given by ( equation ( 1 ) ) . Then the expected number of resistant cells produced is , and the expected number of resistant cells at time is given by We are also interested in calculating the expected number of resistant cells averaged only over those patients who develop resistance . This quantity is clinically relevant since many treatment choices may inevitably lead to resistance; in those cases , the drug should be dosed in such a way that the number of resistant cells is minimum , thereby maximizing the time until detection of resistance and disease progression . Mathematically , this amounts to estimating the expected size of the resistant cell population , conditioned on the event that at least one surviving resistant clone is produced prior to the extinction of sensitive cancer cells . We make the approximation that the expected resistant cell number , conditioned on the complementary event of no surviving resistant cell clones , is negligible . Then the expected number of resistant cells averaged over the cohort of patients who develop resistance is estimated as ( 12 ) Suppose that at the start of therapy there exists a small population of resistant cells . We may then adapt the theory to calculate the probability of resistance and expected size of the resistant clone under various dosing schedules . Let us consider the initial population as two separate populations: sensitive cells and resistant cells , where is the initial fraction of resistant cells ( assume for simplicity that is an integer ) . Then the probability of avoiding resistance is given by the probability that the pre-existing resistant cell clones become extinct times the probability that the initial sensitive cell population does not give rise to any surviving resistant clones during treatment . Let denote the probability , calculated as in equation ( 5 ) , of de novo resistance arising from the initial sensitive population of size . The probability of extinction of the pre-existing clone is given by if and otherwise . ( Note that and may be replaced by and in the case of pulsed schedules with partial resistance ) . Then the total probability of resistance is given by ( 13 ) Let represent the expected number of resistant cells arising from the initial sensitive cell population of size , calculated as in equation ( 10 ) . The expected number of resistant cells at time is given by plus the expected current size of the initially resistant population . Thus we have ( 14 ) where once again the rates and may be replaced by their effective values in the case of pulsed therapy with partial resistance .
Using the estimates derived above , we now propose a method for optimizing dosing strategies to minimize the probability of resistance . In cases where the emergence of resistance is certain , this method will predict a dosing strategy that maximally delays the detection of resistance by minimizing the number of resistant cancer cells . The optimal dosing strategy is selected from a range of tolerated treatment schedules specified by toxicity constraints . In practice , these toxicity constraints , in addition to the growth and death rates of sensitive and resistant cells at varying dose levels , must be determined experimentally for each drug and cancer type . In the following we will construct example toxicity constraints to demonstrate the methodology and test for sensitivity to the constraint profile . A modification of treatment schedules can change the duration of each treatment pulse ( affecting and ) , the intensity of the dose ( affecting growth and death rates of sensitive cancer cells , and ) , or both . When considering complete resistance , the growth and death rates of resistant cells are unaltered by changing treatment strategies . We assume that all other parameters are unaffected by changes in administration schedules as well . Thus , we consider toxicity constraints to provide a bounded domain in the four-dimensional parameter space spanning , and . We can immediately reduce the dimension of the constraint domain to two , since specifies explicitly through the fixed length of the treatment-and-break cycle , , and and are both dependent on the concentration of the drug and thus cannot vary independently . Therefore , we consider toxicity constraints in the form of a function specifying the maximum amount of time , , that a drug can be administered to a patient at a particular concentration before causing dose-limiting toxicities . In the following , we make the simplifying assumption that this drug concentration specifies the death rate of sensitive cancer cells , , and does not alter the growth rate , ; alternatively , we can also investigate treatment strategies that modulate the growth rate rather than the death rate of sensitive cancer cells , or both . We assume that such relationships between and are monotonically decreasing functions of ; see Figure 4A for an example of a toxicity constraint . From clinical trial data we obtain the maximum amount of time for which a range of drug concentrations are tolerated , leading to a relationship between and the drug concentration . The effect of particular drug concentrations on and/or may then be found experimentally by exposing sensitive cancer cells to drug doses and measuring the growth and death rates . Such investigations identify a toxicity constraint relating and . We display example constraint functions in Figures 4B and 4C . We next show some example toxicity data for the targeted drug erlotinib , which is an EGFR inhibitor used in treating solid malignancies such as non-small cell lung cancer . Compiling data from several clinical trials [34]–[36] , we obtain a relationship between the drug dose and plasma concentration ( measured as the maximum concentration achieved after a single dose ) . This data is plotted in Figure 5A; here we observe a relatively linear relationship between dose and plasma concentration . We also compiled data points on the number of days each particular dose was tolerated in continuous daily administration . We converted each dose level to concentration using the linear relationship found , and plot these points in Figure 5B . A conservative toxicity constraint in terms of vs . concentration is plotted , where we assume that any concentration or length of pulse increased beyond what was tolerated in the trials would not be admissible . This toxicity constraint , in conjunction with further experimental data on the growth and death rates of sensitive and resistant cancer cells at various concentrations , would enable us to calculate optimal dosing schedules for this specific system using our model . For our theoretical investigations , we now introduce several example families of toxicity constraints to test for sensitivity of the probability of resistance to several key aspects of the shape of the curve . All of these example constraints are convex , monotonically decreasing functions of . Thus we have implicitly assumed that as the drug concentration increases and the cell death rate increases , the maximum tolerated length of a treatment pulse decreases . In the first family , we vary the maximum dose that can be tolerated for the full treatment schedule of days . In the second family we vary the maximum dose that can be tolerated for just one day , and in the third family we vary the degree of convexity of the constraint curve , or the initial rate of decrease in as the concentration increases . Consider the first family of toxicity curves in Figure 6A , specified by ( 15 ) where for and . In our notation , , the subscript denotes the constraint family and the superscript indicates a specific function belonging to this family . These constraints serve to vary the endpoint representing the maximum dose that can be tolerated for a full treatment cycle ( days ) while fixing the endpoint representing the maximum dose that is tolerated for just one day of a treatment cycle , specified by the death rate . In other words , in this family of constraints we vary the continuous-dose concentration endpoint ( represented by black circles in Figure 6A ) of the toxicity constraint via the parameter , while keeping the form of the constraint and the high-dose concentration endpoint fixed . We also test for sensitivity to two other aspects of the toxicity constraints: the high-dose concentration endpoint ( i . e . the maximum dose that is tolerated for just one day ) and the degree of concavity of the curve . Figure 6B shows a family of constraints varying the high-dose endpoints ( shown in black circles ) . These example constraints are specified by equation ( 16 ) where for and . Likewise , a family of constraints varying the degree of convexity is exhibited in Figure 6C and specified by the following equations: ( 17 ) where , for each function , the and are determined by setting the endpoints to be ( 18 ) Once the toxicity constraint is established ( e . g . Figure 4 ) , the tolerable range of treatment schedules is specified by the area under the curve on the . We then aim to locate the optimal point within this area that minimizes the probability of resistance . In situations in which the optimum probability of resistance is 1 or close to 1 , we aim to locate the optimal point minimizing the expected number of resistant cells conditional on developing resistance , thus maximizing the time until disease progression . We note from the analytical approximations that a change in the mutation rate does not modify the choice of optimal dosing schedule . We also observe from our analytical approximations that the optimizing points must always lie directly on the toxicity constraint curve itself – intuitively , any point lying below the toxicity curve represents a weaker than tolerated dosing schedule and hence cannot minimize the risk of resistance . Once a minimizing point is located , the optimal treatment schedule is entirely specified since the duration of treatment pulses are given by , the length of the drug holiday is given as the remainder of the cycle duration , and the intensity of the dose is specified by the death rate of sensitive cells , . To illustrate this concept , let us consider the toxicity constraint from equation ( 15 ) . Recall the constraint restricting the treatment schedules to viable dosing strategies in which the population of sensitive cells decreases overall in time . This constraint may restrict the domain of the toxicity curve to a limited range of . For the current example , this restriction is shown in Figure 5 . We can then calculate the probability of resistance , the expected number of resistant cells , and the conditional expected number of resistant cells over the range of treatment schedules specified by this restricted constraint curve . Note again that in the formula describing the expected number of resistant cells , equation ( 10 ) , the growth rate of is dominated by the growth rate of resistant cells , , at later times , since the other time-dependent term in the expression , , approaches zero as increases . Thus we can neglect the latter term when considering the long-term growth of the resistant cell population . Rewrite equation ( 10 ) as . Here is the time-independent constant comprised of the remaining terms in equation ( 10 ) except for . Analogously , the expected number of resistant cells conditional to the emergence of resistance is approximated by . In Figures 7A–C , we show the probability of resistance , , the time-independent term of the equation describing the number of resistant cells , , and over the range of treatment schedules specified by the restricted constraint curve from Figure 6D . As the drug concentration and hence the death rate of sensitive cancer cells , , increase , we move along the constraint curve from the continuous-dose endpoint towards the high-dose endpoint . As a particular numerical example , consider an initial number of sensitive cancer cells of , a mutation rate conferring resistance of , and a neutral resistance mutation ( ) . Then the probability of resistance , shown in Figure 7A , is minimized when . This result is subsequently used to identify the corresponding optimal treatment schedule in Figure 6D , which in this case is given by days , days , and a drug concentration achieving . When this optimal treatment schedule is used , the probability of resistance is below 10% . However , if a higher dose is chosen , the probability of resistance may increase up to . This example illustrates the importance of locating the optimal dosing regime for the clinical management of patients . The values proportional to the expected number of resistant cells , , and to the conditional expected number of resistant cells , , are displayed in Figures 7B and 7C . Interestingly , in the event that resistance occurs , the optimal treatment schedule for minimizing the resistant cell population is specified by , which differs from the optimal schedule for minimizing the probability of developing resistance . For a general cohort of patients treated with this dosing schedule , the probability of developing resistance would be close to 1; for the subset of patients who do develop resistance , however , this dosing schedule would delay disease progression by the largest amount of time . Let us now examine the dependence of these optimal dosing regimens on variations in parameters and toxicity constraints . Specifically , we investigate the sensitivity of the optimal dosing strategies to several characteristics of the toxicity curves: the maximum dose that can be administered for the whole treatment cycle of days ( the continuous-dose endpoint ) , the maximum dose that can be administered for one day only ( the high-dose endpoint ) , and the degree of concavity of the toxicity curve . The optimal dosing regimens are identified over a range of parameter values of and . First , we consider the family of curves for ( equations ( 15 ) and ( 14 ) as shown in Figure 6A ) . The optimal dosing strategy minimizing the probability of resistance and/or the conditional number of resistant cells is displayed in Figure 8 . In column ( A ) , we show the value of that corresponds to the dosing schedule which minimizes the probability of resistance for a given and . The corresponding minimal probability of resistance is shown in column ( C ) . Column ( B ) displays the value of that specifies the dosing schedule minimizing the conditional expected number of resistant cells , i . e . maximizing the amount of time until disease progression in patients who develop resistance . The rows show the results for constraints , and , respectively . Note that the optimal dosing schedules in the first and second column are not identical , reflecting the fact that the recommended dosing regimens for these two clinical goals are different . In addition , we observe that as the continuous-dose endpoint is varied , the minimal probability of resistance changes ( in column ( C ) ) while the optimal dosing schedules remain relatively unchanged . In particular , the minimal probability of resistance decreases as the continuous-dose endpoint shifts to the right . Next we consider the family of curves , for ( equations ( 15 ) and ( 16 ) , shown in Figure 6B ) . We plot the results in Figure 9 , where the columns show the optimal treatment schedules and the probability of resistance for constraints , and , respectively . For both clinical goals of minimizing the probability of resistance and maximizing the time until detection of resistance , we observe that as the maximum dose tolerated for one day ( the high-dose endpoint ) is increased , the optimal dosing schedule shifts slightly to a more high-dose pulsed regimen in some regions of the parameter space ( particularly when is small ) . However , the minimal probability of resistance changes only slightly as this endpoint is increased . Lastly , we consider the family of curves , for ( equation ( 17 ) , shown in Figure 6C ) . We plot the results in Figure 10 . The columns again show the optimal schedules and the probability of resistance for constraints , and . The results for the first two constraints , and , differ markedly from those of . In particular , for functions with a lower degree of convexity , a high-dose pulsed treatment is optimal for both clinical goals . For these cases a minimal probability of resistance near zero can be achieved . However , for the optimal dosing schedule shifts more towards the continuous end of the dosing spectrum , and in certain parameter ranges the minimal probability of resistance reaches higher values . So far we have only considered treatment strategies during which the total number of sensitive cancer cells declines on average , i . e . when holds . However , for some therapies and cancer types it is impossible to reduce the number of sensitive cancer cells . Then the goal of therapy becomes to slow or even halt the rate of tumor growth . For these cases , the probability of resistance is always one . However , we can still identify treatment schedules that maximally delay progression of disease by controlling the number of resistant cells . The approximations for the expected number of resistant cells derived above remain valid , except when . In this case , we revisit the calculation of and estimate the total number of births during on- and off-treatment phases as ( 19 ) Once again making the approximation , we obtainAfter taking the convolution of the derivative with the expected growth rate of resistant cells , we obtain the expected number of resistant cells at time asNote that for cases when , the formula for the expected number of resistant cells , equation ( 10 ) , experiences a singularity in the denominator when , i . e . when the net growth rate of the resistant cancer cells equals the net growth rate of the sensitive cancer cells . However , the range of therapies considered should be restricted to those in which the net growth rate of sensitive cancer cells is less than that of resistant cancer cells; otherwise , the problem of resistance is secondary to the problem of controlling the sensitive cell population . In these cases , the singularity does not occur .
In this paper , we have constructed a simple mathematical model using birth and death processes to describe the evolution of resistance during targeted anti-cancer therapy . We have derived and validated analytical approximations to this model , which provide a useful tool for predicting the risk of resistance and the growth of resistant cell populations under various dosing strategies . We have used our model and estimates to develop a methodology for designing optimal drug administration strategies to minimize the risk of resistance . In cases in which the risk of resistance is high for any treatment schedule , these strategies are modified to maximize the time until the progression of disease . The probability of resistance is shown to be largely dependent on , the rate of sensitive cell division , which is the product of the current sensitive cell population size and its growth rate . Drugs whose main goal is to increase the death rate of sensitive cells can decrease the sensitive cell population , thus decreasing and reducing the probability of resistance; however , if the initial tumor size is large , it may take a significant amount of time to deplete the sensitive cell population . During this delay , there is still a high probability of generating resistant mutants since the sensitive cell proliferation rate is unchanged . On the other hand , for drugs that inhibit sensitive cell proliferation and effectively reduce the growth rate of sensitive cells , the quantity is immediately reduced to zero regardless of the initial size of the tumor . This implies that drugs that inhibit cancer cell proliferation could be promising for the prevention of resistance in the absence of pre-existing resistant cell clones . Combination therapies in which an inhibitor of sensitive cell proliferation is dosed continuously while short , high pulses of a drug that increases the death rate of resistant cells are administered may also be of interest , as are any combination strategies which separately target the sensitive and resistant populations . We have also extended the theory to incorporate pre-existing resistant cells at the start of therapy . The effect of pre-existing resistant clones on the optimal dosing strategy is highly dependent upon system parameters including the growth and death rates of sensitive and resistant cells , the initial tumor size , and initial number of resistant cells . Consider the probability of resistance in this scenario , given by equation ( 13 ) . We note that the term , denoting the probability of extinction of the pre-existing clone , consists of the power of a quantity usually less than one . Thus even a small population of pre-existing resistant cells can cause the total probability of resistance to be effectively equal to one . For example , if the growth rate of resistant cells is twice their death rate , then the probability of extinction for an initial population consisting of only resistant cells evaluates to . Then the total probability of resistance , given by equation ( 13 ) , is approximately one . Therefore , the presence of even a small number of resistant cells at the start of therapy can effectively prevent a cure . In these cases , we may instead attempt to delay disease progression by controlling the number of resistant cells . Equation ( 14 ) describes the current size of the resistant population as the sum of the average de novo and pre-existing resistant clone sizes . Observe that both terms in this expression grow at the same exponential rate; the term for pre-existing resistance starts at time zero with the value , while the term for de novo resistance starts with value zero at time zero . This fact has implications for treatment schedules in the case of pre-existing resistance: as long as an eventual decline of sensitive cancer cells is achieved , high-dose strategies which slow the effective net growth rate of resistant cells may be more effective than low-dose strategies aimed at maximal continuous inhibition of sensitive cells . By testing several families of toxicity constraints , we have observed that the optimal dosing strategies are strongly affected by the degree of convexity of the toxicity curve , thus delineating a clear priority in experimental efforts to determine the parameters of this constraint . In our experience of studying published results of Phase I clinical trials of molecularly targeted anti-cancer therapies , patient toxicity reports are usually not detailed enough to accurately determine toxicity curves . In light of our observations , we would like to stress the importance of publishing detailed quantitative data on toxicity in clinical trials , so that statistical analyses can be performed to inform these constraint curves . It is also important to estimate the growth and death rates of sensitive and resistant cancer cells during administration of diverse drug concentrations . These curves can be estimated by studying the growth and death kinetics of cancer cells , either in vivo or in vitro . For example , in vitro net growth rates can be determined by subjecting sensitive and resistant cell populations to drug at varying concentrations and counting viable cells at multiple time points . Then , through fluorescence-activated cell sorting techniques , the amount of cell death at multiple time points can be observed , providing the cell death rate at each drug concentration . If the parameters of the model are also estimated for treatment with conventional cytotoxic chemotherapeutics , then our model can be applied to these treatment choices as well . This methodology , together with key parameters derived experimentally , can aid in the design of optimum administration strategies of treatment options for all cancer types that evolve resistance via a single ( epi ) genetic alteration .
|
Recently , the field of anti-cancer therapy has witnessed a revolution by the discovery of targeted therapy , which refers to compounds targeting specific pathways causing abnormal growth of cancer cells . The clinical success of such drugs has been limited by the evolution of acquired resistance to these compounds , which leads to a relapse after initial response to therapy . Current dosing procedures are not designed to optimally delay the emergence of resistance; the identification of such optimal dosing schedules represents an important challenge in clinical cancer research . Here , we design a novel methodology to identify the optimum drug administration strategies that reach this clinical goal . Our model describes the evolutionary dynamics of a tumor cell population during therapy . We consider drug resistance emerging due to a single ( epi ) genetic alteration and calculate the probability of resistance arising during specific dosing strategies . We then optimize treatment protocols such that the risk of resistance is minimal while considering drug toxicity and side effects as constraints . Since this methodology can be extended to describe situations arising during administration of cytotoxic chemotherapy as well , it can be used to identify optimum drug administration schedules to avoid resistance for any cancer and treatment type .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"computational",
"biology/evolutionary",
"modeling",
"mathematics",
"oncology"
] |
2009
|
Evolution of Resistance to Targeted Anti-Cancer Therapies during Continuous and Pulsed Administration Strategies
|
Cerebrospinal fluid ( CSF ) 42 amino acid species of amyloid beta ( Aβ42 ) and tau levels are strongly correlated with the presence of Alzheimer's disease ( AD ) neuropathology including amyloid plaques and neurodegeneration and have been successfully used as endophenotypes for genetic studies of AD . Additional CSF analytes may also serve as useful endophenotypes that capture other aspects of AD pathophysiology . Here we have conducted a genome-wide association study of CSF levels of 59 AD-related analytes . All analytes were measured using the Rules Based Medicine Human DiscoveryMAP Panel , which includes analytes relevant to several disease-related processes . Data from two independently collected and measured datasets , the Knight Alzheimer's Disease Research Center ( ADRC ) and Alzheimer's Disease Neuroimaging Initiative ( ADNI ) , were analyzed separately , and combined results were obtained using meta-analysis . We identified genetic associations with CSF levels of 5 proteins ( Angiotensin-converting enzyme ( ACE ) , Chemokine ( C-C motif ) ligand 2 ( CCL2 ) , Chemokine ( C-C motif ) ligand 4 ( CCL4 ) , Interleukin 6 receptor ( IL6R ) and Matrix metalloproteinase-3 ( MMP3 ) ) with study-wide significant p-values ( p<1 . 46×10−10 ) and significant , consistent evidence for association in both the Knight ADRC and the ADNI samples . These proteins are involved in amyloid processing and pro-inflammatory signaling . SNPs associated with ACE , IL6R and MMP3 protein levels are located within the coding regions of the corresponding structural gene . The SNPs associated with CSF levels of CCL4 and CCL2 are located in known chemokine binding proteins . The genetic associations reported here are novel and suggest mechanisms for genetic control of CSF and plasma levels of these disease-related proteins . Significant SNPs in ACE and MMP3 also showed association with AD risk . Our findings suggest that these proteins/pathways may be valuable therapeutic targets for AD . Robust associations in cognitively normal individuals suggest that these SNPs also influence regulation of these proteins more generally and may therefore be relevant to other diseases .
Cerebrospinal fluid ( CSF ) contains promising biomarkers for neurological and psychiatric diseases such as Alzheimer's disease ( AD ) , schizophrenia , and Parkinson's disease [1]–[4] . The brain directly and rapidly influences the composition of CSF , and as such , CSF analytes may provide insights into neurological and psychiatric disease pathways that may not be identifiable using blood or other biological fluids . The use of endophenotypes in genome-wide association studies ( GWAS ) has provided novel insights into pathways and proteins that are associated with AD onset and progression [5] , [6] . We have demonstrated the utility of CSF amyloid-beta ( Aβ ) , apolipoprotein E ( ApoE ) and tau levels as endophenotypes for genetic studies of AD [7]–[16] . In our most recent work we used nearly 1 , 300 samples to conduct a GWAS with CSF tau levels [9] . In that study , Cruchaga et al . identified three genome-wide significant loci , including rs9877502 , which also showed a consistent association with AD risk , tangle pathology , and global cognitive decline in independent datasets . The success of these and other similar efforts has led to broader efforts to develop and leverage datasets of this type [17]–[27] . While amyloid plaques and neurofibrillary tangles are the primary pathological features of AD , genetic , clinical , and animal studies demonstrate that endocytosis , cholesterol metabolism , and inflammatory and immune responses also play an important roles in AD pathogenesis [28] . To further leverage the advantages of the endophenotype based approach , we have sought to use analytes related to these other aspects of AD pathology . For this work we have obtained data from the Rules Based Medicine , Inc . ( RBM ) ( Austin , TX ) Human DiscoveryMAP Panel . This panel includes over 175 analytes selected from the constellation of known cytokines , chemokines , metabolic markers , hormones , growth factors , tissue remodeling proteins , angiogenesis markers , acute phase reactants , cancer markers , kidney damage markers , and central nervous system biomarkers . Analytes were quantitatively measured in CSF samples from 574 samples ( including both cognitively normal and demented individuals ) from two independent datasets to identify novel phenotypes that may contribute to the pathogenesis of AD . After careful quality control and evaluation of the literature we selected 59 AD-related analytes for analysis in genome-wide association studies . We identified genome-wide significant associations between putatively functional SNPs and five phenotypes ( Angiotensin-converting enzyme ( ACE ) , Chemokine ( C-C motif ) ligand 2 ( CCL2 ) , Chemokine ( C-C motif ) ligand 4 ( CCL4 ) , Interleukin 6 receptor ( IL6R ) and Matrix metalloproteinase-3 ( MMP3 ) ) . The genetic basis of variance in these important disease-related analytes may provide insights into the mechanisms contributing to AD and other human diseases .
We observed a significant association between CSF MMP3 and CCL2 and gender , where both analytes were lower in females relative to males in cognitively normal samples from both ADNI and the Knight ADRC . We also observed significantly increased MMP3 and CCL2 levels with increasing age in cognitively normal samples from both ADNI and the Knight ADRC . We failed to detect consistent association in both the ADNI and Knight ADRC samples with plasma levels of these analytes and age or gender . Full results including slopes of the regression models can be found in table S2 ( CSF results ) and table S3 ( plasma results ) .
Angiotensin converting enzyme ( ACE ) is encoded by the ACE gene ( 17q23 . 3 ) and has been previously implicated in AD pathogenesis . In vitro , ACE inhibits Aβ aggregation in an activity-dependent manner by slowing the rate of fibril formation [38]–[40] . ACE may inhibit Aβ aggregation by converting the highly amyloidogenic Aβ42 peptide into the more stable Aβ40 peptide [41] . In vivo , inhibition of ACE activity in an AD mouse model promotes Aβ42 deposition in the hippocampus [41] . Studies in mouse and human brain homogenate demonstrate that ACE causes Aβ degradation in a two-step process . First , ACE cleaves Aβ42 into Aβ40 and then Aβ40 undergoes degradation [41] . ACE activity is elevated in the CSF of AD patients [42] . Neuroblastoma cells exposed to synthetic Aβ42 oligomers , but not monomeric Aβ42 , produce elevated ACE protein levels and ACE activity , suggesting that Aβ aggregates may stimulate the up-regulation of ACE in AD brains as a mechanism of combatting the accumulation of these protein aggregates . Our findings suggest that SNPs within the ACE locus alter CSF and plasma ACE levels . Several SNPs in the ACE gene region have significant associations with increased levels of ACE . Conditional analyses suggest a single signal in this region is responsible for the association . While we are unable to determine the specific functional allele in this region , rs4343 and rs4316 are in nearly perfect LD with each other ( D′ = 0 . 99 , r2 = 0 . 91 ) and both are inferred to have regulatory function ( table 3 ) . Rs4316 has not been studied in AD or reported to be associated with other phenotypes to date . Rs4343 is a widely studied marker in ACE , and we found a significant association of this SNP with increased CSF ACE levels . The minor allele “G” of rs4343 has been previously reported to be significantly associated with increased plasma ACE activity levels and plasma ACE protein levels [21] , [29] . Our observation of increased CSF ACE levels with the G allele of rs4343 is consistent with these reports . In relation to AD , the findings with rs4343 have been varied with some studies suggesting association with CSF Aβ42 levels and AD risk while others do not [14] , [43]–[49] . Recent data from 1269 samples with CSF and genetic data failed to detect association between CSF Aβ42 , Tau or pTau181 levels and these SNPs in ACE ( table 3 ) [9] . Results from the recent International Genomics of Alzheimer's Project provide evidence of association between the SNPs in ACE we report to be associated with higher CSF ACE levels and reduced risk for AD [30] . In addition , recent work using 600 samples with CSF ACE measurements found a significant increase in ACE levels with increasing ptau/Aβ42 ratio ( which was used as a predictor of AD status ) [50] , further reinforcing the relationship between ACE levels and AD status . Our results demonstrate in human subjects that SNPs in the ACE gene are associated with elevated CSF and plasma ACE protein levels and that these SNPs are also associated with reduced AD risk . Taken with in vitro and in vivo studies that demonstrate that ACE cleaves and clears Aβ in an activity-dependent manner , these findings suggest that individuals carrying polymorphisms that increase ACE protein , and possibly ACE activity , may be better able to clear accumulating Aβ aggregates and are thus at reduced risk for developing AD . Matrix metalloproteinase 3 ( encoded by MMP3; located on 11q22 . 3 ) is hypothesized to contribute to endogenous , physiologic clearance of amyloid plaques . MMP3 is expressed in neurons , astrocytes , microglia and vascular cells [51] . MMP3 is preferentially localized in senile plaques in the parietal cortex of AD brains , while hippocampal plaques are relatively spared of MMP3 [52] . Aβ treatment causes upregulation of MMP3 expression in primary astrocyte cultures and in mixed hippocampal cultures [53] . MMP3 also degrades extracellular Aβ [54] . The closely related MMP2 and MMP9 proteins have been well studied in the context of AD pathogenesis . Astrocytes that surround plaques in AD mouse brains show enhanced MMP2 and MMP9 expression [55] . Conditioned astrocyte media is sufficient to reduce synthetic Aβ levels , which is abolished with treatment of inhibitors specific to MMP2 and MMP9 [55] . MMP9 can degrade fibrillar Aβ in vitro and plaques in hippocampal slice cultures [52] , [56] , [57] . In an AD mouse model , knocking out MMP2 and MMP9 results in increased steady-state Aβ [55] . Thus , MMP proteins may contribute to Aβ clearance by promoting Aβ catabolism . Supporting the relationship of MMP3 with Alzheimer's disease , CSF MMP3 levels are increased in individuals with a ptau181/Aβ42 ratio indicative of AD [50] . In a previous study , we failed to detect an association between significant SNPs in this study and CSF Aβ42 , Tau or pTau181 levels ( table 3 ) [9] . Studies testing the association of MMP3 SNPs and haplotypes and risk for AD have produced mixed results [58]–[60] . There appears to be two tiers of association in this locus , one group of SNPs with p-values less than 1E-38 , which includes the non-synonymous SNP rs679620 , and another which p-values between 1E-08 and 1E-12 , including several variants with inferred regulatory function . Conditional analyses of our data in this region indicate that the more strongly associated group of variants tags a single association signal and that no independent associations are detected in this region . While we cannot definitively identify the causal variant , rs679620 , a non-synonymous SNP in the MMP3 region , was significantly associated with increased CSF MMP3 levels . This suggests a possible protective effect of this variant with respect to AD . We found that rs573521 , rs645419 and rs679620 are associated with increased CSF MMP3 levels in this study and with reduced risk of AD in the IGAP study . These associations provide additional support for the role of MMP3 in AD pathology . In addition , the identification of SNPs in the intergenic region between MMP3 and MMP1 that are associated with both MMP3 and MMP1 levels suggest a common regulatory locus or close functional relationship between these members of the MMP gene family . The identification of SNPs near MMP3 that are associated with increased CSF MMP3 protein levels and reduced AD risk supports the protective role of MMP3 in clearing Aβ from human brains . CCL2 , also called monocyte chemotactic protein-1 or MCP-1 , is encoded by the CCL2 gene , located on chromosome 17q11 . 2-q12 . It is a chemokine that is involved in immunoregulatory and pro-inflammatory processes . Amyloid plaques in AD brains are surrounded by activated immune cells that produce CCL2 among other chemokines [61] . In the absence of CCL2 , amyloid pathology is accelerated in an AD mouse model , illustrating its important role in amyloid plaque clearance and pointing to a potentially reparative role in AD [62] . However , overexpression of CCL2 in an AD mouse model resulted in marked accumulation of reactive microglia and enhanced diffuse plaque accumulation , suggesting a role in Aβ aggregation [63] . In vitro work suggests that inhibition of CCL2 synthesis , reduces Aβ25–35- and Aβ1–42-induced toxicity in primary neuronal cultures [64] . Interestingly , treatment of primary astrocyte cultures with synthetic Aβ42 causes astrocytes to increase CCL2 synthesis and release [64] and astrocyte migration in response to CCL2 is reduced in the presence of Aβ42 [65] . Thus , Aβ42 in AD brains may stimulate astrocyte-mediated CCL2 release and result in increased neuronal susceptibility to Aβ42 toxicity . The immune system involves a delicate and perfectly coordinated balance to function well; so , it is conceivable that CCL2 could play reparative and deleterious roles in AD pathogenesis . A 2006 study evaluated CCL2 levels in serum samples from 48 individuals with Mild Cognitive Impairment ( MCI ) , 94 AD patients and 24 age-matched controls [66] . Significantly increased plasma CCL2 levels were found in MCI and mild AD , but not in severe AD patients , as compared with controls . It has also been reported that increased CSF CCL2 levels at baseline in patients with prodromal AD correlated with a faster cognitive decline during the study's follow-up period [67] . The largest study to date examined 600 samples with CSF CCL2 measurements and observed a significant increase in CCL2 protein levels with ptau/Aβ42 ratio indicative of AD [50] . As is the case with most pro-inflammatory cytokines and cytokine receptors , the levels increase in AD cases . Rs2228467 , located within the CCBP2 gene , is significantly associated with increased CSF CCL2 protein levels . The CCBP2 gene encodes the chemokine-binding protein 2 and demonstrates a high affinity for binding to CCL2 [68] . Conditional analyses suggest that this marker accounts for the entirety of the association signal in this region . Previous studies suggest that chemokine receptors can demonstrate high affinity binding to chemotactic proteins [68]; however , how polymorphisms in one chemokine affect expression and function of associated chemokines is poorly understood . PolyPhen2 and SIFT both predicted this amino acid change to be damaging . A recent study found that rs2228467 is significantly associated with lower circulating monocyte counts in the blood ( p = 1 . 57×10−7 ) [69] . If the rs2228467 polymorphism affects the chemokine function of CCBP2 or CCL2 , then this could have downstream effects on the recruitment of macrophages and dendritic cells and , in turn , monocyte development . CCL2 is known to be a necessary component in monocytes crossing the blood brain barrier [61] , [70] . Our findings suggest that increased CCL2 associated with variation at rs2228467 may cause a chemotactic response that results in lower levels of circulating monocytes in the blood . While rs2228467 has strong effects on CCL2 levels and CSF CCL2 levels change in Alzheimer's disease , this SNP does not appear to impact risk for AD or CSF Aβ42 levels ( p = 0 . 45 ) [9] . However , as CCL2 has been implicated in the pathogenesis of diseases characterized by monocytic infiltrates , like psoriasis [71] , rheumatoid arthritis [72] and atherosclerosis [73] further investigation of rs2228467 with regard to these and related diseases is clearly warranted . Here we have demonstrated in human subjects that SNPs in the CCBP2 gene are significantly associated with elevated CSF CCL2 protein levels . While , CSF CCL2 protein levels are not significantly associated with AD risk , evidence in mouse and cell models of AD suggest that increasing CCL2 levels increases microgliosis , amyloid plaque accumulation , and neuronal toxicity associated with Aβ . Taken together , these findings implicate CCBP2 and CCL2 as risk factors for AD pathogenesis . The CCL4 protein is encoded by the CCL4 gene ( 17q12 ) . Studies evaluating the relationship between AD and inflammation have shown that CCL4 is expressed in subpopulations of reactive astrocytes and in microglia [74] . Neuritic plaques in AD are surrounded by activated microglia and astrocytes , which may produce inflammatory products when stimulated with Aβ [75] . Macrophages showed an increased secretion of CCL4 when treated with Aβ . Current information concerning CCL4 and other plaque-associated chemokines suggests that their production plays a role in the recruitment and accumulation of astrocytes and microglia in senile plaques [76] . These data suggest a possible relationship between CCL4 and AD pathogenesis . We identified association between several SNPs , including one non-synonymous SNP , one synonymous SNP and five markers with predicted regulatory effects , and CSF CCL4 levels . Conditional analyses suggest that these SNPs tag a single association signal in this region . These SNPs do not show association with AD in the IGAP dataset or with CSF Aβ42 , Tau or pTau181 levels ( table 3 ) . In addition , CSF CCL4 levels are not significantly associated with ptau181/Aβ42 ratio , a predictor of AD status [50] . These results do not indicate a role for CCL4 levels in risk for AD . However , CCL4 may play a role in HIV Type 1 transmission , AIDS disease progression , and acute kidney injury [77] , [78] . Thus it will be important to evaluate the impact of rs6441977 ( V168M polymorphism in CCRL2; associated with decreased CCL4 levels ) and other markers with regulatory effects , on these and related diseases . The interleukin 6 receptor ( IL6R ) is a protein encoded by the IL6R gene ( 1q21 ) . Interleukin 6 is a potent pleiotropic pro-inflammatory cytokine that regulates cell growth and differentiation and plays an important role in the immune response and may also play a role in hippocampal neurogenesis [79] . We identified association with sIL6R levels for several SNPs in the IL6R region . Conditional analyses suggest that this is a single signal is driven by rs61812598 and other SNPs in high LD with these markers . Among these , rs2228145 , rs3811448 and rs4129267 have predicted functional effects ( see table 3 ) . Rs3811448 , a non-synonymous marker in the associated region is not significant when conditioning upon rs61812598 , rs4129267 or rs2228145 . Conversely , both rs61812598 rs4129267 and rs2228145 remain highly significant upon conditioning for rs3811448 . This makes it clear that there is a single signal in this region , tagged by rs61812598 , rs4129267 and rs2228145 , which are in complete LD with each other . Rs2228145 is a non-synonymous polymorphism in the IL6R gene , ( D358A ) . While both SIFT and Polyphen 2 predict this change to be benign , rs2228145 has recently been shown to significantly increase plasma concentrations of sIL-6R , and reduce concentrations of membrane-bound IL-6R , resulting in impaired IL-6 responsiveness [80] . These results demonstrate that consequential changes in protein levels , likely resulting from the rs2228145 polymorphism , may translate into a functional impairment in IL-6R signaling . The rs2228145 polymorphism and other SNPs in this region have previously been shown to be significantly associated with plasma sIL-6 levels [21] , [81] Trans-signaling is important for IL6-mediated cellular communication with molecular targets and that blocking trans-signaling lessen the deleterious effects of IL6 signaling [79] . Trans-signaling occurs via sIL6R , which is the proteolyzed product of IL6R . IL6R is proteolyzed by γ-secretase , which also cleaves APP [82] . Thus , polymorphisms in IL6R that modify IL6 signaling may result in modulation of signaling to many downstream targets that directly or indirectly influence AD pathogenesis . Additionally , rs2228145 has been implicated previously as significantly increasing the risk of sporadic AD in a Chinese Han population in subjects without the APOE ε4 allele [83] . While sIL6R levels have been previously reported to decrease in AD cases [84] , a recent study using a much larger sample found a significant increase in CSF sIL6R levels and increasing ptau181/Aβ42 ratio [50] , suggesting a possible relationship between sIL6R levels and AD pathogenesis . It has been proposed that there is a reciprocal relationship between IL-6 and Aβ . The IL-6/sIL-6R complex is reported to enhance APP transcription and expression [85]–[87] . Based on these data , rs2228145 , which is associated with increased sIL6R levels , would be predicted to alter CSF Aβ42 levels and risk for AD . Unfortunately , we did not detect association of markers in IL6R with AD risk in the IGAP analysis and association with AD biomarkers was weak and inconsistent ( table 3 ) . Rs4129267 is located within the intronic region of IL6R . This SNP is inferred to be “likely to affect binding” of the Olf-1 transcription factor using RegulomeDB . Like rs2228145 , this marker is associated with sIL6R levels [21] , [81] . In addition , this marker has been reported to be associated with levels of fibrinogen and C-reactive protein in blood as well as asthma and pulmonary function [31] , [32] , [88]–[90] . While it remains unclear what the causal marker is for this association signal due to the high levels of LD , the putative functional effects of both rs2228145 and rs4129267 make them top candidates for future investigation . Dysregulated production of IL6 and its receptor are implicated in the pathogenesis of many diseases , including multiple myeloma [91] , autoimmune diseases [92] , and prostate cancer [93] . In addition , the association of several significant SNPs in our study with asthma , C-reactive protein levels and coronary heart disease highlights the relationship between the inflammatory response and these disorders . This information suggests a central role for the IL6/sIL6R complex in these and possibly other diseases and suggests that further characterization of the effects of rs2228145 and rs4129267 on human disease phenotypes is warranted . In conclusion , we have identified significantly associated SNPs for five different AD-related analytes . These associations are robust across different biological fluids , dementia status and inferred presence of AD pathology both within and between independent sample sets . The SNPs observed to be associated with CSF ACE and MMP3 levels also appear to show association with AD in the predicted direction , providing support for previous hypotheses of involvement of these genes and their function in amyloid clearance for risk for AD . While inflammation is known to play an important role in AD , the pro-inflammatory markers investigated here , and their associated SNPs , do not appear to alter AD risk or disease progression . However , because the immune system is an exceedingly complex set of signaling cascades that must be perfectly regulated in order to function properly and because this regulation involves constant flux , we may not be able to fully capture the subtle effects in the function of these proteins that have a cumulative effect on AD pathogenesis over the course of a lifetime . The reproducibility of our findings in cognitively normal individuals and in plasma levels of the respective proteins as well as putative functional effects of these variants suggest that these SNPs may directly affect their respective proteins . Thus , insights into the genetic basis of variance in important pro-inflammatory protein levels are relevant to other diseases that are modulated by those processes . Finally , our findings demonstrate the continued utility of an endophenotype-based approach to finding functional alleles and disease-associated loci .
CSF and plasma samples from the Knight ADRC were evaluated for levels of 190 analytes using the Human DiscoveryMAP Panel and a Luminex 100 platform . CSF and plasma samples from the ADNI sample were assessed using the same Human DiscoveryMAP Panel and measurement platform [21] , [94] . After filtering each set independently for phenotypes that had valid measurements in at least 90% of the samples , the intersection resulted in 76 analytes . For each of the 76 analytes that passed quality assurance in both datasets we performed a PubMed search on August 11 , 2013 . The purpose of this literature search was to reduce the phenotype list to those that are relevant to our AD centered samples from the Knight ADRC and ADNI , thus reducing the dimensionality of the data and concentrating statistical power on the most relevant phenotypes . Search terms included any of the following three terms , the name of the analyte on the chip , the official gene name and the official protein name and the term “Alzheimer's disease” . Analytes with more than 50 search results are considered to be “AD-related” . For analytes with fewer than 50 search results we inspected the manuscripts manually to determine whether there was evidence for a relationship with AD . A list of all analyte names on the chip along with official gene and protein names and results of the literature search is provided in supplementary Table S5 . All samples were genotyped using the Illumina 610 or the Omniexpress chip . Prior to statistical analysis , sample data were filtered using rigid quality control ( QC ) criteria by array: minimum call rate ( 98% ) , minimum minor allele frequency ( 2% ) , and exclusion of SNPs out of Hardy-Weinberg equilibrium ( p<1×10−6 ) . Unanticipated duplicates and related individuals were prioritized after calculating pairwise genome-wide estimates of identity-by-descent . Eigensoft was used to calculate principal component factors for each sample and confirm ethnicity [96] . These calculations were included as covariates in our analysis to adjust for possible confounding effects of population stratification . The 1000 genome data ( June 2012 release ) and the Beagle software were used to impute genotypes in the combined ADNI and Knight ADRC samples [97] . SNPs with a Beagle r2 of 0 . 3 or lower , a minor allele frequency ( MAF ) lower than 0 . 05 , out of Hardy-Weinberg equilibrium ( p<1×10−6 ) , a call rate lower than 95% or a Gprobs score lower than 0 . 90 were removed . A total of 5 , 815 , 690 SNPs passed the QC process . The Kolmogorov-Smirnov goodness-of-fit test was performed to evaluate normality of the 59 phenotypes of interest in the Knight ADRC samples . When deviations were observed , phenotypes were log transformed to approximate a normal distribution . The ADNI data for these samples and phenotypes had already been adjusted to fit normal distribution patterns by the ADNI biomarker core . Associations reported for age and gender were performed in cognitively normal samples only to reduce potential confounds of dementia . We performed a genome-wide association for each of the 59 phenotypes to identify genetic loci associated with protein levels in CSF . For the initial association analysis in each series we used PLINK to perform linear regression and evaluated the association between the additive model for 5 . 8M SNPs and each phenotype [98] . Age , gender , and the principal components from Eigensoft analysis were included as covariates . Variance explained by each marker is reported as the difference in the model r2 between full models with and without the SNP included as a variable . Association of SNPs of interest with plasma analyte levels in the ADNI and Knight ADRC samples was calculated using the same approach . Analysis of each sample separately reduces the possible confound of demographic or ascertainment differences between the ADNI and Knight ADRC samples . Genome-wide association results from the two datasets were meta-analyzed using the default settings in METAL [99] . Genomic inflation factor scores ( GIF ) were estimated using the R package GenABEL [100] . We set a strict and extremely conservative study-wide alpha level of 1 . 46×10−10 for the combined analysis . This was calculated by applying a Bonferroni correction for 5 . 8 million SNPs and 59 analytes , or 342 . 2 million tests . SNPs that met the initial significance criteria were further filtered using the following criteria . First , we rejected SNPs where the direction of the effect was different in the Knight ADRC and ADNI datasets . Second , we removed all SNPs where the minor allele frequency was less than 5% ( unless they were directly genotyped ) . Finally , we rejected all associations with phenotypes where the genomic inflation factor was greater than 1 . 03 ( GIF was calculated without SNPs where MAF is <0 . 05 ) . Conditional analyses on each of the genome-wide significant SNPs were conducted using the –condition function in PLINK . We also performed additional analyses in the genome-wide significant loci to determine the stability of the results when stratified by clinical AD status and by CSF AB42 levels . CSF AB42 strata were based on levels that approximate amyloid deposition detected in PET scans using Pittsburgh Compound B ( PIB ) . For the KADRC samples values less than 500 pg/ml indicate PIB retention/Aβ deposition , while values greater than 500 pg/ml indicate PIB negativity and the absence of Aβ deposition [95] . For the ADNI samples values less than 192 pg/ml indicate retention/Aβ deposition , while values greater than 192 pg/ml indicate PIB negativity and the absence of Aβ deposition [101] . We used the R package Multiphen , which performs a multivariate test of the linear combination of phenotypes most associated with the genotypes at each SNP , to evaluate each of the top hits for joint effects on multiple phenotypes in the study [37] . Analysis was carried out in the KADRC and ADNI samples separately using default settings as described here ( http://cran . at . r-project . org/web/packages/MultiPhen/vignettes/MultiPhen . pdf ) . Initial results of IL6R indicated a GIF statistic greater than 1 . 03 suggesting the p-values were inflated by confounding variables . By removing a 500 kb window on either side of the strongest signal we identified that the inflation was due to the large number of highly significant p-values surrounding the IL6R gene ( adjusted GIF = 1 . 017 ) . We did not remove this phenotype as it appears the inflation is due to a strong and replicable association signal in this single region . We analyzed other phenotypes that failed GIF quality control using the same strategy and did not observe similar phenomena . For each locus where association was detected with the CSF endophenotypes we obtained data from the International Genomics of Alzheimer's Project ( IGAP ) association study of AD Stage 1 results [30] . IGAP is a large two-stage study based upon genome-wide association studies ( GWAS ) in individuals of European ancestry . In stage 1 , IGAP used genotyped and imputed data for 7 , 055 , 881 single nucleotide polymorphisms ( SNPs ) to meta-analyse four previously-published GWAS datasets consisting of 17 , 008 Alzheimer's disease cases and 37 , 154 controls ( The European Alzheimer's disease Initiative – EADI the Alzheimer Disease Genetics Consortium – ADGC The Cohorts for Heart and Aging Research in Genomic Epidemiology consortium – CHARGE The Genetic and Environmental Risk in AD consortium – GERAD ) . In stage 2 , 11 , 632 SNPs were genotyped and tested for association in an independent set of 8 , 572 Alzheimer's disease cases and 11 , 312 controls . Finally , a meta-analysis was performed combining results from stages 1 & 2 . We used ANNOVAR to annotate SNPs of interest with location and functional information [102] . RegulomeDB was used to annotate SNPs within known and predicted regulatory elements [103] . We used SIFT and POLYPHEN2 for preliminary assessments of the functional consequences of amino acid changes [104] , [105] . All data collection was conducted under approval by the appropriate Institutional Review Boards . Analyses presented here were approved by the Institutional Review Board at Brigham Young University ( E110252 ) .
|
The use of quantitative endophenotypes from cerebrospinal fluid has led to the identification of several genetic variants that alter risk or rate of progression of Alzheimer's disease . Here we have analyzed the levels of 58 disease-related proteins in the cerebrospinal fluid for association with millions of variants across the human genome . We have identified significant , replicable associations with 5 analytes , Angiotensin-converting enzyme , Chemokine ( C-C motif ) ligand 2 , Chemokine ( C-C motif ) ligand 4 , Interleukin 6 receptor and Matrix metalloproteinase-3 . Our results suggest that these variants play a regulatory role in the respective protein levels and are relevant to the inflammatory and amyloid processing pathways . Variants in associated with ACE and those associated with MMP3 levels also show association with risk for Alzheimer's disease in the expected directions . These associations are consistent in cerebrospinal fluid and plasma and in samples with only cognitively normal individuals suggesting that they are relevant in the regulation of these protein levels beyond the context of Alzheimer's disease .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"quantitative",
"trait",
"association",
"studies",
"statistical",
"analysis",
"of",
"genetic",
"association",
"inflammation",
"medicine",
"and",
"health",
"sciences",
"genetic",
"association",
"studies",
"genome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"immunology",
"computational",
"biology",
"genetics",
"of",
"disease",
"neuroscience",
"immune",
"response",
"human",
"genetics"
] |
2014
|
Genome-Wide Association Study of CSF Levels of 59 Alzheimer's Disease Candidate Proteins: Significant Associations with Proteins Involved in Amyloid Processing and Inflammation
|
Exotic pathogens and pests threaten ecosystem service , biodiversity , and crop security globally . If an invasive agent can disperse asymptomatically over long distances , multiple spatial and temporal scales interplay , making identification of effective strategies to regulate , monitor , and control disease extremely difficult . The management of outbreaks is also challenged by limited data on the actual area infested and the dynamics of spatial spread , due to financial , technological , or social constraints . We examine principles of landscape epidemiology important in designing policy to prevent or slow invasion by such organisms , and use Phytophthora ramorum , the cause of sudden oak death , to illustrate how shortfalls in their understanding can render management applications inappropriate . This pathogen has invaded forests in coastal California , USA , and an isolated but fast-growing epidemic focus in northern California ( Humboldt County ) has the potential for extensive spread . The risk of spread is enhanced by the pathogen's generalist nature and survival . Additionally , the extent of cryptic infection is unknown due to limited surveying resources and access to private land . Here , we use an epidemiological model for transmission in heterogeneous landscapes and Bayesian Markov-chain-Monte-Carlo inference to estimate dispersal and life-cycle parameters of P . ramorum and forecast the distribution of infection and speed of the epidemic front in Humboldt County . We assess the viability of management options for containing the pathogen's northern spread and local impacts . Implementing a stand-alone host-free “barrier” had limited efficacy due to long-distance dispersal , but combining curative with preventive treatments ahead of the front reduced local damage and contained spread . While the large size of this focus makes effective control expensive , early synchronous treatment in newly-identified disease foci should be more cost-effective . We show how the successful management of forest ecosystems depends on estimating the spatial scales of invasion and treatment of pathogens and pests with cryptic long-distance dispersal .
The invasion of ecosystems by non-native plant pathogens and insects [1] , [2] , [3] , [4] poses a growing threat to ecosystem function and conservation as global trade , travel and environmental change create opportunities for introduction and establishment of exotic organisms [5] , [6] , [7] , [8] . Cryptic infection ( i . e . asymptomatic or undetectable for a period of time ) and long-distance dispersal ( i . e . with a fat-tailed probability distribution ) of transmissible pathogens and pests present two serious impediments to the effective control of these organisms: the epidemics only become apparent once symptoms develop , by which time the outbreak will have grown , and new , sometimes distant foci may have been established through long-distance dispersal [9] . When the invading agents are unknown , they are likely to spread unnoticed and unchecked for even longer if their identification is difficult and their transmission poorly understood . The combination of delayed detection of cases with long-distance dispersal has the potential to sustain invasive spread even under modestly favourable conditions for the invading organism . In fact , management strategies that are restricted to the treatment of symptomatic hosts are likely to fail without offering much return for the resources deployed [10] , [11] , [12] . The challenge in devising epidemiologically- and economically-viable management strategies lies instead in matching the temporal and spatial scales of control with often poorly understood temporal and spatial scales of epidemic spread . Adopting such a scale-matching approach at a landscape level requires estimation of the actual spatial extent of the epidemic , including the location and speed of its expanding front . The degree of matching that can be achieved is determined by governing principles , regarding the location and nature of management actions , illustrated schematically in Fig . 1 . The problem of cryptic infection is not restricted to natural communities . For example , cryptic infection has frustrated efforts to eradicate the AIDS pandemic [13] and regional endemics of malaria [14] . However , in natural communities there are specific difficulties in matching scales of control with intrinsic epidemic scales . These difficulties include accessing sites to detect and control new infections , identifying the range of host species of a pathogen , and surveying the spatial distribution of hosts in a heterogeneous landscape . Within this context of uncertainty on multiple scales , the use of computational models for linking epidemiological , landscape and weather dynamics over large regions [15] , [16] , [17] , [18] can assist in assessing the effectiveness of disease management strategies . The use of models to inform disease management across landscapes can be divided into a number of interlinked stages: ( i ) construction of robust models that capture enough biological features to exhibit realistic dynamics; ( ii ) estimation of parameters such as dispersal distances and rates of spread from incomplete observations of infection; ( iii ) use of the model to predict the current extent of symptomatic and asymptomatic infection in order to assess the current status of damage and risk; ( iv ) given the estimated current status , use of the model to predict future pathogen spread under different management scenarios . Unavoidably , these steps are taken under uncertainties about host distribution and density , abiotic forcing such as weather , and responses of pathogen and host life-cycles to treatments . In this paper , we study the efficacy of management strategies for controlling epidemics of forest pathogens with cryptic infection and long-distance dispersal , focusing on the emerging water mould Phytophthora ramorum , the cause of sudden oak death [19] . The current epidemics of sudden oak death , particularly in California , USA , are similar in extent and severity to historical outbreaks of white pine blister rust [20] , chestnut blight [21] , [22] , [23] , and Dutch elm disease [24] , [25] . Limited understanding of pathogen transmission and biology , lack of multiple-scale epidemiological insight , and failure to recognize cryptic pathogen spread , contributed to the failure of attempts to manage these historical outbreaks [22] . We use computationally-intensive approaches for modelling and parameterizing spatio-temporal stochastic population dynamics in order to examine scenarios for the control of emerging epidemics in natural forest landscapes . We consider how to design efficient control strategies that account for uncertainties associated with cryptic infection and long-distance dispersal in heterogeneous host landscapes . Specifically , we focus on the control and management of sudden oak death in redwood-tanoak and Douglas-fir-tanoak forests of northern California . Our purpose is twofold . First , to illustrate a case study where availability of epidemiological and landscape data - typical of datasets that are or can be collected in natural ecosystems - allows modelling of pathogen spread and control , and offers advisory messages on current and future invasive organisms that are difficult to control . Our second purpose is to provide practical guidance for planning of control and prevention of further spread of P . ramorum , both in California and in temperate and coastal areas of Europe and the eastern USA where outbreaks have occurred in nurseries but spread in the wild has apparently been limited [26] , [27] , [28] , [29] . Recent rapid spread in larch plantations in Britain and Ireland has been causing great concern and offers a cautionary example of the unpredictable impacts of this pathogen in new environments [30] . Phytophthora ramorum has been expanding its range in coastal California since the mid-1990s , killing millions of trees , including oak ( Quercus spp . ) and tanoak ( Lithocarpus densiflorus , recently attributed to a new genus as Notholithocarpus densiflorus ) [27] , [31] , [32] . This epidemic has caused damage to public and private property , economic impact on nursery , gardening and logging industries , and increased the cost of implementing regulatory activities [33] , [34] . Many are also worried that large-scale tree mortality will have profound long-term environmental consequences , by changing the structure of plant and microbial communities , altering landscape ecological structure and function , and increasing forest-fire hazards [3] , [27] , [31] , [35] , [36] , [37] , [38] . Phytophthora ramorum is known to infect over one hundred species of forest shrubs and trees [27] . On oak and tanoak trees , P . ramorum causes bleeding bole cankers that can lead to relatively rapid mortality; therefore the disease name of sudden oak death . Other hosts such as California bay laurel ( Umbellularia californica ) suffer mild leaf-blight or twig-dieback symptoms and are major sources of inoculum for infection of oaks and tanoaks [39] . There is no evidence for sporulation of this pathogen from true oak species ( Quercus sp ) in Californian ecosystems [39] . Transmission of inoculum is thought to occur both locally and over long distance via rain splash , stream and river currents , wind and mist , and human-mediated transport [27] , [28] , [40] , [41] , [42] . The earliest symptoms of P . ramorum–invasion of a site are often small lesions on the foliar hosts , with minor effects on tree health and similar to lesions caused by native pathogens . Hence , confirmation of invasion by P . ramorum requires extensive on-the-ground sampling and laboratory isolation of the pathogen , which makes early diagnosis over large areas impractical . The alternative method of aerial surveying ( Fig . 2 ) can only detect host mortality , which may be preceded by pathogen establishment by several years . These limitations in surveying are a primary cause for cryptic infections of this pathogen . Control of P . ramorum poses significant epidemiological challenges: in addition to its cryptic and long-distance spread , the long infectious period and generalist nature of the pathogen aid its transmission across heterogeneous landscapes . Moreover , current measures for controlling P . ramorum at the landscape scale consist mostly of host removal , as no effective chemical treatment or biological control exists [27] , [43] , [44] . These measures are restricted by economic cost , logistics , and limited options for coordination with private landowners [27] , ; they are also complicated by the disparate epidemiological , commercial , and amenity importance of the different hosts . In California , control of P . ramorum has been largely limited to state-wide quarantine [46] , [47] and small-scale treatments [48] as the epidemic has grown rapidly on multiple fronts . Subsequently , desirable epidemic control via coordinated scaled-up treatment and prevention has been hampered by uncertainty about how to act effectively with limited resources , allowing the scale of the problem to continue to aggravate . In the late 1990s , the P . ramorum epidemic took a geographic leap from the main focus around the San Francisco Bay Area , probably through human-mediated pathogen transport [40] , to establish two disjunct outbreaks in northern California and southern Oregon ( Fig . 2 ) [49] . The Curry County , Oregon outbreak was first reported in 2001 [41] . Since the early stages of its detection , this outbreak has been kept under intensive control through extensive monitoring and removal of both infected host material and surrounding hosts as a buffer [41] , . These aggressive treatments have contained , but not eradicated , P . ramorum , which has continued to spread within a relatively small geographic region , probably due to cryptic infection that makes early detection difficult [41] , [48] . The pathogen has now been detected in scattered clusters in Curry County that add up to an area of about 80 ha [44] . However , the eradication attempts have prevented local intensification of the disease and minimized damage to the forest [44] . The other isolated outbreak , in Redway , Humboldt County , California was reported in 2002 [45] . Unlike the Oregon outbreak , eradication was not attempted and management has been of limited extent at the Humboldt County site; this outbreak has expanded in each subsequent year [51] ( Fig . 2A–B ) . We develop and parameterize a mathematical model for forecasting the spread of P . ramorum and assessing control options in the Humboldt focus , the northern forefront of the Californian epidemic where host and environmental conditions [49] favour spread over a large stretch of forest ( ∼200 km by 75 km ) extending up to Curry County , Oregon ( Fig . 2A ) . The model combines the key aspects of P . ramorum epidemiology with data on vegetation distribution [49] and weather variation , and shares features with a model we have developed for the spread of P . ramorum in California in the absence of disease control [32] . We use this model to explore the following control scenarios being considered by policy makers [27] , [45]: removal of hosts , protective aerial spraying ( an experimental technique [51] , [52] ) , and construction of a host-free “barrier” [53] . Within each scenario , we adopt strategies with differing degrees of match between the scales ( spatial and temporal ) of the control and the spread of P . ramorum . The outcomes of all these measures are uncertain but critical , given the amount of resources at stake and the risk in case of failure or lack of implementation of control . A consequence of the non-intuitive dynamics arising from the intertwining of multiple scales of pathogen dispersal with cryptic and symptomatic infections ( e . g . , Fig . 1 ) , is that some conclusions about P . ramorum control in California may not follow our expectations . For example: Should we remove hosts at or ahead of the outbreak focus ? When would protective spraying be most effective ? Which size of barrier would work ? More generally , we demonstrate conditions for achieving effective control ( either delay in spread or eradication ) of plant pathogens or forest insects with cryptic and long-distance spread .
Landscape epidemiology uses concepts from epidemiology and landscape ecology in order to understand natural and managed disease dynamics on a large scale , such as regional or continental scales [8] , [15] , [16] , [17] . In applying strategies for the control of invading pathogens at the landscape level , we need to consider limiting principles that determine the maximum gain achievable and the minimum effort ( and economic expenditure ) required given the current state of the epidemic . The limits posed depend on the degrees of cryptic and long-distance spread of the specific pathogen within the host landscape , as well as on the goals of the intervention ( Fig . 1 ) . First , if the aim is eradication of a local outbreak , it is essential to match the spatial extent of the control area to that of the pathogen . For pathogens with cryptic infection and long-distance dispersal the extent of the epidemic is likely to be larger and increase faster than what estimates based on observed symptoms suggest . Second , if the aim of the intervention is control in a particular area of the outbreak , it is necessary to re-apply treatment to make up for partial coverage and partial effectiveness of each round and clearing reinvasion from non-treated infected areas . Finally , if the aim is to protect a target area ( at-risk but not infested ) we need to assess how extensively to treat in and around that area in relation to the distance to the advancing front . In all cases , it is essential to estimate the full extent of infection , including its moving front , and the rate of pathogen spread in order to control disease effectively [10] , [12] . Other modelling studies have examined principles of pathogen invasion in heterogeneous landscapes [4] , [17] , [54] , [55] , but addressed animal diseases or pathogens that pose a different challenge to control than the combination of cryptic infection and long-distance aerial dispersal that we have considered . We focus on Humboldt County as a case study for three reasons . First , this outbreak of P . ramorum is geographically isolated [40] and has grown with minimal intervention , which offers an opportunity to estimate the natural spread of P . ramorum in the wild . Second , we estimate dispersal and transmission parameters from aerial survey data on pathogen spread ( provided by the USDA Forest Service , Forest Health Protection ) that are unique in California and elsewhere , in that they cover a whole , mostly non-managed outbreak area for several years ( 2004–2009 ) and avoid some of the incompleteness and biases that ground surveys inevitably encompass . This aerial survey dataset was verified in two ways . Field verification of aerial detections of the pathogen was done through ground monitoring by local scientists trained in the identification of P . ramorum disease followed when necessary by laboratory confirmation . Although it was not possible to apply this approach to every detected symptomatic tree , a high degree of confidence in the surveys is conferred by the following factors: verification was done after every annual survey; priority was given to edges and outliers in the spatial pattern of detection; patchiness in tree mortality provides a strong signature of true detection at stand level . The aerial records were also checked by comparison with field observations from a permanent study-plot network and an early-detection watershed-level survey based on pathogen baiting in streams and rivers [38] , [56] . In addition , we did a systematic validation of aerial detections against the presence of suitable hosts of P . ramorum using vegetation distribution databases such as CALVEG [57] . The third reason for using Humboldt County as a case study is that the isolation of the outbreak offers opportunities for control and requires management decisions that are specific to this outbreak . Phytophthora ramorum infections were first reported in Humboldt County in 2002 around the town of Redway [45] . The affected area in Humboldt has since then grown at an increasing rate with the mortality of tanoak and oak trees scattered over thousands of hectares ( Fig . 2B , and Fig . S4 in Text S1 ) . Tanoak mortality peaked in 2007 and has slowed since , likely due to low spring-rainfall from 2007 until 2009 ( Fig . S1B in Text S1 ) . The disease has spread predominantly northward of the initial focus near Redway , probably due to prevailing winds . To date , only moderate , localized control measures have been applied in Humboldt with evidence that they might have had an impact at ( but possibly not beyond ) the scale of the treated individuals and plots [45] , [51] . The geographic isolation of this focus from the wider epidemic initially raised hopes of eradication [45] , but the current size of the focus suggests that amelioration and containment are more realistic goals . No direct measures of the area with cryptic infection in Humboldt ( which is wider than the area with symptoms ) exist , because of the large spatial extent of the region that would need ground surveying for the presence of the pathogen , the limited resources to do so , and the spatial heterogeneity in landownership and in landowner cooperation with monitoring efforts . Options for controlling P . ramorum are currently limited to removal of inoculum ( i . e . , culling and burning of diseased hosts ) , removal of hosts ( i . e . pre-emptive host culling with herbicide or cutting ) , and chemical protection; but no curative chemical treatment or biological control exists [27] , [43] . In each of these approaches there is difficulty in field identification of infected hosts , treatment costs are high , treatment permits are slow to obtain , and the logistics of working in areas with many small landowners complicate the implementation of treatment . We explore the following control strategies initiated in 2010 and implemented in differing spatial areas ( Fig . 2D ) . 1 ) Removal at the origin , in an area containing the focus ( Area 1 , Fig . 2D ) about once per year; this strategy includes follow-up monitoring ( more frequently in cells with more abundant hosts ) , partially-effective detection of symptoms , and removal of inoculum and hosts in symptomatic ( and adjacent ) stands using host removal , herbicide treatment and pile burning [27] , [51] . 2 ) Removal ahead of the origin , in an area north of the focus ( Area 2 , Fig . 2D ) ; is otherwise identical to ‘removal at the origin’ . 3 ) Mixed strategy: Aerial spraying with Agri-Fos® ( a phosphate compound ) [43] on a large scale [52] to provide temporary partially-effective protection of hosts ( e . g . tanoak ) and prevent northern spread ( to the Target ) ; here we combine inoculum ‘removal at the origin’ ( Area 1 ) and , with the same frequency , spraying ‘ahead of symptoms’ in areas with lower human-population density ( Area 2 ) . While some aerial spraying experiments are ongoing in Oregon , the long-term efficacy and practicality of these treatments has not yet been established . In California , it is likely there would be limited willingness of landowners to approve aerial spraying , which would impede large-scale host protection treatments in Humboldt . Therefore , we explore this control scenario as a hypothetical investigation of the impact of altering forest susceptibility at landscape level . 4 ) A host-free ‘barrier’ , an approach initially proposed for the vicinity of Redway when the disease focus was smaller [45] , but here located further north just south of Grizzly Creek , a tributary of the Van Duzen River watershed , to prevent northern spread ( to the Target , Fig . 2D ) ; a similar barrier has been under construction a few kilometres north of the location we consider in the model [53] . For all scenarios , we concentrate on a region containing the initial focus near the south edge and extending ∼85 Km north ( Fig . 2C–D ) , the predominant direction of spread . Control is implemented and ‘northern invasion’ defined according to a breakdown of this region into Area 1 ( comprising the focus ) , Area 2 ( north of the focus and predicted to contain less or no infection at the start of control ) , and the Target area ( predicted not to be infected at the start of control and to be protected from invasion ) . We study different spatial scales of control , i . e . , the size of Area 1 ( equal to that of Area 2 ) in relation to the spatial extent of cryptic infection ( set by the location of the epidemic front ) . The ‘barrier’ , located at the north edge of Area 2 , extends from east to west , is either 5 km or 10 km wide north to south , and is managed in order to remain host free . We run the control scenarios from 2010 to 2017 and , with an earlier start date , from 2005 to 2017 . Host removal and spraying are implemented roughly synchronously across the control area to optimize impact , and followed up to account for incomplete detection and partial coverage and effectiveness of treatments . We developed a probabilistic , spatially-explicit metapopulation model for the transmission dynamics of P . ramorum in a landscape of mixed-host stands represented by square cells ( 250 m by 250 m ) . Each cell has a susceptibility and infectivity that were evaluated based on its composition and density of host species , estimated using the CALVEG database of plant community distributions [57] implemented in a geographic information system ( GIS ) [49] . At each time , a cell can be in one of four states: Susceptible; Infected and asymptomatic ( cryptic ) ; Infected and symptomatic ( detectable ) ; or Removed ( where treatments are applied ) . Removed cells can be re-colonized via host re-sprouting or host re-invasion . Infected cells can transmit inoculum to susceptible cells according to a dispersal kernel ( a probability function of relative distance ) and have an average infectious period of 10 years . Several studies suggest that the infectious period of P . ramorum is limited ( albeit long ) and varies among species and environmental conditions [39] , [58] , [59] . In bay laurel , leaf shedding rates increase in the presence of foliar infection with a greater increase in dry than in cool and humid conditions [58] . These observations suggest a mechanism by which non-lethal hosts can recover from infection and limit their infectious period . In tanoak twigs and stems , no mechanism of recovery from P . ramorum infection is documented , but systemic infection is lethal in this host and observations suggest that no sporulation occurs on dead tanoak tissue [59] . Therefore , we assume a finite infectious period that is longer than the time since annual surveys of P . ramorum were initiated in California [38] . The effects of variable spring-rainfall and temperature on pathogen transmission through sporulation and infection [39] , [58] , [59] are accounted for in the estimation of the model parameters and in the predictions . The model was parameterized using aerial surveys of tanoak mortality in the Redway area between 2004 and 2009 . We applied Bayesian Markov chain Monte Carlo , data-augmented inference [60] to estimate the time and location of the index case ( year 2001 , 2–3 km south of Redway ) , the rate and ‘spatial scale’ of transmission , and the rate of disease-induced tanoak mortality . The estimated average time between tanoak infection and mortality is about 2 . 5 years ( [2 . 3 , 2 . 9] 95% credible interval ) . In addition , we used this inference procedure to choose among candidate dispersal-kernel functions , which potentially can greatly influence the predictions of pathogen spread and of the impact of management strategies , as hypothesized at the beginning of the paper . We found that P . ramorum can disperse over large distances with a long tail of low probability: a power-law function fitted the data significantly better statistically than a negative-exponential function . We also contrasted the goodness-of-fit of the models based on each of the dispersal kernels through a visual comparison of predicted and actual progress of disease in space and in time ( Fig . S3 and S4 in Text S1 ) . We note that it was not possible to cross-validate the model against independent representative data because no such data were available . For some of the control scenarios explored , we calculated a local basic reproduction number ( R0 ) to assess the impact of treatment in the area where control is applied ( Fig . 1 ) . The local R0 is determined by the estimated dispersal kernel and transmission rate of the pathogen and by the post-treatment host-landscape . See Text S1 for further detail on model assumptions and formulation , estimation methods , and predictions . In order to probe the generality of the model outcomes , we considered three scenarios representing a likely range of ability or risk of the pathogen to spread in the host landscape: “high” , “medium” and “low” pathogen-spread scenarios . We defined these scenarios using the inferred uncertainty about the three estimated parameters characterizing pathogen transmission and the period of cryptic infection . We may think of this ability or risk to spread as a pathogen trait encompassing the joint effect of several traits and factors . The medium-spread scenario corresponds to the median of the posterior distribution of the estimated parameters; this is the case considered in all results presented , unless stated . The high ( low ) spread scenario corresponds to a combination of parameter values leading to greater ( lower ) potential of the pathogen to spread and lower ( higher ) efficacy of practitioners to detect infection; we chose these values based on 95% credible regions of the parameters . We use these pathosystem scenarios to study how the impact of control strategies depends on our estimates of epidemiological parameters ( Text S1 ) , and to assess whether the predictions in the “medium-spread” scenario are representative , or could change under uncertainty about parameters or conditions presented by other host-pathogen systems .
First , we forecast the current size of the epidemic ( including cryptic and symptomatic infection ) and the current and future speed of its moving front under natural conditions , i . e . , without management actions . This step is essential as observations of disease symptoms do not reveal the full extent of the infection focus . We define the epidemic front as the stretch of landscape where the probability of invasion changes from 95% to 5% as the distance from the focus increases . Assuming that the weather pattern in each year after 2010 equals the average of annual patterns during 2000–2009 , we predict that the epidemic front will advance northward at a speed of about 4 km/year ( Fig . 3A ) . The speed of the infection front is driven by weather and landscape conditions that affect the pathogen: it was slower before 2004 when the focus contained few unit cells , and during 2007 and 2008 when weather ( Fig . S1B in Text S1 ) and local landscape conditions were less favourable for infection , but faster in 2005 and 2009 when these conditions were favourable; the predicted slow down in 2013–14 is due to naturally-lower landscape-level contiguity of hosts in the northern part of the study region ( Area 2 , Fig . 2 ) . We forecast that in 2010 the front of the epidemic is situated 28 to 35 km north of Redway , between Miranda and the Van Duzen River in Humboldt County . These predictions do not account for heterogeneity in topography , which could affect spread . In addition , the predictions are likely to be sensitive to future change in annual weather and climate [39] , [59] , e . g . , caused by changes in the strength and duration of future El Niño/La Niña oscillation cycles , as suggested by the effects of past weather variability on model output ( Fig . 3A ) . It is possible , therefore , that future surveys and weather would yield different estimates of epidemic front dynamics . For example , our preliminary estimates based on data up to 2007 [61] yielded a faster advance of the front . Nevertheless , the predictions in Fig . 3A provide the best estimates available on current evidence . Sustained removal of inoculum on a smaller scale than the size of the epidemic focus at the start of control – either “at” or “ahead of” the origin ( Area 1 or Area 2 , <16 km or <40 km north of Redway ) – is effective locally but fails to contain or delay invasion of the Target area ( Fig . 4C–D ) due to spread from undetected cryptic infection . Despite the local basic reproduction number R0 dropping from >10 to <1 in either control area , elimination is thwarted by re-infection from non-controlled areas ( Fig . S5 in Text S1 ) . There is a marginal advantage in treating ahead rather than at the origin , because only part of Area 2 is infected in 2010 , which allows for a delay in spread of infection within and beyond it . Supplementing the removal at the origin with host protection ( Agri-Fos® spraying ) that stretches a few kilometres beyond the epidemic front ( Fig . 4E ) slows down the epidemic front from 4 to ∼1 km/year but fails to contain it . Indeed , the front re-gains speed in 2016 ( Fig . 3B ) as protection wanes ( and some susceptible forest is infected before the next spraying round ) and mounting inoculum disperses over this thinning “barrier” . Fig . 4D–E , 4C and 4F provide examples in support of the principles stated in Fig . 1C , 1D and 1E , respectively , i . e . , invasion ahead of the treatment area , re-invasion of the treated area , and spread over barriers . Sustained removal ( either alone or with host protection ) on a scale larger than the size of the epidemic focus at the start of control – either by increasing the size of the control area ( Fig . 5 ) or through early monitoring and treatment ( Fig . 6 ) – controls infection locally and delays invasion of the Target area significantly ( ≥1 year ) . First , removal starting in 2010 in an expanded area stretching well beyond the epidemic front ( ∼15 km ) reduces the overall level of inoculum and delays invasion of the Target for ∼3 years ( Fig . 5D ) . If removal had started in 2005 in the original , smaller area it would have delayed invasion of the Target by ∼1 year ( Fig . 6C ) . As above ( Fig . 4 ) there is a marginal advantage in treating ahead of ( Fig . 6D ) rather than at the origin ( Fig . 6C ) because inoculum is reduced nearer the front . The likely reason why removal yields only modest delays , even when expanded in space or time , is the delay in detection of infection within the control area due to the cryptic-infection period ( >2 years ) . Removal treatments alone do not cause a sufficient drop in inoculum to slow down the front significantly , although the drop is greater with control starting in 2005 ( Fig . 4 and 6C–D ) . Second , large-scale protection/spraying ahead of the origin , together with removal at the origin , starting in 2010 in an expanded area slows down the front speed to ∼0 . 5 km/year and prevents invasion of the Target for >6 years ( Fig . 5E ) . If this mixed strategy had started in 2005 in the original , smaller focus the entire host-protected area would have been infection-free initially . While the front speed would have decreased to about the same level as with an intervention starting in 2010 ( ∼0 . 5 km/year ) , the extent of host protection would have been maximized and invasion of the Target contained for a much longer period , well over 10 years ( Fig . 6E ) . A 5 km wide host-free “barrier” , just south of the Van Duzen River ( Fig . 4F ) , is ineffective at containing spread because inoculum builds up behind the barrier and occasional long-distance dispersal eventually succeeds in establishing new infection foci north of the host-free zone . A 10 km ( rather than 5 km ) wide barrier ( located 3 km further north ) ( Fig . 5F ) , also fails to contain spread overall but is successful in delaying spread for about one year . Overall , the control strategies involving removal or chemical protection of hosts slow down rather than interrupt spread due to the partial coverage and efficacy , and the limited temporal duration of the treatments , e . g . , not all hosts in a control area are treated and the effect on those that are treated is partial and temporary . Regarding where to apply curative treatment , the best location for removal depends on whether the goal is reduction of existing infection or containment of its spread . Both preventive measures , host chemical protection and the “host-free” barrier , need to be applied ahead of the front and the bigger the protected area or the wider the barrier the greater the impact . Removal is the only curative treatment for P . ramorum in the ecosystems of the western USA , and as such is the only treatment capable of reducing inoculum where it is already present ( e . g . Fig . S5 in Text S1 ) . The predicted epidemic growth over time , in the absence of interventions , over the range of pathogen-spread scenarios ( high , medium and low spread ) distributes approximately evenly about the medium-spread scenario and the survey data ( Fig . S6A in Text S1 ) . This evenness suggests these scenarios are representative of a likely range of epidemic potential associated with the inferred uncertainty in the estimated parameters . The differing ability of the pathogen to spread in each of these scenarios influences the relative impacts of control strategies in an expected way . As the spread potential of the pathogen increases , the invasion of the non-infected area ( Area 3 , Fig . 1 ) is delayed increasingly more ( Fig . S6B in Text S1 ) . Moreover , the ranking of the different control strategies according to their impact is preserved across the range of potential pathogen-spread scenarios ( Fig . S6C–E in Text S1 ) . We conclude that results comparing the effectiveness of control strategies in the medium-spread scenario ( Fig . 3 , 4 and 5 ) are qualitatively robust and representative of the viability of these strategies over more general conditions , including potentially other host-pathogen systems . Note that the measures of impact of the two removal strategies , “at” and “ahead of the origin” , crossover in the course of time because the outcomes of the strategies are case sensitive , as already stated . The mixed strategy is sustainable ( i . e . , the infection level remains stable in the long term ) in the medium- and low-spread scenarios ( Fig . S6C in Text S1 ) , while removal over an enlarged area ( larger than the cryptic epidemic , Fig . 5 ) is sustainable in the low-spread scenario .
Faced with an invading plant pathogen , it is vital for the success of control to identify the pathogen's biological and epidemiological features early on by collecting adequate field and laboratory data . The invading pathogen may be the cause of an established outbreak , such as P . ramorum in California and more recently in Western Europe , or an emerging threat , such as the risk of P . ramorum establishing in other areas . If the pathogen has cryptic spread and/or long-distance dispersal it presents non-intuitive multiple-scale dynamics that make it difficult to anticipate the impact of landscape management strategies and require re-evaluation of conventional approaches to regulatory and control activities [10] , [11] , [12] , [46] , [47] , [61] . For example , are expectations about the impact , timing , and location of treatments ( which are likely to have logistical delays and sparse spatial coverage ) justified ? For P . ramorum , cryptic infection makes it very difficult to identify the actual extent of outbreaks; which , combined with the pathogen's long-distance dispersal and long infectious period , leads to much uncertainty about an effective strategy for eradication or at least for containment of this emerging pathogen [41] , [48] . Historic epidemics of single-host pathogens with similar ability to spread cryptically have frustrated management actions in Europe and North America and caused extensive changes in natural and urban forest landscapes [21] , [23] , [24] , [25] . Currently , emerging generalist pathogens such as P . ramorum and P . cinnamomi , the cause of jarrah dieback [62] , [63] , endanger plant species in North America , Europe , Australia , and South Africa , prompting a global need to understand their dynamics and to identify effective management . By exploring options for the control of P . ramorum in northern California , we have demonstrated general principles for effective landscape control ( containment or eradication ) of forest pathogens characterized by cryptic and long-distance dispersal: 1 ) Continued monitoring of an at-risk target area is essential for early detection and prompt action . Our model shows that if treatment is not followed up its benefits will not be sustained ( Fig . S5 in Text S1 ) . 2 ) Curative treatment ( e . g . , removal ) or preventive treatment ( e . g . , chemical or pre-emptive culling ) , should , respectively , be applied rapidly on the scale of the whole infested area , including cryptic infections , or in a large-enough non-infested area that includes the host landscape at significant risk . Our model shows that that treatment in a limited area can be rapidly overcome by re-invasion through long-distance dispersal from non-treated areas . 3 ) If the control area is smaller than the infected area and there is long-distance dispersal , removal can be more effective “ahead of the origin” or “at the origin” depending on multiple factors , such as the extent of invasion “ahead of the origin” , the heterogeneity of the landscape , the tail of the dispersal kernel , and the efficacy of treatment . The choice of where to target treatment is highly debated by practitioners , including forest managers; our model suggests that this choice has to be made on a case by case basis . 4 ) The scale of the pre-control cryptically-infected area must be predicted using epidemiological data , and preferably also a parameterized model; this scale depends on the dispersal kernel and on how long the pathogen has been established . 5 ) Treatments should be applied as synchronously as possible across the control area to maximize the impact of resources and minimize pathogen escape , and should have repeated rounds to fight re-lapse of infection due to partial effectiveness , partial coverage , and ( if applicable ) re-infection from outside the control area . Our results indicate that partial coverage and lack of coordination ( Fig . 4–6 ) , or delay in follow up of treatment ( Fig . S5 in Text S1 ) can drastically and rapidly reduce the impact of management strategies , resulting in potentially very low cost-effectiveness of these actions . 6 ) Early treatment , when the expanding focus is smaller , is more cost-effective in achieving local control and protecting non-infected areas . Our results show that the growth in the extent of the cryptic epidemic can be much faster than what the visible epidemic suggests ( Fig . 4 and 6 ) . 7 ) If control is applied late when the epidemic focus has grown significantly large , the more feasible goals are local reduction of inoculum and containment ( e . g . , years of delay in spread , Fig . S6B in Text S1 ) , both of which can ameliorate local damage , may involve re-forestation , and could allow time for development of more-effective control tools . 8 ) Host-free “barriers” of plausible width can be ineffective at containing long-distance dispersal , unless there are additional buffers of spread ( e . g . , topographic features ) . However , wider barriers ( ∼10 km , in the current study ) can delay the epidemic front . Barriers have been proposed for controlling animal diseases [55] , but are less likely to be successful with aerially-dispersed plant diseases . The above principles extend , in essence , to the management of other forest pests , such as wood-boring insects , which have had increasing economic and ecological impact [64] , [65] . Unreliability in parameter estimates can affect our confidence in the predicted viability of control and management strategies . We found the results from our specific study , however , to be qualitatively robust to the inferred uncertainty in the parameters . For a generic host-pathogen system , there is uncertainty in the predicted efficacy of control strategies that involve removal of inoculum or host protection due to uncertainty in the pathogen dispersal kernel and transmission rate ( parameters α and β ) and in the model components representing the effects of the heterogeneous landscape and variable weather , all of which determine the pathogen's potential to spread and the severity of outbreaks . There is also uncertainty in the predicted impact of host protection , which depends on the duration of cryptic infection ( related to parameter rC ) , and in the predicted impact of removal of inoculum , which depends , in particular , on the rate of host re-invasion of treated stands . We expect the efficacy of a host-free barrier to depend chiefly on the tail of the dispersal kernel . The viability of a barrier would be affected also by directionality and extreme strength of winds , but these factors would most likely reduce the efficacy of the barrier even further , while topographic features could have the opposite effect ( c . f . Results ) . For forest pathogens , there are specific challenging steps in designing management strategies , such as acquiring host and pathogen landscape distribution data [40] , [56] , determining the effects of environmental conditions on inoculum production and establishment [39] , [66] , and developing techniques to estimate pathogen dispersal parameters and the extent of cryptic infection . However , forest diseases have some simplifying features relative to , for example , annual herbaceous-plant communities or contact structures ruled by individual movement and behaviour in animal and human populations: forest trees are long lived and do not move . These factors lead to comparatively slower changes in community-level inoculum production and host composition as hosts die [67] and to more straightforward short term forecasting . Yet , an important limitation in modelling forest diseases is the lower volume and greater biases in case recording compared with standard collection of clinical and veterinary data on human and livestock diseases . Similar issues apply to the predictive modelling of cryptically spreading forest pests . In relation to the sudden oak death outbreak in Humboldt County in northern California , our results suggest that P . ramorum will continue to spread north relatively rapidly in the medium and long term in the absence of effective landscape-level interventions . Spread on such scale could cause great damage in northern California , and eventually foil management attempts in Oregon [41] through the import of inoculum from uncontrolled epidemic foci . If extensive interventions were implemented , removal of inoculum on a sufficiently-large scale and frequency could delay the northern spread of the pathogen by several years . If this measure were supplemented with effective host protection ( a form of “vaccination” ) applied repeatedly ahead of the epidemic front , it could contain the spread for even longer . Large-scale chemical protection against P . ramorum is only at the very early stages of efficacy evaluation [52] and there would be substantial social , legal , and economic obstacles to its application in California . However , we explored this hypothetical control scenario to illustrate the potential impacts of changes in the epidemiological characteristics and spatial arrangement of hosts on the spread of P . ramorum . Our study suggests that the removal of infected hosts could be much more effective in ecosystems where landscape-level host communities are ( or have been made ) less susceptible to infection or support lower rates of sporulation . While we have demonstrated the importance of the epidemiological characteristics of host communities by considering reductions in susceptibility and sporulation of hosts that result from hypothetical chemical treatments in northern California , the implications of our results extend to other locations at risk of P . ramorum emergence such as eastern USA forests and parts of Europe . In such locations host characteristics might differ and/or it could become feasible to apply a form of extensive protection treatment in the future . While containment of the pathogen in southern Humboldt County is possible in theory , the estimated large size of the focus and potential long-distance dispersal of P . ramorum make the scale , nature of treatment , and coordination needed to do so a major challenge . Moreover , we find no evidence that a host-free “barrier” would contain the pathogen's dispersal for a significantly long time , at least under the assumption of similar topographic and weather conditions to those near Redway , the source of the epidemiological data used to parameterize the model . Nevertheless , although our results suggest that full containment is not likely , they also suggest that removal of infected hosts can reduce inoculum effectively within a control area and yield local benefits . This outcome is important for the implementation of policy on disease management and regulatory control , because removal of infected hosts is the only established means of treating infection by this pathogen . Moreover , applying the above measures on a more modest scale than we have considered could still delay epidemic growth sufficiently to allow time for ecosystem adaptation and management , therefore reducing the ecological , economic , and social impacts of disease [48] . Such delay would also ‘buy’ time for the development of chemical and biological control tools . Looking more widely into the benefits of disease management , large-scale control measures in Humboldt County should be designed also with the goal of achieving , or maintaining , forestry and other economic enterprises currently impacted by the presence of P . ramorum . Finally , the model suggests that the most viable strategy epidemiologically and economically is to control new , smaller foci through early detection , removal of inoculum , and host protection ahead of the epidemic front . These epidemiologically-based control insights should be linked to an understanding of how the viability of management actions is also shaped by pre-existing factors such as economic , social ( e . g . , patterns of land ownership , acceptability of specific treatment methods ) , and legal ( e . g . , state and federal permitting and environmental compliance ) constraints [48] , [68] . Our results suggest that it is possible to reduce inoculum and to contain the spread of P . ramorum , but also indicate that early and aggressive interventions alone might not achieve eradication of this pathogen . These findings are consistent with the epidemiological patterns observed in the northernmost focus of P . ramorum incidence in southwest Oregon , where aggressive treatments have contained but not eradicated the pathogen [41] , [48] . A new find of P . ramorum in northern California should allow us to test the approaches outlined in this paper . In spring 2010 , we detected an additional P . ramorum outbreak approximately 100 km north of the Redway infection site ( Valachovic et al . , unpublished ) . Although small ( ∼10 ha ) , the new site is extensive enough to suggest that it has been active for a number of years , but with a long cryptic-infection period . During fall 2010 and spring 2011 the majority of inoculum producing hosts was removed in and around this site . Our overall conclusions address several challenges about the management and control of emerging plant pathogens in heterogeneous host populations in natural landscapes . Large- scale dispersal , high local and regional sporulation , and a broad host range produce a host landscape with high connectivity that facilitate rapid and extensive invasion [8] . Our study demonstrates that many management actions are ineffective in achieving their stated goal of limiting pathogen spread , but also suggests that efforts to control emerging plant pathogens should be encouraged . Fragmentation of suitable habitat through disturbances such as logging , wildfire , disease , or disease control efforts , may lead to aggregated host distributions . Understanding how landscape structure influences invading species is critical to identifying appropriate management actions to reduce their impacts [4] . At the scale of the Humboldt study area ( Fig . 2 ) , the variation in host communities is not sufficient to limit the spread of P . ramorum in the long term . However , larger distances between patches of suitable habitat can reduce the likelihood of establishment of invasive species [8] . For example , across California there are regional variations in host availability and weather conditions responsible for spatial refugia , that could remain infection-free for many years despite the potential long-distance dispersal of P . ramorum [32] . In addition , disease management actions against emerging pathogens in Humboldt County and elsewhere are likely to be applied unevenly across the landscape because individual landowners assign different value to their forest resources . Further research is needed to understand how the spatiotemporal variation introduced by social dynamics would affect the impact of management treatments and pathogen spread . In other forests worldwide , where environmental conditions are less suitable for P . ramorum and related pathogens , the spatial arrangement of treatments could be particularly influential on the efficacy and cost effectiveness of pathogen management . We hope to have shown in this paper that the adoption of informed control measures is , at least , more likely to ameliorate local economic , ecological , and social impacts of disease , while making rational use of limited resources . Moreover , by linking disease control with management practices , it may even be possible to convert challenges into opportunities for shaping ecosystem composition and function for the benefit of communities and the environment .
|
We discuss principles governing the spread and management of diseases in natural forest ecosystems . Invasive organisms are damaging world forests and agricultural crops at an increasing rate and severity due to global trade and environmental disturbances . While prevention is the best option , practitioners must decide whether and how to act once a pathogen emerges in a new environment . But , how do we know that an invasion is occurring , its current extent , and its future spread ? Emerging pathogens are often observed too late because they are unknown or difficult to detect before causing damage . Once we detect the invader , what can we do to manage and hopefully eliminate it ? As many invasions occur rapidly on large geographic scales , small-scale affordable experimentation is not an option , so we need predictive models to gain insight into these questions . Invaders are more challenging if they disperse cryptically and over long distances . We study the case of sudden oak death in California , estimating where and how fast it is spreading , and showing that resources for management must be deployed rationally and early in order to succeed . A promising strategy is curative treatment at the core with preventive protection stretching far around the focus of the outbreak .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"public",
"health",
"and",
"epidemiology",
"forestry",
"infectious",
"disease",
"epidemiology",
"plant",
"biology",
"statistics",
"population",
"dynamics",
"spatial",
"epidemiology",
"integrated",
"control",
"plant",
"science",
"mathematics",
"pest",
"control",
"plant",
"pathology",
"population",
"biology",
"epidemiology",
"biology",
"nonlinear",
"dynamics",
"agriculture",
"spatial",
"and",
"landscape",
"ecology",
"ecology",
"disease",
"dynamics",
"plant",
"pathogens",
"statistical",
"methods"
] |
2012
|
Landscape Epidemiology and Control of Pathogens with Cryptic and Long-Distance Dispersal: Sudden Oak Death in Northern Californian Forests
|
The brassinosteroids ( BRs ) represent a class of phytohormones , which regulate numerous aspects of growth and development . Here , a det2-9 mutant defective in BR synthesis was identified from an EMS mutant screening for defects in root length , and was used to investigate the role of BR in root development in Arabidopsis . The det2-9 mutant displays a short-root phenotype , which is result from the reduced cell number in root meristem and decreased cell size in root maturation zone . Ethylene synthesis is highly increased in the det2-9 mutant compared with the wild type , resulting in the hyper-accumulation of ethylene and the consequent inhibition of root growth . The short-root phenotype of det2-9 was partially recovered in the det2-9/acs9 double mutant and det2-9/ein3/eil1-1 triple mutant which have defects either in ethylene synthesis or ethylene signaling , respectively . Exogenous application of BR showed that BRs either positively or negatively regulate ethylene biosynthesis in a concentration-dependent manner . Different from the BR induced ethylene biosynthesis through stabilizing ACSs stability , we found that the BR signaling transcription factors BES1 and BZR1 directly interacted with the promoters of ACS7 , ACS9 and ACS11 to repress their expression , indicating a native regulation mechanism under physiological levels of BR . In addition , the det2-9 mutant displayed over accumulated superoxide anions ( O2- ) compared with the wild-type control , and the increased O2- level was shown to contribute to the inhibition of root growth . The BR-modulated control over the accumulation of O2- acted via the peroxidase pathway rather than via the NADPH oxidase pathway . This study reveals an important mechanism by which the hormone cross-regulation between BRs and ethylene or/and ROS is involved in controlling root growth and development in Arabidopsis .
Roots are important plant ground organs , which absorb water and nutrients to control plant growth and development . In higher plants , root growth is maintained by coordinating cell proliferation and differentiation [1–3] . Plant hormones have been known to play a crucial role in the regulation of root growth [4] . Recent studies in the Arabidopsis root have shown that different hormones control organ growth by regulating specific growth processes such as cell proliferation , differentiation or expansion in distinct tissues . Plant hormones such as auxin , cytokinin , abscisic acid , brassinosteroids , ethylene and gibberellins have been shown to be involved in root growth through a range of complex interactions . The activities of these hormones during root growth progression depend on cellular context and exhibit either synergistic or antagonistic interactions . For example , ethylene enhances inhibition of root cell elongation through upregulating the expression of ASA1 and ASB1 to enhance auxin biosynthesis in Arabidopsis seedlings [5] . Furthermore , ethylene regulated root growth was also mediated through modulating the auxin transport machinery [6] . In addition , cytokinin was also found to control root growth through transcriptional regulation of the PIN genes and thus influencing auxin distribution [7] . The balance between auxin and cytokinin signaling is crucial during root growth . In Arabidopsis , cell division and cell differentiation largely determines root meristem size , which is under the control of cytokinin and auxin through an ARR1/SHY2/PIN circuit [1] . All these studies suggest that hormonal cross-talk plays a pivotal role in the regulation of root growth . The brassinosteroids ( BRs ) represent a class of phytohormones involved in a wide variety and developmental processes including root development [8–12] . BR , detected by the BRI1 receptor , activates the transcription factors BES1 and BZR1 , which in turn govern the transcription of a large number of genes [13–16] . BRs are known to participate in root growth and development , because mutants impaired with respect to either the synthesis or signaling of BR develop foreshortened roots [17 , 18] . However , excessive application of bioactive BR hampered normal development of plants [19] . Therefore , a finely tuned cellular regulation of BR levels is important for the development of plant . It has been found that BR deficient conditions elicit the expression of BR biosynthesis genes , while increase in endogenous BR concentration lead to feedback regulation of the expression of BR metabolic genes to maintain the homeostasis of BR [20] . Recent studies demonstrate that BR interacts with plant hormones such as abscisic acid , gibberellins , auxin and cytokinin to regulate plant growth and development [21–23] . BR interacts with ethylene to regulate the gravitropic response of the shoot , and is involved in ethylene-controlled processes in the hypocotyl of both light- and dark-grown seedlings [24 , 25] . Exogenously supplied BR enhances the stability of type 2 of the enzymes 1-aminocyclopropane-1-carboxylate synthase ( ACS5 and ACS9 ) and thus increasing ethylene production , thereby modulating the hypocotyl growth of etiolated seedlings [26] . Though both BRs and ethylene have been reported to regulate root growth and development , it is still unknown if there is a cross-regulation between BRs and ethylene in this process . In addition to plant hormones , the regulation of root growth has also been tightly linked to reactive oxygen species ( ROS ) . Root growth is profoundly affected by endogenously generated ROS . While ROS were initially believed to merely represent a damaging by-product of the plant’s stress response [27] , they have been now recognized as signaling molecules [28] . For example , ROS have been shown to be important for balancing cell proliferation and differentiation during root growth , and have been proposed to adopt a signaling role during lateral root formation [29 , 30] . It has been reported that ROS produced in mitochondria of root tip cells in response to the hormone abscisic acid ( ABA ) are responsible for regulating the root’s meristematic activity [31] . A BR receptor-mediated increase of the cytosolic concentration of calcium ions ( Ca2+ ) regulates ROS production , thereby reducing the length of the hypocotyl in dark-grown seedlings [32 , 33] . Though BRs have been reported to regulate many plant biotic and abiotic stresses through the regulation of ROS homeostasis [27 , 34] , the role of the cross-regulation between BRs and ROS in root growth is largely unknown . Here , the participation of BR in root growth and the extent of its cross-regulation with both ethylene and ROS signaling were investigated by characterizing a novel A . thaliana det2 mutant allele ( det2-9 ) selected on the basis of its short-root phenotype , which proved to be defective with respect to BR synthesis . A key observation was that the det2-9 mutant accumulated more ethylene and ROS than the wild type . The increased accumulations of both ethylene and ROS caused the short root phenotype in det2-9 . This study reveals a mechanism about how BRs regulate root growth through a cross-regulation with ethylene and ROS signaling .
To identify novel determinants involved in the control of root growth , an ethyl methane sulfonate ( EMS ) -mutagenized Arabidopsis population was screened by monitoring root length and elongation . One mutant was subsequently named as short root 5 ( sr5 ) ( Fig 1A and 1B ) . The length of the mutant root was only 23% of the one of a wild type ( WT ) seedling at 7–8 days post germination . A longitudinal zonation pattern analysis showed that the size of its root apical meristem ( RAM ) was significantly smaller than the WT control ( Fig 1C ) . Both meristem zone ( MZ ) and transition zone ( TZ ) in the RAM were substantially reduced in size . Cortical cells in the mutant mature zone were significantly shorter than those in WT , and cell number in the MZ was strongly reduced ( Fig 1D and 1E and Table 1 ) . The number of cells formed by the RAM in sr5 was 1 . 8 fold fewer than that in the WT control . The length of the mutant’s RAM was 67% of WT’s , and both its MZ and TZ were reduced in size ( Table 1 ) . The compromised RAM in the mutant was accompanied by an increased cell cycle time which displayed 1 . 4 times longer than that in the WT control ( Table 1 ) . The signal obtained from the mitotic cyclin B1;1 G2/M transition marker pCYCB1;1::GUS was much weaker in the mutant than the WT control ( Fig 1F ) , an indication that cell proliferation was inhibited in sr5 . The conclusion was that the mutant’s short root derived from both a reduced MZ cell number and a smaller cell size in the mature zone . When positional cloning was employed to identify the site of the sr5 mutation , a position on chromosome 2 flanked by the markers W20 and W22 was identified ( S1A Fig ) . Sequencing of the genes present in the critical genomic region revealed that the mutant had a point mutation causing a G-to-A transition at nucleotide position 107 after ATG in At2g38050 ( DET2 ) . The root growth and seedling morphology of the reported det2-1 mutant were indistinguishable from those of sr5 ( S1B and S1E Fig ) . Since the F1 hybrid sr5 x det2-1 retained the short-root phenotype ( S1C Fig ) , it was concluded that the sr5 mutation likely involved a lesion in DET2 . Moreover the DET2 promoter driving DET2 cDNA fused to GFP-GUS ( pDET2::DET2-GFP-GUS ) complemented the short-root phenotype in sr5 ( S1D Fig ) , suggesting that the G107A mutation in DET2 led to the short-root phenotype in the sr5 mutant . In seedlings carrying the transgene , GUS activity was detected both in the shoot and the RAM ( S2 Fig ) . DET2 encodes a steroid 5α-reductase acting in the BR synthesis pathway . The phenotype of sr5 , which was similar to that observed in det2-1 grown in darkness , was rescuable when the plants were treated with exogenous BR ( eBL ) ( S1B Fig ) . Since there are eight alleles of det2 mutant have been reported , we renamed the sr5 mutant as the det2-9 which was used for most of the analysis in this study . We compared the expression levels of some BR induced genes between det2-9 and det2-1 through Q-PCR analysis . The results showed that , though both mutants displayed reduced expression levels of TCH4 , BAS1 , IAA17 and IAA19 , the det2-9 mutant has a higher expression level of these BR-induced genes than the det2-1 mutant ( S3 Fig ) . Consistently , the det2-9 mutant had a weaker phenotype compared with the det2-1 mutant ( S1 Fig ) , indicating that the point mutation at position 107 might not be a null allele . A RNA-seq approach was applied to compare the det2-9 root transcriptome with that of the WT , a total of 1 , 480 and 1 , 116 genes were found to be , respectively , up- and down-regulated ( Fig 2A ) . Among the differentially expressed genes , based on the GO analysis we found that there is a statistically significant enrichment in genes annotated as being linked to secondary metabolic process and response to stimulus ( P<0 . 01 ) . It is not surprising for this enrichment considering the dwarf phenotype of the mutant . Though the previous research has shown that exogenously supplied BR can enhance the production of ethylene [26] , the ethylene biosynthesis and ethylene response factors were up-regulated in det2-9 according to our RNA-seq analysis ( Fig 2B ) . These genes included 1-aminocyclopropane-1-carboxylate synthase encoding genes , 1-aminocyclopropane-1-carboxylate oxidase encoding genes and ethylene response factor encoding genes ( S1 Dataset ) . According to our GO analysis , we also found that many of genes belong to GO:0000302 ( response to reactive oxygen species ) were up-regulated significantly in det2-9 ( Fig 2C ) , which was in contrast with the previous reports showing that BR could induce the generation of H2O2 [27 , 34] . To confirm the RNA-seq results , we performed a quantitative real time-PCR ( qRT-PCR ) assay on a selection of 20 ethylene related genes which were differentially transcribed in WT and det2-9 seedlings grown in light and dark growth conditions ( Fig 2D and S4 Fig ) . Though the expression changes of most of ethylene related genes were confirmed , we also found that some genes , for example ACS6 , ERF6 and ERF17 , had little agreement between the transcript abundance by RNA-seq and qRT-PCR analysis . Considering three independent repeats were done for the confirmations , the results of qRT-PCR analysis are more reliable . These results suggest that both ROS and ethylene signaling were enhanced in the det2-9 mutant . Ethylene signaling in the det2-9 mutant was monitored by the expression of the pEBS::GUS ethylene signaling reporter . The strength of the GUS signal was considerably higher in the mutant than that in the WT ( Fig 3A ) , suggesting that an enhanced level of ethylene signaling occurred in det2-9 . The increased ethylene response in det2-9 was abolished by the presence of 10 nM eBL during seedling growth ( Fig 3A ) . The ethylene content was considerably higher in the det2-9 mutant than that in WT seedlings ( Fig 3B and 3C ) . The transgene line pDET2::DET2-GFP-GUS/det2-9 complemented the higher level of ethylene observed in det2-9 ( Fig 3C ) . Treatment with the BR synthesis inhibitor propiconazole ( PPZ ) also resulted in higher ethylene content in light-grown WT seedlings , while eBL ( 10 nM , a concentration which partially rescued the short-root phenotype in det2-9 ) treated WT or bes1-D ( a mutant which displays an enhanced BR signaling response ) light-grown seedlings both showed a reduction in ethylene content ( Fig 3B ) . A similar profile of ethylene content was also observed when seedlings were grown in darkness ( S5 Fig ) . All these above chemical treatment experiments and mutant analysis suggest that both exogenous applied low levels of BR and native BR signaling negatively regulated ethylene biosynthesis . In addition , we also observed that root growth was inhibited gradually by eBL at concentrations ranging from 10 to 5000 nM ( Fig 4A and 4B ) . While the hypocotyl length was unchanged when treated with low concentration of eBL ( <500 nM ) but reduced sharply when the concentration of eBL greater than 500 nM ( Fig 4A ) . Furthermore , dark-grown seedling hypocotyls treated with higher concentration of eBL ( ≥500 nM ) displayed a typical “triple response” , indicating the enhanced ethylene response ( Fig 4A ) . Therefore , we further examined the effects of BR on ethylene production using different concentrations of eBL . The results showed that ethylene content was greatly reduced in seedlings treated with low concentration of eBL ( 10 or 100 nM ) while it was strongly increased when the concentration of eBL greater than 500 nM ( Fig 4D ) . Consistently , both GUS staining analysis with the pEBS::GUS transgene line and an examination of ethylene response factors ( ERFs ) expression using qRT-PCR analysis show that low concentrations ( 10–100 nM ) of BR inhibits ERF expression while high concentrations ( ≥500 nM ) of BR enhanced expression ( Fig 4C and 4E ) , consistent with the change ethylene levels . In summary , BR either positively or negatively regulate ethylene biosynthesis depends on the levels of BRs . When det2-9 mutant seedlings were grown on a medium containing either silver nitrate ( AgNO3 , an antagonist of ethylene signaling ) or 2-aminoethoxyvinyl glycine ( AVG ) ( an inhibitor of ethylene synthesis ) , the root growth of det2-9 mutant was partially rescued , producing root lengths almost double than those developed by the non-treated mutant seedlings . However , both treatments inhibited the root growth of WT seedlings ( Fig 5A and 5B ) . In addition inhibition of ethylene signaling by AgNO3 rescued the cortical cell length in det2-9 ( S6 Fig ) . On the other hand , the root cell elongation and root growth of the mutant seedlings was found to be more sensitive to ACC ( a precursor of ethylene synthesis ) ( Fig 5A and 5B and S6A Fig ) . In addition , both WT and det2-9 mutant seedlings displayed similar cell numbers in root meristem under the same treatment with either AgNO3 or ACC ( S6B Fig ) . This result suggests that the BR deficiency caused short-root phenotypes in det2-9 was mediated by the effect of ethylene signaling on root cell elongation . Consistently , the octuple acs mutant CS16651 ( acs2-1/acs4-1/acs5-2/acs6-1/acs7-1/acs9-1/amiRacs8acs11 ) , ein2-5 and the ein3/eil1-1 double mutant , which have defects in either ethylene biosynthesis or ethylene signaling , were less affected by the PPZ treatment than WT ( Fig 5C ) . The short-root phenotype of det2-9 was partially recovered in the det2-9/acs9 double mutant and det2-9/ein3/eil1-1 triple mutant ( Fig 5D–5G ) . These results indicate that the short-root phenotype in det2-9 partly result from enhanced ethylene biosynthesis and ethylene signaling . A promoter analysis showed that promoters of ACS6 , 7 , 9 , 11 , along with ACO1 and 3 ( all these genes were strongly up-regulated in det2-9 , Fig 2 ) contained a BRRE and/or an E-box , the binding sites for BES1 and BZR1 ( Fig 6A ) . The direct interaction of the ACSs by BES1 or BZR1 was confirmed by a chromatin immunoprecipitation ( ChIP ) /qPCR analysis in FLAG-tagged BES1 or YFP-tagged BZR1 transgenic lines ( Fig 6A ) . A series of yeast one-hybrid assays were conducted to further verify whether any of these promoters was regulated directly by either BES1 or BZR1 . The outcome was that in yeast BES1 interacted with the promoters of ACS7 and 9 , while BZR1 did so with the promoters of ACS9 and 11 ( Fig 6B ) . Neither of the two transcription factors interacted definitively with the ACS6 ( Fig 6B ) , ACO1 or the ACO3 promoter ( S7 Fig ) . The trans-activity of BES1 or BZR1 with the ACS promoters was further demonstrated in a transient dual LUC expression assay in A . thaliana mesophyll protoplasts . The over-expression of both BES1 and BZR1 strongly repressed the activity of ACS promoters ( Fig 6C ) , confirming that either BES1 or BZR1 can repress ACS7 , 9 and 11 gene expression in vivo . A previous study reported that short-term treatment with eBL resulted in dephosphorylation of BES1 ( its active form ) in WT and in det2-1 , but not in bri1-5 or bin2-1 signaling BR mutants [35] . Therefore we further investigate the expression of ACSs in BR signaling mutants including bri1-116 and bin2-1 . qRT-PCR results showed that the expression of ACS6 , 7 , 9 and 11 increased in the det2-9 ( see also Fig 2D ) , bri1-116 and bin2-1 mutant compared with WT ( S8 Fig ) . The expressions of these four ACS genes decreased when treated with eBL to a lower extent in bri1-116 or bin2-1 mutants compared with the det2-9 mutant ( S8 Fig ) . These results indicate that BR signaling pathway is required for the BR-mediated repression of ACS gene expression , via direct regulation by the BES1 and BZR1 transcription factors . The transcriptomic analysis in det2-9 mutant roots identified genes responding to ROS as BR-targets ( Fig 2C ) . Therefore , we further analyzed det2-9 mutant for defects in ROS using the nitroblue tetrazolium ( NBT ) staining method to detect the presence of O2- in vivo [36] . The NBT signal was higher in the det2-9 mutant than that in the WT control ( Fig 7A ) , while there was no clear difference when 3 , 3’-diaminobenzidine ( DAB ) staining was used to visualize the level of H2O2 present [37] ( S9 Fig ) . This suggests that det2-9 accumulated O2- but not H2O2 . Treatment with eBL substantially reduced the extent of the O2- hyper-accumulation in det2-9 ( Fig 7A ) . Meanwhile , the BR-signaling defective mutant bri1-116 hyper-accumulated O2- , while BR-signaling enhanced plants ( p35S::BRI1-GFP or bes1-D ) accumulated less superoxide anion in their roots compared with WT plants ( Fig 7B ) , indicating that BR signaling suppresses the accumulation of O2- . Therefore , root growth analysis was done using det2-9 mutant seedlings were exposed to two different O2- scavengers , namely superoxide dismutase ( SOD ) [38] and 1 , 3-dimethyl-2-thiourea ( DMTU ) [39] . The root length in det2-9 was significantly increased in the presence of 0 . 65U/ml SOD , while the same treatment inhibited root growth in WT seedlings ( Fig 7C ) . Similarly , a concentration of 0 . 1 to 2 mM DMTU treatment , which has no effect on root growth in WT seedlings , could significantly increase root lengths in det2-9 ( Fig 7D ) . The accumulation of O2- in det2-9 can be the result of the activation of two signaling pathways: peroxidase or NADPH oxidase . When the NADPH oxidase pathway was blocked by the presence of either diphenylene iodonium ( DPI ) [40] or ZnCl2 [41] , det2-9 mutant roots were insensitive to any treatment ( Fig 8A and 8B ) . Consistent with this result , the abundance of transcripts of the four NADPH oxidase genes ( RBOHC , D , F and G ) was identical in det2-9 and WT ( S10A Fig ) . The root growth response to PPZ treatment of three mutants rbohD , rbohF and rbohD/F was also similar to the one of WT ( S10B Fig ) . NBT staining showed that the BR deficiency-induced O2- hyper-accumulation by PPZ treatment was unaffected in both plants harboring p35S::NADPHD-GFP and the rbohD/F double mutant ( S10C Fig ) . And O2- hyper-accumulates in det2-9/rbohD and det2-9/rbohD/F mutants similarly to det2-9 , compared with WT ( S10D Fig ) . These experiments allow us to conclude that the hyper-accumulation of O2- in det2-9 did not involve the NADPH oxidase pathway . So attention was focused on the peroxidase pathway [29] , by treating seedlings with either salicylhydroxamic acid ( SHAM ) [42] or 1 , 10-phenanthroline ( 1 , 10-Phe ) [43] , inhibitors of peroxidase activity . The root length of det2-9 was significantly increased by both treatments , whereas root growth of WT was slightly inhibited ( S11A and S11B Fig ) . NBT staining showed that the levels of O2- in det2-9 reduced sharply when treated with SHAM or 1 , 10-Phe but no obvious changes were observed when treated with DPI or ZnCl2 ( Fig 8C ) , which was consistent with the NBT staining observed in det2-9/rbohD and det2-9/rbohD/F mutants compared with WT and det2-9 ( S10D Fig ) . When the transcription of genes encoding peroxidase was investigated , no clear-cut differences were visible between the mutant and WT ( S12 Fig ) , but peroxidase activity was much stronger in the det2-9 mutant and was reduced when seedlings were treated with exogenous BR ( Fig 8D ) . Thus the hyper-accumulation of O2- in det2-9 was likely the effects of an increased peroxidase activity . Given that the level of both ethylene and O2- was enhanced in det2-9 , the question arose as to whether ethylene and ROS interacted with one another . O2- accumulation was initially assayed in WT and det2-9 plants treated with either AVG or ACC ( Fig 9A ) . NBT staining showed that the ACC treatment had a positive and AVG had a negative effect on superoxide anion accumulation in WT roots ( Fig 9A ) . This indicates that ethylene induces an accumulation of O2- in Arabidopsis . NBT staining also showed that p35S::EIN3-GFP accumulated more O2- than WT , but when treated with PPZ , the extent of O2- accumulation was similar among p35S::EIN3-GFP , ein3/eil1-1 and WT ( S13 Fig ) , indicating that there was another pathway independent from ethylene participating in O2- accumulation when BR synthesis was blocked with PPZ treatment . In the det2-9 mutant , there was no clear increasement for ACC-induced superoxide anion accumulation , but the AVG treatment reduced it , which is also an indication that the increase in superoxide anion accumulation was at least partially dependent on ethylene production in det2-9 . Since peroxidase activity in det2-9 was higher than that in WT ( Fig 8D ) , an experiment was conducted to compare peroxidase activity in plants carrying p35S::EIN3-GFP , the ein3/eil1-1 double mutant and WT . The result showed that the peroxidase activity was not clearly affected in p35S:EIN3:GFP , ein3/eil1-1 compared with the wild-type control ( Fig 9B ) , indicating that ethylene signaling pathway is unlikely to activate the POD pathway for BR-regulated accumulation of O2- in det2-9 mutant . To investigate whether the O2- accumulation can alter normal ethylene signaling , we compared the expression level of pEBS::GUS when treated or not with methyl viologen ( MV , a superoxide anion propagator ) treatment . As shown in Fig 9C , the expression level of pEBS::GUS reporter was considerably induced when treated with MV , suggesting that an O2- accumulation can increase ethylene content . We then measured the primary root growth of ein2-5 , ein3/eil1-1 and wild type when treated or not with MV . Mutants in ethylene signaling were more resistant than WT to the negative effects of MV on root growth ( Fig 9D ) . The expression of genes encoding ACS and ACO , analyzed by qRT-PCR , increased when treated with MV ( S14 Fig ) . These results further indicated that O2- accumulation can alter normal ethylene production .
The BRs are well recognized as promoters of cell elongation , also in addition to their involvement in the de-etiolation response , where the opening of the apical hook is thought to require a decrease in the level of ethylene synthesis [17 , 44 , 45] . It was found that BR enhances ethylene production through the synergistic interaction with eto1 and eto3 [46] . Another study from the same lab showed that supplying BR exogenously promotes ethylene synthesis in the A . thaliana seedlings via stabilizing ACS5 and ACS9 protein [26] . This BR-induced ethylene production was also observed in mung bean and maize [47 , 48] . However , in jujube fruit , 5 μM BR-treated fruits caused a significantly lower level of ethylene during storage and the inhibition fruit ripening [49] . These contradictory results indicate the complicated effects of BR on ethylene synthesis . However , all these observations are based on the chemical treatment with BR . In this study , a new mutant allele of DET2 , det2-9 , was identified based on the short-root phenotype ( Fig 1 and S1 Fig ) . DET2 encodes a steroid 5α-reductase involved in BR biosynthesis , catalyzing the formation of campestanol with campesterol as substrates . Since another allele det2-1 and other mutants such as cpd and dwf4 which have defects in different steps of BR biosynthesis also displayed short-root phenotype [50–52] , it is unlikely that campesterol accumulation caused the short-root phenotype . The short-root phenotype is most likely a result of the reduced BRs synthesis in det2-9 since externally applied BRs could largely rescued the mutant phenotypes ( S1 Fig ) . Through genetic analysis and chemical treatment , we found that the short-root phenotype in det2-9 was partly the resulted of over accumulation of ethylene leading to enhanced ethylene signaling ( Figs 3 and 5 ) . The ethylene content was considerably higher in the det2-9 mutant than that in WT seedlings ( Fig 3B ) . Treatment with the BR synthesis inhibitor propiconazole ( PPZ ) also resulted in higher ethylene content in light-grown WT seedlings , while eBL ( 10 nM , a concentration which enhances the growth of root in det2-9 ) treated WT or bes1-D ( a mutant which displays an enhanced BR signaling response ) light-grown seedlings both showed a reduction in ethylene content ( Fig 3B ) . A similar profile of ethylene content was also observed when seedlings were grown in darkness ( S5 Fig ) . Transcriptional profiling showed that a number of ACS genes were up-regulated in the det2-9 mutant both in light and dark growth conditions , consistent with its increased level of ethylene ( Figs 2B , 2D and 3 and S4 and S5 Figs ) . Since the BR signaling transcription factors BZR1 or BES1 bind to ACS promoters to repress their expression ( Fig 6 ) , we analyzed BR signaling pathway by using bri1-116 or bin2-1 mutants and found that the BR-mediated down-regulation of ACS genes was greatly reduced in these two mutants compared with det2-9 ( S8 Fig ) , further indicating BZR1 or BES1 mediated BR signaling negatively regulates the expression ACS transcription factors . This result indicates that native physiological levels of BRs negatively regulate ethylene production through BZR1 or BES1 mediated transcriptional regulation of ACSs . Since our results and Zhu et al . ’s observations [49] are in contrast to other reports which showed that BRs enhanced ethylene biosynthesis [26 , 46–48] , we further did dosage-dependent assay to test the effects of BR on ethylene productions and root growth . Not surprisingly , root growth was inhibited gradually by eBL at concentrations ranging from 10 to 5000 nM ( Fig 4A and 4B ) . However , ethylene content was greatly reduced in seedlings treated with low concentration of eBL ( 10 or 100 nM ) while it was strongly increased when the concentration of eBL greater than 500 nM ( Fig 4D ) . Root growth analysis under both treatment suggests that both high and low levels of ethylene cause a short-root phenotype ( Fig 4D ) , which is consistent with the previous reports . In the octuple acs mutant ( CS16651 ) , which has only 10% ethylene level compared with WT , a reduced root growth phenotype was observed [53] . The acs9 mutant also displays a short root phenotype ( Fig 5E ) . The high levels of BR ( 500 nM or 1000 nM BR ) induced ethylene production is also consistent with the previous reports in Arabidopsis [26 , 46] . This study together with previous reports clearly showed that BRs either positively or negatively regulate ethylene biosynthesis in a concentration-dependent manner to control root growth . Certainly , since BR can also interact with other plant hormones such as auxin , ABA , cytokinin and jasmonic acid to regulate myriad aspects of plant growth and developmental processes in plants [54 , 55] , externally applied BR treatment caused root-growth phenotype might be also result from the interaction between BR and other plant hormones . ROS represent not only a by-product of stress response , but also influence growth and development in response to both internal developmental signals and external environmental cues [28] . The contrasting ROS status in the cell proliferation and the cell differentiation zones has recently been shown to be an important driver of root growth [29] . A mitochondria localized P-loop NTPase was also reported to regulate quiescent center cell division and distal stem cell identity through the regulation of ROS homeostasis in Arabidopsis root [56] . It has been pointed out that ABA-promoted ROS regulates root meristem activity [31] . In cucumber plants exposed to exogenous BR , H2O2 accumulates as a result of an increased activity of NADPH oxidase [27] , while in tomato , the same result is achieved by the up-regulation of RBOH1 [57] . BR has been documented as inducing a receptor-dependent increase in cytosolic Ca2+ , which stimulates NADPH oxidase-dependent ROS production [32 , 58] . Thus , although the participation of BR in root growth and development is accepted , its interaction with ROS signaling has not been systematically explored to date . Here , a key finding was that the det2-9 mutant hyper-accumulated O2- , which in itself likely contributed to the short root phenotype ( Fig 7 ) . BR inhibited the synthesis of O2- via the peroxidase ( Fig 8C and 8D and S11 Fig ) rather than via the NADPH oxidase ( Fig 8A and 8B and S10 Fig ) pathway . These results suggest that H2O2 and superoxide anion respond dissimilarly to BR in A . thaliana seedlings . While the level of H2O2 rises rapidly upon exposure to exogenous BR , the one of the superoxide anion is repressed . In addition , the hyper-accumulation of ethylene displayed by det2-9 contributed to a rise in the superoxide anion content in a peroxidase-independent manner ( Fig 9A and 9B ) . In summary , according to this study together with the previous reports , a proposed model was given in Fig 10 . We suggest that BR inhibits ethylene synthesis by activating the transcription factors BZR1 and BES1 under low levels . These transcription factors bind directly to the ACS promoters , thereby suppressing ACS expression and damping the level of ethylene synthesis under normal growth conditions . While high levels of BR induce ethylene biosynthesis either through increasing the stability of ACSs or influencing auxin signaling regulated ethylene production [47 , 59 , 60] . The possible regulation mechanism of BES1/BZR1’s activity under different levels of BR maybe refer to the regulation mechanism of ARF3 under different levels of auxin . Recent study has found that ARF3 acts as a repressor or activator depends on auxin concentration [61] . At the same time , BR inhibits the synthesis of O2- via the peroxidase pathway , but not NADPH oxidase pathway , which serves to regulate the growth of the A . thaliana seedling root . The accumulation of the O2- is also partially controlled by ethylene signaling in a peroxidase-independent manner and the O2- accumulation can enhance ethylene signaling by increasing the expression of ACSs and ACOs . Understanding how ethylene mediates BR signaling to control the accumulation of the O2- represents a logical follow-up research target .
All of the A . thaliana mutants and/or transgenic lines utilized are in a Col-0 background; the following have been described elsewhere: det2-1 [62] , bes1-D [13] , bri1-116 [63] , bin2-1 [64] , p35S::BRI1-GFP [65] , ein2-5 [66] , ein3/eil1-1 [67] , acs9 [68] , p35S::BZR1-YFP [69] , pNP::BES1-FLAG [70] , p35S::EIN3-GFP [71] , and p35S::NADPH-GFP [72] . And rbohD , rbohF , rbohD/F all described in Torres’ paper [73] . The octuple acs mutant ( CS16651 , acs2-1/acs4-1/acs5-2/acs6-1/acs7-1/acs9-1/amiRacs8acs11 ) [53] was obtained from the Arabidopsis Biological Resource Center ( ABRC , Columbus , OH , USA ) , and the marker lines pCYCB1;1::GUS [74] and pEBS::GUS [75] from early research . The 1501-bp upstream region from the DET2 start cordon and the cDNA of DET2 were amplified and linked to the GFP-GUS reported in gateway vector PKGWFS7 . 1 [76] to obtain pDET2::DET2-GFP-GUS reporter construct . Prior to germination , the seed was surface-sterilized by fumigation in chlorine gas , held for two days at 4°C on solidified half strength Murashige and Skoog ( MS ) medium , then transferred to a growth room providing a 16 h photoperiod and a constant temperature of 20°C . Root tips were imaged by laser-scanning confocal microscopy . The number ( obtained from a count of cells in the cortex file extending from the quiescent center to the TZ ) and length of cortical and mature epidermal cells were obtained from microscope images using ImageJ software . The criteria for defining the MZ and TZ were those described by Napsucialy-Mendivil et al . [77] . The cell production rate was based on the rate of root growth and the length of fully elongated cells , and the cell cycle time on cell production and the number of cells present in the MZ , as described by Napsucialy-Mendivil et al . [77] . The number of cells displaced from the cell proliferation domain ( Ntransit ) during a 24 h period was estimated from the equation Ntransit = ( 24 ln2 NMZ ) /T , where NMZ represents the number of cells in the RAM MZ and T means cell cycle time in hours . Histochemical GUS staining was performed according to the method described by Gonzalez-Garcia et al . [78] . The mapping population was the F2 generation of the cross sr5 x Landsberg erecta . Genomic DNA was extracted from each F2 seedlings showing the sr5 phenotype . Simple sequence length polymorphism markers were used for the initial genome-wide linkage analysis , following Lukowitz et al . [79] . To enable fine mapping , 22 PCR-based markers were designed to target the relevant region of the A . thaliana genome sequence . RNA was isolated from the roots of six day-old det2-9 and WT seedlings using the TRIzol reagent ( Invitrogen , Carlsbad , CA , USA ) and treated with DNase I to remove contaminating genomic DNA . The preparation was enriched for mRNA by introducing magnetic beads coated with oligo ( dT ) . The resulting mRNA was fragmented into fragments of about 200 nt , and the cDNA first strand was then synthesized via random hexamer priming . After synthesizing the second strand with DNA polymerase I , the ds cDNA was purified using magnetic beads coated with oligo ( dT ) and End reparation is then performed . Adaptors were then ligated to each end of the fragments , and the products were size-selected by gel electrophoresis . Finally , the fragments were amplified based on the adaptor sequences , purified using magnetic beads coated with oligo ( dT ) and dissolved in the appropriate amount of Epstein-Barr solution . The concentration and integrity of the ds cDNA was monitored using a 2100 Bioanaylzer device ( Agilent Technologies Japan Ltd . ) . The cDNA was then sequenced using an Ion Proton platform ( www . thermofisher . com ) . Low quality and adaptor sequences were removed and the remaining sequences were then aligned to the A . thaliana genome sequence using SOAP2 software . Individual transcript abundances were expressed in the form of the number of reads per kilobase per million reads ( RPKM ) , and differentially transcribed genes were identified using the thresholds FDR≤0 . 001 and |log2|≥1 [80] . The RNA template required for qRT-PCR was isolated using an RNeasy PlantMini kit ( Qiagen , Hilden , Germany ) following the manufacturer’s protocol . After treating with DNase I to remove contaminating genomic DNA , a 2 μg aliquot was reverse-transcribed using a Transcriptor First Strand cDNA Synthesis kit ( Roche , Basel , Switzerland ) , following the manufacturer’s protocol . The subsequent qRT-PCRs were run on a MyiQTM Real-time PCR Detection System ( Bio-Rad , Hercules , CA , USA ) using FastStart Universal SYBR Green Master mix ( Roche , Basel , Switzerland ) . Each sample was represented by three biological replicates , and each biological replicate by three technical replicates . The reference sequence was AtACTIN2 ( At3g18780 ) . Primer sequences are given in S2 Dataset . Ten seedlings were placed in a 100 mL vial containing 50 mL solidified half strength MS either with or without eBL or PPZ , and immediately capped . The vials were held under a 16 h photoperiod and a constant temperature of 20°C . After seven days , a 10 μL sample of the headspace was subjected to gas chromatography using a GC-6850 device equipped with a flame ionization detector ( Agilent Technologies Japan Ltd . ) . The coding sequences of BES1 and BZR1 were inserted separately into the EcoRI-XhoI cloning site of pGADT7 ( Takara , USA ) , while the promoter sequences of ACS6 , 7 , 9 , 11 , ACO1 and 3 were inserted into the cloning site of pAbAi . The primer sequences used in the construction of the various constructs are given in S2 Dataset . Each of the constructs ( including an empty vector for control purposes ) was transferred separately into yeast Y1HGold using the PEG/LiAc method . The yeast cells were plated onto SD/-Ura/-Leu medium containing various concentrations of Aureobosidin A to allow for a highly stringent screening of interactions . The procedure followed the manufacturer’s protocol given for the Matchmaker Gold Yeast One-Hybrid Library Screening System ( www . clontech . com ) . Ten day old transgenic plants were used for the ChIP assay following Gendrel et al . [81] . The quantity of precipitated DNA and input DNA was detected by qPCR . For each ACS promoter , primers were designed to amplify a fragment of length ~70–150 bp lying within the 2 kbp of sequence upstream of the transcription start site . The relevant primers are given in S2 Dataset . Enrichment was calculated from the ratio of bound sequence to input . The BES1 or BZR1 coding sequences were amplified and the resulting sequences introduced into pBI221 to place them under the control of the CaMV 35S promoter . The ACS promoter sequences were amplified and introduced into the pGreenII0800-LUC reporter vector . Both recombinant plasmids were then transferred into A . thaliana protoplasts . Firefly luciferase ( LUC ) and renillia luciferase ( REN ) activities were measured using the Dual-Luciferase Reporter Assay System ( www . promega . com ) . LUC activity was normalized against REN activity [82] . Details of all primers used are given in S2 Dataset . The roots of five day-old seedlings were immersed for 15 min in 2 mM NBT in 20 mM phosphate buffer ( pH 6 . 1 ) . The reaction was stopped by transferring the seedlings into distilled water . The material was then imaged under a light stereomicroscope . Tissue peroxidase activity was measured by a spectrophotometric analysis ( 420 nm ) of the formation of purpurogallin from pyrogallol in the presence of H2O2 . The roots of nine day-old seedlings were harvest and weighted . Tissue homogenate was prepared using 9 times phosphate buffer and then centrifuged for 10 min in 3500 rpm . The supernatant was used for peroxidase activity measurement . A single unit of enzyme was defined as the amount catalyzed and generated 1 μg pyrogallol by 1 . 0 mg fresh tissues in the reaction system at 37°C . Peroxidase activity was calculated from the formula provided with the peroxidase assay kit ( Jiancheng Bioengineering Institute , Nanjing , China ) . Sequence data for genes used in this study can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: DET2 ( At2g38050 ) , BES1 ( At1g19350 ) , BZR1 ( At1g75080 ) , ACS1 ( At3g61510 ) , ACS2 ( At1g01480 ) , ACS4 ( At2g22810 ) , ACS5 ( At5g65800 ) , ACS6 ( At4g11280 ) , ACS7 ( At4g26200 ) , ACS9 ( At3g49700 ) , ACS11 ( At4g08040 ) , WOX5 ( At3g11260 ) , ARF10 ( At2g28350 ) , ARF16 ( At4g30080 ) , ARR1 ( At3g16857 ) , SHY2 ( At1g04240 ) , BRI1 ( At4g39400 ) , CYCB1;1 ( At4g37490 ) , ACO1 ( At2g19590 ) , ACO2 ( At1g62380 ) , ACO3 ( At2g05710 ) , ACO4 ( At1g05010 ) , ACO5 ( At1g77330 ) , ERF6 ( At4g17490 ) , ERF13 ( At2g44840 ) , ERF17 ( At1g19210 ) , ERF104 ( At5g61600 ) , ERF105 ( At5g51190 ) , EBS ( At4g22140 ) , EIN3 ( At3g20770 ) , EIL1 ( At2g27050 ) , RBOHC ( At5g51060 ) , RBOHD ( At5g47910 ) , RBOHF ( At1g64060 ) , RBOHG ( At4g25090 ) , TCH4 ( At4g57560 ) , BAS1 ( At2g26710 ) , IAA17 ( At1g04250 ) , IAA19 ( At3g15540 ) , ACTIN2 ( At3g18780 ) .
|
Both brassinosteroids ( BRs ) and ethylene have been known to control root growth and development . ROS have been also reported to play an important role in root development . However , the relationship between BRs and ethylene or ROS in root growth and development was not addressed before . In this study , a det2-9 mutant defective in BR synthesis was identified from an EMS mutant screening , displaying a short-root phenotype which is result from the hyper-accumulation of ethylene and superoxide anions ( O2- ) . Exogenous BR apply showed that BRs either positively or negatively regulate ethylene biosynthesis in a concentration-dependent manner . Different from the BR induced ethylene biosynthesis through stabilizing ACSs stability , we found that the BR signaling transcription factors BES1 and BZR1 interacted with promoters of ACS7 , ACS9 and ACS11 to repress their expression , indicating a native regulation mechanism under physiological levels of BR . The BR-modulated control over the accumulation of O2- acted via the peroxidase pathway rather than via the NADPH oxidase pathway . This study provides new insights into how brassinosteroids control root growth through the cross-regulation with ethylene synthesis and ROS .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"peroxidases",
"chemical",
"compounds",
"plant",
"growth",
"and",
"development",
"ethylene",
"signaling",
"cascade",
"ethylene",
"enzymes",
"brassica",
"enzymology",
"organic",
"compounds",
"hormones",
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"plant",
"hormones",
"experimental",
"organism",
"systems",
"seedlings",
"plants",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"proteins",
"chemistry",
"biochemistry",
"plant",
"biochemistry",
"signal",
"transduction",
"eukaryota",
"plant",
"and",
"algal",
"models",
"organic",
"chemistry",
"root",
"growth",
"cell",
"biology",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"biosynthesis",
"physical",
"sciences",
"cell",
"signaling",
"organisms",
"signaling",
"cascades"
] |
2018
|
Brassinosteroids regulate root growth by controlling reactive oxygen species homeostasis and dual effect on ethylene synthesis in Arabidopsis
|
The community , the assemblage of organisms co-existing in a given space and time , has the potential to become one of the unifying concepts of biology , especially with the advent of high-throughput sequencing experiments that reveal genetic diversity exhaustively . In this spirit we show that a tool from community ecology , the Rank Abundance Distribution ( RAD ) , can be turned by the new MaxRank normalization method into a generic , expressive descriptor for quantitative comparison of communities in many areas of biology . To illustrate the versatility of the method , we analyze RADs from various generalized communities , i . e . assemblages of genetically diverse cells or organisms , including human B cells , gut microbiomes under antibiotic treatment and of different ages and countries of origin , and other human and environmental microbial communities . We show that normalized RADs enable novel quantitative approaches that help to understand structures and dynamics of complex generalized communities .
The community , i . e . the assemblage of organisms co-existing in a given space and time , is central to much of ecology [1] , and since Darwin’s “entangled bank” [2] one of the great challenges of biology is to explain the observed species diversity in communities mechanistically as a consequence of interactions and evolution . Modern experimental methods of high-throughput sequencing have brought us closer to complete inventories of community diversity . Moreover , these methods enable us to widen the scope of the community concept to generalized communities , that we define as assemblages of genomically diverse entities , which include , apart from communities in classical ecology , for instance B or T cell repertoires of the adaptive immune system , viral quasi-species , tumors , or human microbiomes . An intuitive description of a community composition is a table with columns species and abundance , possibly ordered from most to least abundant species ( we use the term species here in a loose sense for operational taxonomic units or other genomically distinct biological entities ) . A visually more accessible graphical representation of this table would be a plot that arranges the species along the horizontal axis and the abundances as vertical bars , sorted from highest to lowest bar . While such a plot is expressive for a specific community , it does not lend itself to quantitative comparisons between communities . To illustrate this point , consider a comparison of a community of South-American animal species with one of Sub-Saharan African animal species from regions with otherwise similar conditions . The two species columns of our table would have practically no overlap so that a direct comparison of these tables or plots is not possible . The same lack of overlap has to be expected for other generalized communities . For instance if we compare high-throughput sequencing data of B cell receptors of two persons , it is unlikely that there are receptors on mature B cells that occur in both persons . Nevertheless , it is a meaningful biological question whether the abundance structures of the two sets of B cells differ , e . g . whether the B cell repertoire is dominated by a few clones with high cell numbers , or whether it is distributed over many different clones with low cell numbers . A popular method for community comparison even in the absence of species overlap is to compute for each community a diversity index [3] , i . e . a single number that characterizes one aspect of the community , for instance the species richness , the evenness of the distribution , the Shannon entropy , or one of many related measures [4] , and then to compare the values of these indices between communities . The main disadvantage of this index approach is that it reduces a feature-rich abundance distribution to a single number , which may neglect important characteristics of that distribution . An alternative approach that had a major influence on the development of the theoretical foundations of modern ecology is to discard the species labels of the species-abundance table , which then becomes a so-called Species Abundance Distribution ( SAD; for excellent reviews see [5] , or [6] , chapter 9 ) . There are several established ways of presenting the information contained in a SAD ( see Fig 1 of [5] ) , for instance as straightforward histogram with species abundance as function of a species index , or as a binned histogram , typically with doubling bin widths with decreasing abundance . Here we focus on the Rank Abundance Distribution or RAD ( Fig 1 ) as SAD representation . RADs are simply vectors of species abundances sorted in decreasing order , usually visualized as two dimensional plots , possibly with one or both axes scaled logarithmically . In comparison to the mentioned simple and binned histograms , RADs are more smooth due to their component sorting [7] , and they retain the full biological resolution of the sampling experiment . The information content in RADs and probability distribution functions is the same , and one can be transformed into the other ( see e . g . [8 , 9] ) . Obviously , RADs retain the complete shape information of species-abundance tables . The abstraction of species information means that RADs enable analysis of generic abundance distribution features of generalized communities , independently of the actual species composition . RADs , and more generally SADs have been a key conceptual tool in the development and benchmarking of mechanistic models of ecological communities [10–15] . The mathematical functions resulting from mechanistic or statistical models , such as the log-normal distribution , were usually fitted to empirical RADs or SADs to identify the best community model . This basic research has paved the way for the application of these distributions to community comparisons , for instance to the characterization of community changes with changing environmental conditions ( see e . g . [16 , 17] . In these cases , the community comparison is typically a parameter comparison between fits of generic mathematical models to different RADs or SADs , or it is a purely visual comparison of these distributions [5] . In real-world samples , RADs are often not adequately described by a uniform mathematical model , e . g . a single log-normal distribution [18] . In macroecology , knowledge about the properties and relations between the observed animals and plants can be used to deconstruct multimodal distributions , and to fit simple models to fractions of the samples [3] . This knowledge is generally not available for high-throughput sequencing data of complex generalized communities , so that RAD analyses based on simple parametric models are difficult . This calls for a non-parametric approach . A further problem is that in practice the number of sampled species or sequences usually differs between samples . This means that RADs often have different dimensions and cannot be compared directly . It is possible to test for arbitrary pairs of RADs the null hypothesis that they originate from the same distribution using the Kolmogorov-Smirnov test [3 , 19] , but this is usually not helpful for quantitative comparisons . Technically , differences between a pair of RADs of different richness m , n could be also quantified by the Kolmogorov-Smirnov statistic D evaluated for the corresponding pair of cumulative distribution functions . However , it is problematic to interpret D in such cases since we are forced to equate non-reporting of |m − n| ranks in the shorter RAD with zero abundance , although the non-reporting may have technical reasons , e . g . limited sequencing depth . Thus , the question arises whether quantitative RAD comparisons are possible between samples of different richness . We quote from the widely cited SAD review by McGill et al . [5]: “How do we compare SADs ? Nearly all comparisons of SADs along gradients , deconstructions or time trajectories to date have been purely by visual inspection ( … ) . Most particularly , these visual inspections have been performed on rank-abundance plots which , by using an x-axis that runs from 1 to S ( i . e . species richness ) , seriously confounds the effects of species richness per se with other changes in the shape of the SAD ( … ) . Changes in species richness are a legitimate factor that should be considered a change in shape of the SAD . However , changes in richness so strongly dominate in rank-abundance plots that no other changes are easily considered . Is there any other change in the shape of an SAD after controlling for the fact that productivity affects richness ? We cannot say at the present time ( … ) More rigorous multivariate methods are needed . ” Here , we introduce MaxRank normalization of RADs , a new method that enables quantitative comparison of RADs , including their shapes . The approach is non-parametric and allows for the direct quantitative comparison of complex RADs without deconstruction and model fitting . An essential component of the method is the re-sampling of RADs up to a given richness . Consequently , the resulting normalized RADs ( NRADs ) are largely agnostic about the true richness of the original sample . The fact MaxRank normalization uses re-sampling may lead to conflation with rarefaction and rarefying ( on the distinction between the terms see [20] ) , two other techniques that also use re-sampling . Rarefaction [21] is typically used to estimate and compare richness between samples , i . e . exactly the quantity that is not of interest in MaxRank normalization . Rarefying ( e . g . [22] ) is applied to normalize OTU counts between samples . However , it is precisely the purpose of RADs and MaxRank normalization to abandon OTUs , and thus to make quantitative comparisons of abundance structures of different communities possible , irrespective of OTUs . We show here that results from the quantitative comparison of normalized RADs reflect biological differences between samples . To emphasize the versatility of the method , we have chosen a diverse set of high-throughput sequencing data representing different types of generalized communities , namely , human B cell receptor repertoires , and various human and environmental microbiomes . As one example of generalized communities we use human B cell receptor ( BCR ) repertoires . The diversity of BCRs in an individual is crucial for the recognition of antigens and the adaptive immune response [23] . Here we focus on the so-called heavy part of the receptor encoded by combinations of gene segments of the Ig heavy chain ( IGHV ) locus ( Fig 2A; [24] ) , and especially on the diversity of the “variable” VH segments in that part ( Fig 2B ) . Based on sequence homology , the VH segments are grouped into seven families ( VH1—VH7 ) , with members of a family having more than 80% sequence homology [25] . The sizes of the families vary from 1 ( VH6 family ) to 18–21 ( VH3 family ) . The primary repertoire of rearranged IGHV genes among naïve , antigen-inexperienced B cells ( Fig 2C ) typically encompasses all available VH segments , although abundances can vary considerably between VH gene segments . Exposure to antigens leads to a selective adaptation of the BCR repertoire that results in individual-specific sets of memory B cells ( Fig 2D ) . In the course of this complex maturation process , the usage of VH segments in BCR rearrangement repertoires may change . On top of this layer of complexity , several classes of BCRs with different biological functions , such as IgG or IgM , are generated by class switching . Depending on the chronological order of these different processes , and on the individual immune histories , we can expect more similar or more divergent IGHV gene diversity between receptor classes and individuals in memory B cells . We study this question with RADs computed from High-Throughput Sequencing ( HTSeq ) data . HTSeq technology is also transforming the study of microbial communities , because it allows us for the first time to see these complex assemblages in their full diversity [26 , 27] . For instance , we now start to see the diverse composition of human gut microbiomes , and we begin to understand the links between the human microbiome , health and disease [28] . However , the deluge of data makes us also aware of the need for new ways to analyze and model such complex systems , e . g . with methods developed in ecology [29] , such as RADs . We have selected three HTSeq data sets to demonstrate the potential and limitations of RADs for the analysis and the modeling of microbiomes: the considerable effect of antibiotics on gut microbiomes [30] , a large collection of gut microbiomes from countries where different life styles prevail [31] , and a diverse set of human and environmental microbiomes [32] . In these examples , we use RADs as an analytic tool to generate easily interpretable results , and as a basis for quantitative models .
In the work presented here we used high-throughput sequencing ( HTSeq ) amplicon data from four different sources to compute and analyze NRADs , as described in the following . For each dataset we followed the flowchart in Fig 3 . The last box of the flowchart Fig 3 indicates that sets of RADs normalized to a common R can be analyzed in numerous ways . In this article we used methods from three branches of data analysis: ordination , clustering , and classification . Since many ordination and clustering methods require a distance between the studied objects , we first describe how we computed distances between pairs of NRADs , and then the actual analysis methods .
With their highly diverse repertoire of antigen-binding receptors , human memory B cells are an example of what we have earlier termed generalized community . This diversity is achieved by a process that is only partly understood and currently subject of intense research [43 , 44]: it starts with the genetic recombination of triplets of specific VH , DH , and JH gene segments from genetic pools of these segments , followed by various mutation and selection steps , and eventually leads to distinct classes and sub-classes of memory B cells . We assume as a working hypothesis that all memory B cells underwent the same diversity generating process . If this is true , we should see a very similar VH gene rearrangement pattern ( Fig 2 ) in all sub-classes of memory B cells , leading to the same normalized RADs ( NRADs ) of memory B cells in all sub-classes . To test this hypothesis , we used HTSeq data of the immunoglobulin heavy-chain variable ( IGHV ) regions of four large memory B cell sub-classes , IgG+ , IgG+CD27+ , I g M o n l y + C D 27 + , IgM+IgD+CD27+ , from two donors . The IGHV regions derive from the genomic pool of 38–46 VH segments ( Fig 2A ) and have been modified by mutation and selection steps . If we interpret the original set of VH segments as the “species” to be ranked according to their abundances , we should under our working hypothesis see the same NRADs in all memory B cell sub-classes . To exclude distortions of abundances due to primer bias , we collapsed abundances of VH segments from the measured read numbers to the numbers of distinct VH sequence variants . Thus , in this case an abundance is the number of distinct sequences that originate from the same VH segment , and that have been diversified by somatic mutations and clonal expansions . Fig 4 summarizes the results . The non-normalized RADs ( top left of Fig 4 ) have similar , boomerang-like shapes , although direct comparisons is difficult since differences in abundance span more than one order of magnitude , and maximum ranks differ between 35 and 40 . For direct comparison we therefore normalized the RADs to a MaxRank R = 35 ( top right of Fig 4 ) . The resulting NRADs have overall very similar shapes , lending support to our working hypothesis of a common generation and selection process . However , there are notable features that differentiate between groups of NRADs . For instance at rank 1 the most diverse IGHV regions in IgM receptors of donor 1 ( green curves in Fig 4 ) are more abundant than all other rank 1 abundances . Conversely , for donor 2 the most diverse IGHV regions in IgM ( blue curves ) are the least abundant of all rank 1 abundances . Towards higher ranks , IgM abundances of both donors ( blue and green ) are more similar to each other . IgG receptors ( red and orange ) have more similar abundance structures throughout all ranks , and have stronger right tails than IgM receptors , indicating a more even VH segment diversity in IgG than IgM receptors . The differences between the NRAD curves are subtle , but they emerge clearly when we quantitatively analyze NRAD distances ( Eq 2 ) . In the hierarchical clustering tree of the distances ( Fig 4 ) we see three main clusters of NRADs , a big cluster of IgG+ and IgG+CD27+ NRADs on the right of the tree in red and orange , and two clusters of IgM related NRADs . These three clusters appear robustly , no matter whether the average linkage or the complete linkage criterion is used for clustering . As expected for replicates , ( A , B ) and ( C , D ) of the same sub-class coming from the same donor yield NRADs that are most similar and thus fall into the same lowest-level clusters . Beyond this , memory B cell sub-classes have remarkably different cluster structures . The big IgG+/IgG+CD27+ cluster ( red and orange ) of eight NRADs has a substructure of an IgG+ cluster and a separate IgG+CD27+ cluster , i . e . here the memory B cell sub-class has a stronger impact on the NRAD than inter-donor differences . This is different for the I g M o n l y + C D 27 + and IgM+IgD+CD27+ clusters . There , each of the two donors forms a cluster of its own that combines both I g M o n l y + C D 27 + and IgM+IgD+CD27+ NRADs , i . e . in these two IgM sub-classes , inter-personal differences in VH diversity patterns are stronger than differences between sub-classes . Our working hypothesis was that we have basically a single , diversity generating process for all memory B cell sub-classes , leading to the same NRADs for VH gene rearrangement pattern in all sub-classes . Overall , our results are consistent with this big picture since all NRADs are variants of the same boomerang shaped template . However , the quantitative analysis of NRAD distances picked up differences that require refinements of this model . This is not surprising since some of the steps that potentially affect VH rearrangement diversity , for instance class switching , cannot apply equally to all memory B cell sub-classes . What is surprising are the specific differences in VH rearrangement diversity between IgM and IgG sub-classes , e . g . the structure of the IgG cluster discussed above suggests that there could be significantly different selection pressures towards the final memory B cells in the two sub-classes IgG+ ( i . e . IgG+CD27− ) and IgG+CD27+ . Recently , we have used the same HTSeq data for a detailed sequence-based analysis of the clonal composition and genealogy of memory B cells [33] . Although our current NRAD-based analysis disregards much of the information used in [33] , results are consistent: First , the VH gene rearrangement composition is mostly very similar across the studied sub-classes of memory B cells , in agreement with a shared generation process [45] . Second , IgM+IgD+CD27+ and I g M o n l y + C D 27 + show almost the same VH gene rearrangement diversity , likely due to their clonal relatedness as described in [33] . Third , there are significant and consistent differences between IgG+CD27+ and IgG+CD27− with respect to mutation load in both donors , also in agreement with [46] or [47] . To conclude this section , we return to the conspicuous boomerang shape that is the template common to all B cell receptor RADs and NRADs ( Fig 4 ) . When testing for similarity to standard model distributions in ecology , we found that the broken stick distribution [13] is a good description for the NRADs of sub-class IgG+CD27+ ( Fig 5 ) . If included in the hierarchical clustering , the broken stick NRAD appears among the branches of the IgG+CD27+ sub-tree ( inset of Fig 5 ) . However , even for the other sub-classes , we cannot reject the broken stick distribution ( p-values from Kolmogorov-Smirnov tests between 1 . 0 and 0 . 87 with a median of 0 . 99 ) , though they deviate more than IgG+CD27+ . Note that the normalized broken stick distribution in Fig 5 has no free parameters and therefore has not been fitted . Several mechanisms in community ecology and elsewhere lead to broken stick RADs [13 , 48–50] . A simple explanation for the observed RADs of VH segment usage could be the following . Assume a fixed number n V H of VH segments in the genome , and a fixed total number Nt of all BCR sequence variants ( i . e . summed over all VH segments ) . The biological purpose of fixing Nt could be to provide a sufficient number of BCR variants to cover the typical antigen diversity . These two assumptions fix the average number N t / n V H of sequence variants per VH segment . If this is all we know , the Maximum Entropy principle states that the geometric distribution is the most parsimonious explanation fulfilling these requirements [51] . The geometric distribution is the discrete equivalent of the continuous exponential distribution , which generates the broken stick RAD [48] . Thus , our argument posits a random process that produces numbers of sequence variants per VH segment with a geometric distribution . In fact , sequence counts of most RADs are compatible with geometric distributions according to Kolmogorov-Smirnov tests at significance level 0 . 05 ( exception: sample D of IgM+IgD+CD27+ with p = 0 . 04 ) . Given that there is good agreement between the RADs of the BCR sub-classes ( Fig 4 ) , and also good agreement in the usage of individual VH segments between human donors [33] , our argument suggests the following two testable hypotheses . First , the random mechanism leading to the broken stick RAD could be encoded in the human genome and conserved among individuals with intact immune systems . Second , since our argument is generic , we should see the broken stick RAD also in other species with similar BCR rearrangement mechanisms . Dethlefsen et al . [30] reported the effects of a short course of Ciprofloxacin ( Cp ) treatment on the gut microbiomes of three healthy human individuals . When comparing gut microbiomes prior to treatment and during treatment , they found markedly perturbed taxonomic composition , richness , diversity , and evenness . These perturbations varied between individuals . After treatment , the community compositions recovered within four weeks to states close to pre-treatment , though with some species lost . We tested whether the dynamics of perturbation and recovery is reflected by changes of NRADs . The NRADs fall into two clusters , one well-defined “off-Cp” cluster of NRADs before and after treatment ( blue and green in Fig 6 ) , and one wider “on-Cp” cluster during treatment ( red in Fig 6 ) . All on-CP NRADs have heavier heads and less weight in the tails , consistent with the decreased gut microbiome diversity under treatment discovered by [30] . After normalization , we could compute distances between all pairs of NRADs and apply multidimensional scaling ( MDS ) to the distance matrix . The MDS plot Fig 6B captures the abundance dynamics from the well-defined off-Cp cluster on the right to the wider on-Cp cluster on the left and back again to the off-Cp cluster on the right . We usually find in MDS analyses a strong correlation of the first coordinate with Shannon entropy . Hence , the dynamics in the MDS plot Fig 6B from right ( pre-Cp ) to left ( Cp ) to right ( post-Cp ) corresponds to a succession of high-low-high entropy . This can also be seen directly from the NRADs: the off-Cp NRADs have a relatively heavy tail and weak head , i . e . a more even distribution with higher entropy , corresponding to a more diverse gut microbiome . Conversely , the on-Cp NRADs have a more heavy head and weaker tail , i . e . a less even distribution with lower entropy , corresponding to a less diverse gut microbiome , partly decimated by the effect of the antibiotic . Dethlefsen et al . [30] remarked that after treatment several taxa failed to recover , while the participants in the study had normal intestinal function , and they argued that the eliminated taxa after treatment may have been replaced by other taxa with similar functions . The fact that pre- and post-Cp NRAD ensembles have the same shapes and form a single compact off-Cp cluster ( Fig 6 ) supports this assessment . It is instructive to compare our analysis based on the abundance structure with an OTU composition analysis as in Fig 6 of Ref [30] . The OTU based PCA in Ref [30] has a cluster structure that is influenced by both the individual microbiome donor and by the treatment state . The conflation of both influences makes the result of the PCA richer but also more difficult to interpret: If we consider individual microbiome compositions , all three individuals have different microbiome dynamics under treatment . Conversely , the NRAD based analysis is blind to individual differences in microbiome composition . This blindness to composition means on the other hand to focus on the abundance structure , which makes the result in our Fig 6 more easy to interpret: In terms of the abundance structure , all three individuals behave in the same way , clearly showing a generic effect of the antibiotic treatment . Yatsunenko et al . [31] found in 528 gut microbiomes from Malawi , United States and Venezuela , that ( 1 ) species richness of gut microbiomes increased with age from birth to about the third year , and then was much less variable , and ( 2 ) taxonomic composition of adult gut microbiomes from the Unites States differed strongly from those of Malawi and Venezuela , while the latter two showed less pronounced differences . In our analysis we do neither use richness ( we normalize to a common richness ) , nor taxonomic labels , but we use solely the abundance vectors ( Eq 1 ) as quantitative descriptors of NRAD shapes . Nevertheless , we will in the following show results consistent with key results from [31] with NRADs . Additionally , we will present a novel NRAD-based quantitative model for the development of gut microbiome entropy as function of age . Prior to normalization , the richness of the samples covered three orders of magnitude , from 4105 to 296214 different ranks . All 528 RADs were normalized to the same MaxRank of R = 4105 to make RAD shapes and quantities computed thereof comparable . Analyses with NRADs are complementary to taxonomy-based analyses: NRADs are blind to taxonomy , which is both a limitation and an advantage . It is a limitation because community biology is a function of taxonomic composition . It is an advantage because it enables quantitative comparison between taxonomically different generalized communities and thus discovery of generic community biology that is independent of actual taxonomic composition . A dataset where these aspects can be explored is the GlobalPattern dataset of Caporaso et al . [32] with 26 samples from human microbiomes , various environments , and mock communities . We transformed OTU tables to RADs and normalized them to MaxRank R = 2067 , the minimum richness in the set , and we computed a distance matrix of all pairs of NRADs . Hierarchical clustering of the distance matrix ( dendrogram in Fig 10A ) led to close clustering of some samples that also form taxonomic clusters [32] , namely the creek samples and a lake sample , the mock communities , or the human tongue and most feces communities . The NRADs in these clusters have low distances and thus are similar ( e . g . Fig 10B ) . Other relationships are more unexpected . The taxonomy-based analysis by [32] clusters tongue and palm microbiomes together and clearly separates human microbiomes from environmental samples . Conversely , NRADs of microbiomes of human palms do cluster closer with environmental samples than with microbiomes of tongues or feces . For instance , the NRAD palm1 does not cluster with the microbiome tongue1 of the same individual , but most closely with the sediment samples . NRADs of both palm1 and sediment2 have a lower abundance of rank 1 than tongue1 but more heavy tails ( Fig 10C ) , were palm1 and sediment2 fit almost perfectly . Reasons for this unexpected clustering are not known . It could be that microbiomes in tongue and feces are more strictly controlled by the host body and the microbiome itself , while microbiomes on palms are more exposed to the environment . In the GlobalPatterns dataset , entropies HR range from 2 . 67 ( tongue2 ) to 6 . 58 ( soil2 ) with a median of 4 . 10 and a standard deviation of 1 . 06 , and one may be tempted to explain the observed clustering by different entropies . However , such a simple approach is not successful here . For instance , the three NRADs in Fig 10D have almost the same entropies ( H R c r e e k 3 = 3 . 98 , H R m o c k 2 = 3 . 95 , H R t o n g u e 1 = 3 . 91 ) , but have completely different shapes , and are members of different clusters . In general , the rich information content of an NRAD cannot be reduced to a single scalar quantity . In the introduction we have mentioned that a theoretical possibility to quantitatively compare pairs of RADs of different richness ( a strength of MaxRank normalization ) is the use of the Kolmogorov-Smirnov statistic D , without RAD normalization . This corresponds to using the Chebyshev distance between the corresponding two cumulative distribution functions with different support . Although it is obvious that this approach treats the tail regions in a problematic way , it is unclear whether this problem is of practical relevance . To test this , we treated D as a distance and reran the hierarchical clustering with this distance . In fact , this D-based tree recovers some of the features of Fig 10A , but in general shows less biologically meaningful clusters ( S1 Fig ) . We conclude that a D-based analysis is in practice no alternative to MaxRank normalization . As mentioned in the introduction , one commonly used procedure for quantitative RAD comparisons is a parametric approach . Typically , a standard distribution such as the log-normal is fitted to a set of RADs and the comparison is then reduced to a comparison of the fitted parameters . We have tested the viability of this approach by fitting to the RADs of all GlobalPatterns samples five standard distributions , broken stick as null model , preemption , log-normal , Zipf , and Mandelbrot [41] . For most samples , with the exception of some soil and sediment samples , we found clear qualitative deviations from all five fitted distributions ( S2 Fig ) , or , in other words , only few fitted distributions reflected the actual RAD shapes . This means that this parametric approach is in general not an option for quantitative analysis of HTSeq data . In contrast , with our non-parametric approach we can quantitatively compare all information-rich NRADs in a consistent and detailed way . The key feature of MaxRank normalization is that it enables quantitative RAD comparisons by mapping RADs with diverse richness values to NRADs of a common richness R . It is clear that the lower R , the less information can be carried by NRADs . This could make analyses based on NRADs sensitive to R . To test this , we have tested how the choice of R affects NRAD based classification , NRAD distances , and NRAD entropies , as used in the previous sections . First , we address the question whether NRAD based classification is sensitive to R using the classification of gut microbiomes described earlier . We use the same random forest classification and threefold cross validation as before to classify NRADs of gut microbiomes of individuals older than 3 years into MV ( Malawi/Venezuela ) or US ( United States ) . The results for R = 4105 , 1000 , and 250 are summarized in Table 2 . The table shows a weak reduction in accuracy and κ statistic with decreasing R and increasing errors , both explainable by a loss of information stored in the NRADs with lower R . However , the reduction of R from 4105 to 1000 affects the accuracy of the classifier barely , and even the classifier computed from NRADs with R = 250 has still a good accuracy . The weak dependency is consistent with the importance plot ( Fig 7B ) where the regions of high importance cover two extended rank intervals that can be mapped down to NRADs with R = 250 or lower , though with some losses . Secondly , we test the effect of R reduction on NRAD distances for the GlobalPatterns data set . Fig 11A demonstrates that NRAD distances ( Eq 2 ) are well-conserved even if R is an order of magnitude lower than the richness of the original RADs . Approximately halving R from R = 2067 ( the maximum possible R in the GlobalPatterns set ) to R = 1000 does not affect distances between NRADs: the points are very close to the diagonal ( coefficient of determination r2 = 1 . 000 ) . If we reduce R more drastically to 250 ( green points in Fig 11A ) , deviations appear , but we still have r2 = 0 . 976 . The small deviations of NRAD distances in Fig 11A for R = 250 are all towards smaller NRAD distances ( points shifted to the left of the diagonal ) , because lowering R from 2067 to 250 means that for R = 250 we straighten some of the fine-grained structure of NRADs with R = 2067 that allows for larger NRAD distances . Even for R = 100 we still have r2 = 0 . 913 . In Fig 11B we show for comparison how reduction of R affects NRAD distances for a simpler normalization scheme in which ranks above a given R are cut off ( “cutoff normalization” ) , i . e . information in the tails with ranks higher than R is completely neglected . The scatter of the points is generally wider for all three values of R with r2 = 0 . 995 , 0 . 874 and 0 . 614 , respectively . These lower r2 values point to the importance of the tails that are neglected by cutoff normalization . NRAD pairs with differences predominantly in the neglected tails appear in Fig 11B shifted to the left of the diagonal , while NRAD pairs with more divergent heads and more similar tails appear right shifted . Cutoff normalization clearly is less robust than MaxRank normalization with respect to the choice of R . As a consequence of the robustness of NRAD distances , the down-stream analyses that make use of NRAD distances are also quite robust . For instance , the hierarchical clustering ( Fig 10A ) of the GlobalPatterns dataset is identical between R = 250 , R = 1000 , and R = 2067 up to an agglomeration height of 0 . 98 ( vertical axis in Fig 10A ) . Cluster assignments differ only above this agglomeration height at the highest branching points and thus at the most fuzzy cluster level . Thirdly , we find for each sample i a regular and systematic R dependence of NRAD entropy H R ( i ) ( Eq 5 ) : H R ( i ) / log R ≈ c ( i ) . ( 7 ) with a sample dependent constant c ( i ) . This approximation works reasonably well if R is not varied by more than an order of magnitude . Eq ( 7 ) corresponds formally to the definition of Shannon evenness with R interpreted as richness ( Eq 5 ) , i . e . Shannon evenness changes weakly with R . Moreover , this equation means that entropy or information content will systematically decrease with decreasing R . This decrease will depend weakly on R since log R changes only slowly with R . As the NRAD entropies change systematically with R , this must also affect our model of entropy of gut microbiome NRADs as function of age . For instance if we approximately halve R from 4105 to 2000 and repeat the fitting of the model Eq ( 6 ) we arrive at H 2000 0 , M V = 3 . 34 ± 0 . 16 , H 2000 0 , U S = 3 . 35 ± 0 . 19 , λ 2000 M V = 0 . 70 ± 0 . 09 y r - 1 , λ 2000 U S = 1 . 18 ± 0 . 27 y r - 1 , H 2000 m a x , M V = 5 . 62 ± 0 . 06 , and H 2000 m a x , U S = 5 . 12 ± 0 . 03 . Entropic model parameters H2000 are systematically lower by a small amount than the corresponding H4105 values ( Table 1 ) as expected from an approximate scaling ( Eq 7 ) . If all entropic factors in Eq ( 6 ) scale in the same way , the scaling factor cancels , and the exponential growth term e−λt should stay the same . In fact , the growth parameters λ are almost unaffected by the decrease of R from 4105 to 2000 . Finally , one important aspect of robustness is the following . If we take several samples of different size and therefore different richness of the same , well-mixed generalized community , then determine the RADs of those samples , and finally normalize these RADs to the same R , we should obtain the same NRAD for all samples . Only if this requirement is met can NRADs inform reliably about a generalized community . To test fulfillment of this requirement , we first down-sampled original HTSeq data sets by an order of magnitude in richness . Fig 12A shows as an example the original RAD and the down-sampled RAD of the first sample in the gut microbiome data of [31] ( MG-RAST ID 4 . 4899263e6 ) . We then normalized both to the same MaxRank R = 1000 ( Fig 12B ) . For all samples we found that both RADs , the original and the down-sampled , led to practically the same NRAD . The violin plot Fig 12C makes a quantitative statement about this property: the biologically relevant distance distributions between NRADs are the same for both the original and down-sampled data ( left and middle violin in Fig 12C are equal ) . In comparison , the distances between corresponding pairs of NRADs of original and down-sampled RADs are negligible ( right violin in Fig 12C ) as it should be for a normalization procedure that is robust against the size of the source sample . In summary , we found that NRADs are robust quantitative descriptors of RADs . However , it is also clear that there are critical values of R below which essential RAD structures are lost and analyses become inconclusive . These critical values will depend on the studied RADs and on the analysis method . We recommend to monitor NRAD quality with methods suitable for the respective question . For instance , if the cluster structure of a set of NRADs is of interest , clustering and ordination methods as those presented above can be used to detect loss of structure when results for several R values are compared .
MaxRank normalization makes communities of different richness quantitatively comparable by mapping their RADs to NRADs of a common richness R . This is similar to projecting objects of different higher dimensions to one common lower dimension where the projections can be compared directly . The price to be paid is information loss , especially loss of richness information . However , the remaining information enables new approaches to community analysis . In particular , we could show that the information extracted from NRADs is sufficient to generate quantitative models of the dynamics or composition of generalized communities . MaxRank normalization is not the only possible algorithm that maps higher richness RADs to a common lower richness , but it has crucial advantages over other procedures . For instance , another procedure is to cut off in all RADs ranks beyond a common maximum rank R . Obviously , this simple procedure neglects information in the RAD tails , leading to lower robustness as shown earlier . An alternative would be to scale the rank axis to the same maximum number . This would lead to fractional pseudo-ranks; the corresponding pseudo-rank abundance vectors have different dimensions and thus cannot be compared directly . Another alternative is the coarse-graining of the rank axis to a given number of pseudo-ranks . This is also not satisfying as the coarse-graining would treat samples of different richness differently , and because the pseudo-ranks are not observables . Conversely , MaxRank normalization is conceptually attractive since it corresponds to a real experimental sampling process with the attainment of a given richness as stop criterion . In fact , a similar observational protocol ( “m-species list” ) , i . e . sampling up to a constant maximum rank , has been used in field ecology [53] . MaxRank normalization does not impose a specific model , i . e . the approach is model-free and generally applicable . For instance , we have applied it not only to HTSeq data but also to data from conventional ecological sampling of macro-invertebrates from fresh-water systems ( unpublished ) . MaxRank normalization ( re-sampling up to a given maximum rank ) may be confused with rarefying ( re-sampling up to a given maximum number of individuals ) [20 , 22] . However , the two methods answer different diversity questions: rarefying allows answering questions that relate to sample richness ( and typically also to OTU abundance ) , while NRAD comparison largely eliminates richness ( and OTUs ) and puts emphasis on abundance structure difference . There are a number of limitations of MaxRank normalization . First , it is obvious that most information about richness is lost . Interestingly , it is not completely lost , but partially encoded in the NRADs , as a detailed technical analysis shows ( to be published elsewhere ) . Nevertheless , if mainly changes or differences in richness are of interest , NRADs are not suitable . Second , and related to the previous limitation , it is possible that some of the samples to be compared are richness-limited while others are not . In this case the information loss mentioned above can turn the method useless . To illustrate this point , imagine two systems that should be compared , one with two species , the other with thousands of species . In this case we would normalize the system to R = 2 and lose almost all the information in the RAD of the richer system . This limitation is usually not serious for the analysis of HTSeq data of generalized communities such as microbiomes . One strategy to cope with outlier samples of extremely low richness that would enforce the use of a low R and severe loss of information in NRADs is to discard such outliers , thus sacrificing some breadth for higher accuracy . Third , a more severe problem with HTSeq data is the often implied assumption that read counts are quantitative measures of abundances . Unfortunately , HTSeq data can be biased by the experimental protocol , e . g . by preferential PCR amplification of certain species and non-amplification of others , or it can contain false positives , e . g . error mutants or chimeras produced in the experimental process [54 , 55] . Since RADs are derived from OTU tables , any abundance bias that affects OTU tables will also affect RADs . Although these problems do not limit applicability of MaxRank normalization , they do limit the possible biological interpretation of the results . Related to this point , we found that changes in HTSeq protocols can significantly impact RADs and therefore NRADs . Thus , a stringent control is required if results from different studies are to be compared . Such a control can be implemented e . g . by comparing RADs resulting from the application of the different protocols to common reference samples . As HTSeq technology develops rapidly , some of these experimental problems may be solved in the near future . But even with these imperfections , NRADs can be used as generic quantitative descriptors to discover new community biology . Fourth , analyses based on NRADs alone are blind to taxonomic composition . This can be an advantage because in this way generic effects that influence the abundance structure can become clearly visible . But this blindness to taxonomy makes NRAD based analyses inadequate if differences or changes in taxonomic composition are of major interest .
|
Living things are parts of complex communities , similar to humans living in cities . A quantitative way of describing such communities is to measure the abundance of each species in the community so that a sorted list of abundance numbers is produced , a so-called Rank Abundance Distribution ( RAD ) . With recent breakthroughs in genome analysis this approach can also be applied to very complex communities , such as the community of the myriads of microbes in a human gut ( gut microbiome ) , or the diverse set of human immune cells . One problem with this approach is that it is not trivial to quantitatively compare RADs for different communities , especially if they are highly complex . We show that it is possible to computationally “normalize” RADs so that they can be quantitatively compared across many different communities . In this way , this normalization enables insight into structures and dynamics of arbitrary communities . We demonstrate this with applications to human immune cells , gut microbiomes under antibiotic treatment or under different nutritional regimes , and environmental microbiomes .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"taxonomy",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"immune",
"cells",
"microbiome",
"immunology",
"microbiology",
"data",
"management",
"b",
"cell",
"receptors",
"memory",
"b",
"cells",
"thermodynamics",
"microbial",
"genomics",
"ecological",
"metrics",
"immune",
"system",
"proteins",
"computer",
"and",
"information",
"sciences",
"white",
"blood",
"cells",
"entropy",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"microbial",
"taxonomy",
"species",
"diversity",
"physics",
"community",
"ecology",
"antibody-producing",
"cells",
"biochemistry",
"signal",
"transduction",
"cell",
"biology",
"b",
"cells",
"ecology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"genomics",
"immune",
"receptors"
] |
2017
|
Quantitative Comparison of Abundance Structures of Generalized Communities: From B-Cell Receptor Repertoires to Microbiomes
|
The covalently closed circular DNA ( cccDNA ) of the hepatitis B virus ( HBV ) plays an essential role in chronic hepatitis . The cellular repair system is proposed to convert cytoplasmic nucleocapsid ( NC ) DNA ( partially double-stranded DNA ) into cccDNA in the nucleus . Recently , antiviral cytidine deaminases , AID/APOBEC proteins , were shown to generate uracil residues in the NC-DNA through deamination , resulting in cytidine-to-uracil ( C-to-U ) hypermutation of the viral genome . We investigated whether uracil residues in hepadnavirus DNA were excised by uracil-DNA glycosylase ( UNG ) , a host factor for base excision repair ( BER ) . When UNG activity was inhibited by the expression of the UNG inhibitory protein ( UGI ) , hypermutation of NC-DNA induced by either APOBEC3G or interferon treatment was enhanced in a human hepatocyte cell line . To assess the effect of UNG on the cccDNA viral intermediate , we used the duck HBV ( DHBV ) replication model . Sequence analyses of DHBV DNAs showed that cccDNA accumulated G-to-A or C-to-T mutations in APOBEC3G-expressing cells , and this was extensively enhanced by UNG inhibition . The cccDNA hypermutation generated many premature stop codons in the P gene . UNG inhibition also enhanced the APOBEC3G-mediated suppression of viral replication , including reduction of NC-DNA , pre-C mRNA , and secreted viral particle-associated DNA in prolonged culture . Enhancement of APOBEC3G-mediated suppression by UNG inhibition was not observed when the catalytic site of APOBEC3G was mutated . Transfection experiments of recloned cccDNAs revealed that the combination of UNG inhibition and APOBEC3G expression reduced the replication ability of cccDNA . Taken together , these data indicate that UNG excises uracil residues from the viral genome during or after cccDNA formation in the nucleus and imply that BER pathway activities decrease the antiviral effect of APOBEC3-mediated hypermutation .
The hepatitis B virus ( HBV ) is one of the major causative factors of liver cirrhosis and hepatocellular carcinoma . Chronic inflammation due to persistent HBV infection plays a major causative role in these severe liver diseases . However , it is still unknown how HBV establishes persistent infection and how this infection results in these diseases [1] , [2] . The HBV genome in virions forms a relaxed circular DNA ( rcDNA ) that is converted into covalently closed circular DNA ( cccDNA ) in the nuclei of infected hepatocytes . The cccDNA transcribes all viral RNAs including pregenomic ( pg ) RNA as a replicative RNA intermediate . In the cytoplasm , pgRNA , viral core , and polymerase proteins are assembled into the nucleocapsid ( NC ) , after which the pgRNA is converted into an rcDNA by viral polymerase activity . The mature NCs are transferred to either the endoplasmic reticulum to be secreted after combining with envelope proteins or the nucleus to form cccDNA again for the next replication cycle . Although the host repair system is thought to play a major role in conversion of rcDNA into cccDNA , the molecules responsible for the conversion have not been determined experimentally [3] , [4] , [5] . cccDNA plays a key role in the persistence of viral infection because it is maintained as a stable episome in the nucleus . Moreover , cccDNA is not targeted by anti-HBV drugs and thus enables the re-establishment of viral replication after cessation of antiviral therapy [4] , [6] , [7] . Despite the importance of cccDNA in HBV chronic infection , host factors that control cccDNA are poorly understood in the absence of an efficient experimental system that can produce HBV cccDNA at a level sufficient for analysis . In view of the limitation of HBV in vitro systems , the duck HBV ( DHBV ) model has been commonly used to study HBV infection [8] . DHBV is an avian counterpart of HBV , sharing fundamental features including genomic organization , replication processes , and biological characteristics [9] . Importantly , DHBV produces cccDNA more efficiently than HBV [10] . Previously , we isolated a B-cell-specific gene , activation-induced cytidine deaminase ( AID ) , which is essential for class-switch recombination and somatic hypermutation of immunoglobulin genes [11] , [12] . AID belongs to the APOBEC ( apolipoprotein B mRNA editing catalytic polypeptide ) family of proteins . In humans , this family comprises at least 11 members , including AID and APOBECs 1 , 2 , 3A , 3B , 3C , 3DE , 3F , 3G , 3H , and 4 . These AID/APOBEC proteins have enzyme activity that can deaminate a cytidine base in DNA and/or RNA and thereby generate a uridine base . APOBEC3G ( A3G ) restricts replication of retroviruses , including human immunodeficiency virus type 1 ( HIV-1 ) , and retrotransposable elements [13] , [14] , [15] . A3G has been shown to also restrict other viruses such as HBV [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , but the exact mechanism of restriction of HBV replication remains unresolved . Earlier studies suggested that accumulation of extensive G-to-A ( in an opposite strand of C-to-U ) hypermutation in retroviral DNA might initiate the deamination-mediated restriction pathway of A3G [25] , [26] . However , such hypermutation may not account for rapid reduction of HBV NC-DNA by A3G overexpression because only a limited fraction of NC-DNA accumulates these extensive mutations [16] , [19] , [27] , [28] . Another proposed mechanism is that A3G is encapsidated within NC with pgRNA and interferes in the process of minus-stranded DNA synthesis , such that a catalytically inactive mutant of A3G has been shown to still inhibit viral replication [19] , [21] . Deaminase-independent restriction by A3G has also been demonstrated for HIV-1 [14] , [15] . Human uracil DNA glycosylase ( UNG ) is a base excision repair ( BER ) enzyme that removes uracil residues from DNA following dUTP misincorporations or cytosine deaminations [29] . UNG is also essential for class-switch recombination and somatic hypermutation . In UNG deficiency ( human and mouse ) , class-switch recombination is markedly reduced because uracil bases generated by AID do not produce DNA strand breaks in the absence of UNG . However , UNG-deficient mice and patients accumulate more frequent C-to-T and G-to-A somatic hypermutations because uracil bases generated by AID remain as thymine residues in UNG-deficient condition [29] , [30] . The UNG gene encodes 2 alternative splicing isoforms with unique N-terminal amino acid sequences , mitochondrial type UNG1 and nuclear type UNG2 . Early HIV-1 studies showed that UNG2 is encapsidated into the virion through physical association with the Vpr protein and reduces the mutation rate of viral DNA [31] , [32] . However , the contribution of UNG activity to APOBEC3 ( A3 ) -mediated HIV-1 restriction is controversial . Yang et al . proposed that uracil residues generated by A3G might be eliminated by UNG associated with Vpr and that subsequent DNA cleavage of abasic sites by apurinic/apyrimidinic endonuclease-1 ( APE-1 ) might occur [33] . Meanwhile , other groups reported that A3G-mediated HIV-1 restriction occurs even in the absence of UNG [34] , [35] , [36] . Importantly , HBV does not harbor the Vpr counterpart , and UNG encapsidation in HBV particles has not been reported . Unlike HIV-1 , HBV forms episomal cccDNA , which is a potential target for nuclear UNG activity; however , whether APOBECs can hypermutate the cccDNA and whether UNG has any role in cccDNA maintenance has not been investigated . In the present study , we investigated the possible role of UNG in A3G-mediated antiviral activities on HBV and DHBV . When UNG activity was inhibited by expression of the UNG inhibitory protein ( UGI ) , hypermutation of HBV and DHBV NC-DNA was enhanced in A3G-expressing hepatocytes . We found that more than half of DHBV cccDNA clones accumulated extensive hypermutation by A3G overexpression and UNG inhibition . Moreover , we demonstrated that the cccDNA isolated from cells expressing both A3G and UGI showed decrease in replication activity . These experimental observations indicate that UNG efficiently repairs dysfunctional C-to-U mutations induced by A3G in cccDNA .
To investigate the potential role of UNG in HBV hypermutation , the UNG activity of the human hepatocyte cell line HepG2 was suppressed with UGI , which is an irreversible inhibitor that forms an exceptionally stable complex with the UNG protein [37] . We generated the HepG2 cell line that stably expressed the UGI–estrogen receptor ( ER ) protein by retrovirus-mediated gene transduction following drug selection . We had previously demonstrated that the addition of an ER ligand , 4-hydroxytamoxifen ( OHT ) , to the culture medium activated the enzymatic activity of the fusion partner [38] , [39] . Accordingly , we reasoned that expression of the UGI–ER fusion protein could be applied to control UNG activity in HBV-replicating cells . The UNG assay revealed that OHT stimulation of the UGI–ER protein resulted in very limited UNG activity , amounting to 7% of the activity in the parental HepG2 cells , whereas 97% of activity remained in unstimulated ( EtOH ) cells ( Figure 1A ) . Using this cell line , we estimated the effects of UNG inhibition on A3G-induced hypermutation of HBV NC-DNA . An HBV replicon plasmid , pHBV1 . 5 [40] , [41] , was cotransfected with the FLAG-tagged A3G expression plasmid [or green fluorescent protein ( GFP ) as a negative control] into the UGI–ER HepG2 cells . Three days after transfection , the cells were harvested and cytoplasmic NCs were purified . The NC-DNA was analyzed by differential DNA denaturation polymerase chain reaction ( 3D-PCR ) on a region of the X gene [28] . The 3D-PCR technique is a highly sensitive assay for detecting AT-rich DNA . It was applied to reveal the presence of hypermutation in NC-DNA . Consistent with reports from other groups [18] , [19] , [23] , [24] , [28] , [42] , [43] , A3G expression induced HBV hypermutation , represented as a lower denaturation temperature band ( 83 . 9°C ) than negative controls in 3D-PCR ( Figure 1B ) . OHT addition enhanced the hypermutation in the A3G-expressing cells because the band was detected at the lowest melting temperature ( 83 . 0°C ) . OHT addition did not influence A3G transgene expression in these cells ( Figure 1C ) . The 83 . 9°C PCR products shown in Figure 1B were cloned and sequenced . As predicted by the 3D-PCR assay , the cloned PCR fragments accumulated extensive G-to-A mutations ( Figure 1D ) . We also suppressed UNG activity using a short-interfering RNA ( siRNA ) approach to avoid any artifacts due to the UGI–ER inducible activation system . We used 293T cells for the siRNA experiment because of better transfection efficiency of the siRNA than that afforded by HepG2 cells . For viral replication in human embryonic kidney 293T cells , another replicon plasmid pPB that expresses HBV pgRNA by the CMV promoter was used [44] , [45] . The UNG assay revealed that both UNG-specific siRNAs reduced UNG activity in 293T cells , although at low suppression efficiency ( a maximum of 47% of the control; Figure S1A ) . Nonetheless , 3D-PCR showed amplification at a slightly lower denaturation temperature , indicating the presence of hypermutated HBV DNA from the UNG-specific siRNA-treated cells ( Figure S1B ) . These data indicate that inhibition of UNG activity increases A3G-induced hypermutation of HBV NC-DNA . Next , we investigated whether the enhancement of hypermutation by UNG inhibition was reproduced by endogenous AID/APOBEC3 proteins . We generated a stable HepG2 cell line that constitutively supports both HBV replication and UGI–ER expression in order to establish a transfection-free system ( see Materials & Methods for details ) . Inhibition of UNG activity by OHT addition in this cell line was confirmed by the UNG assay ( Figure 2A ) . IFNγ was used to stimulate the cells to induce endogenous APOBEC deaminases , and changes in deaminase gene expression levels were measured by quantitative reverse transcription-PCR ( qRT-PCR ) . Consistent with that in previous studies [20] , [46] , [47] , A3G was the major responder to IFNγ stimulation among AID/APOBEC3s ( A3s ) in this cell line ( Figure 2B ) . The hypermutation load on NC-DNA in IFNγ-stimulated UGI–ER HepG2 was analyzed by 3D-PCR . Our analyses revealed that UNG inhibition enhanced IFNγ-induced NC-DNA hypermutation ( Figure 2C ) . To evaluate the contributions of endogenous APOBEC3G , we used A3G-specific siRNAs . The efficiency of siRNA knockdown was determined by qRT-PCR ( Figure S1C ) . As shown in Figure 2D , the knockdown of A3G expression counteracted the induction of hypermutation by IFNγ , suggesting that A3G is responsible for HBV hypermutation induced by IFNγ stimulation . The 87 . 2°C PCR products shown in Figure 1D were cloned and sequenced , and confirmed the hypermutation ( Figure S1D ) . These data suggest that UNG counteracts the deamination of HBV NC-DNA triggered by endogenous A3s . To determine the overall hypermutation frequency of HBV DNA , we sequenced the NC-DNA from the A3G-transfected UGI–ER HepG2 . PCR fragments of the X gene partial sequence ( 94°C for denaturation ) of NC-DNA were cloned in a T vector . Fifty clones of each sample were randomly selected for DNA sequencing . Figure 3A shows the mutations found in the sequenced clones . Consistent with 3D-PCR results , the total G-to-A mutation frequency was enhanced by UNG inhibition in A3G transfectants ( indicated as “A3G , OHT” in Figure 3 ) . However , 39 of the 50 sequenced clones were free from hypermutation in both UNG-inhibited and uninhibited cells ( Figure 3B ) . Previous studies [16] , [19] , [27] , [28] also reported few clones harboring A3G-induced mutations in NC-DNA . Since UNG inhibition enhanced hypermutation , we next investigated whether UNG inhibition affects another antiviral activity of A3G , the suppression of NC-DNA production . Cytoplasmic HBV NC-DNA was quantified by native agarose gel electrophoresis ( NAGE ) followed by Southern blotting . NAGE specifically separates intact NC particles , and after the NC particles are transferred to a nylon filter , the DNA content in the NC fraction is measured by hybridization analysis with the HBV DNA probe [21] , [22] , [48] . As shown in Figure 4A , Southern blotting after NAGE revealed that A3G reduced the NC-DNA content , a result consistent with previous reports [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] . We found that the signal intensity of NC-DNA band from A3G-UGI cotransfectants were slightly higher than that from A3G transfectants in HepG2 cells . However , A3G-mediated reduction was not disturbed by UNG inhibition in Huh7 cells . Quantitative PCR ( qPCR ) analyses of the NC-DNA revealed no appreciable difference between the presence and absence of UGI in HepG2 cells ( Figure 4B ) . We concluded that UNG did not affect A3G-mediated reduction of NC-DNA , although enhanced hypermutation was observed in the same experimental culture conditions for 3 days . To observe possible encapsidation of the UNG protein in the HBV NCs , we evaluated the physical association between NC and UNG by immunoprecipitation ( IP ) . Nuclear type of UNG ( UNG2 ) was overexpressed together with the HBV plasmid pPB and FLAG-A3G expression vector in 293T cells . The cytoplasmic fraction containing NCs was used for IP with anti-core ( HBc ) antibody and Western blotting . As shown in Figure S2A , endogenous mitochondrial UNG ( UNG1 ) was clearly detected in input from the cytoplasmic fractions . Leaked UNG2 from the nucleus was also detected in the input . As reported previously [18] , the A3G protein was immunoprecipitated with anti-HBc antibody . However , the signals for UNG2 and GFP disappeared after IP , suggesting that the A3G protein but not the UNG protein is encapsidated into NC . Faint signals for UNG1 were still observed in all lanes of the IP samples at equal signal strength , suggesting nonspecific capture of the UNG1 protein on the protein G sepharose beads . We also investigated the subcellular localization of the UNG2 protein in human hepatocytes . GFP-fused UNG2 localized in the nucleus even in HBV-replicating HepG2 cells ( Figure S2B ) . Taken together , these findings did not reveal evidence for the NC-associated UNG protein , although the NC UNG protein level below detection sensitivity may be sufficient to change the NC-DNA hypermutation frequency . We next investigated whether the nuclear viral intermediate , cccDNA , is a target of UNG activity . Since analysis of cccDNA has been difficult using our HBV in vitro model because of the low abundance of cccDNA ( data not shown ) , we exploited the DHBV replication system , which efficiently produces cccDNA for analysis [9] , [10] . In this study , the surface-deficient DHBV replicon plasmid pCSD3 . 5ΔS was used to transfect a chicken hepatocyte cell line , LMH , because deficiency of surface protein leads to accumulate more cccDNA than wild-type DHBV [10] , [49] . The UNG assay indicated that UGI transfection resulted in efficient decrease in UNG activity even in the LMH cell ( Figure 5A ) . Inhibition of chicken UNG by UGI has been reported previously [50] , [51] . Western blotting confirmed that the expression levels of A3G and catalytically inactive mutant ( mutA3G ) transgenes were not influenced by UGI ( Figure 5B ) . The DHBV plasmid was cotransfected with A3G and UGI vectors into LMH cells . After 3 days , the cells were harvested and NC-DNA was analyzed by 3D-PCR [23] ( primer position is indicated in Figure S3A ) . Data indicate that amplification occurred at the lowest temperature ( 83°C ) from A3G transfectants both with and without UGI expression ( Figure 5C ) , indicating that the A3G protein can hypermutate DHBV NC-DNA . NC-DNA fragments amplified by a standard PCR ( 94°C ) were cloned into the T vector and 10 clones were randomly selected for DNA sequencing . These sequences are shown in Figure 5D ( indicated as “NC-DNA” ) and Figure S3B . Consistent with other HBV experiments ( Figure 3 ) , these data demonstrate the enhanced hypermutation of NC-DNA by UNG inhibition . Next , cccDNA was isolated by nuclear Hirt extraction and further treated with DpnI to digest transfected plasmids . A cccDNA fragment ( 1 . 4 kb ) was amplified by standard PCR ( 94°C ) using cccDNA-selective primers spanning the gap region of rcDNA ( Figure S3A ) [52] , [53] . We tested the amplification efficiency of DNAs from nuclear Hirt extract , secreted viral particle , and the replicon plasmid to verify the specificity of cccDNA-selective PCR ( Figure S3C ) . As expected , the 1 . 4 kb fragment was predominantly amplified from the nuclear Hirt extract containing cccDNA . We also evaluated the specificity of the cccDNA-selective PCR with a replication-defected DHBV replicon plasmid ( pCSD3 . 5ΔP ) . The cccDNA-selective PCR amplified the 1 . 4 kb fragment from transfectants of the replication-competent plasmid but not from those of replication-defective plasmid ( pCSD3 . 5ΔP ) ( Figure S3D ) . The result clearly demonstrates that our cccDNA-selective PCR amplifies the 1 . 4 kb from nuclear viral DNA but not from replicon plasmid . We compared the mutation frequency of cccDNA with those of NC-DNA and pre-C mRNA . Ten randomly selected clones from the A3G-UGI cotransfectants and A3G alone were sequenced , and the mutation frequencies were compared ( Figure 5D ) . Surprisingly , cccDNA clones from the A3G-UGI cotransfectants were hypermutated much more extensively than NC-DNA . Increase in the G-to-A mutation frequency was observed not only in the mutation load per clone but also in the number of clones harboring hypermutation . Eight of 10 cccDNA clones from A3G-UGI cotransfectants carried at least 1 G-to-A/C-to-T mutation , whereas hypermutation was much less frequent on the cccDNA derived from the A3G transfectants . Interestingly , enhanced hypermutations by A3G-UGI were also observed in cDNA clones derived from pre-C mRNA ( Figure 5D ) . The pre-C mRNA is transcribed from cccDNA but not from the replicon plasmid . Overlapping sequences of 2 PCR-amplified regions for NC-DNA and cccDNA ( Figure 5E ) showed that hypermutation was distributed throughout the P gene and that its distribution patterns were similar between samples . Extensive G-to-A hypermutation by A3G and UGI in this region was confirmed . As expected , mutations were biased toward the GpG dinucleotide , a preferential target of A3G ( the underlined nucleotide represents a mutation position ) [28] . Considering the dinucleotide preference , the in-frame TGG codon must be susceptible to nonsense mutations ( TGA , TAG , and TAA ) . In this sequence analysis , 6 of 10 clones of cccDNA from A3G-UGI cotransfectants had premature stop codons in the viral P gene open reading frame ( ORF ) , suggesting an effect on downregulation of viral replication . To estimate the overall mutation frequency of the cccDNA , 2 . 9-kb PCR fragments corresponding to 98% of the whole viral genome were amplified by standard PCR ( 94°C ) and cloned in the T vector . Seven clones were randomly selected from each group and sequenced . Mutation matrices of cccDNA are shown in Figure 5F . As expected from Figure 5D , much more frequent G-to-A mutations were detected in the cccDNA from A3G-UGI cotransfectants than in those from other samples . We observed similar mutation frequencies between the C and P genes ( data not shown ) . Only a few mutations ( 2 mutations in 7920 nt ) in the neomycin-resistant gene of the transfected plasmid ( Figure S3E ) were detected . The results indicate that cccDNA was extensively hypermutated , while the transfected plasmid was not , in A3G-UGI cotransfectants . Since nuclear cccDNA was highly mutated by A3G and UGI coexpression , we investigated the subcellular localization of A3G in LMH cells . Microscopic observation revealed that the majority of GFP–A3G fusion proteins were localized in the cytoplasm and that the nuclear A3G protein was not obvious even in DHBV-replicating LMH cells ( Figure S4 ) . The data indicate that even the A3G protein in LMH cells localizes in the cytoplasm in a manner similar to that in mammal cells [18] , [54] , [55] , [56] and that the extensive hypermutation in cccDNA is not due to misregulation of A3G intracellular localization . To assess the role of UNG in the A3G-induced suppression of DHBV replication , cytoplasmic and nuclear viral DNAs were analyzed by Southern blotting at days 3 and 6 after transfection . A3G-induced suppression was obvious in all samples of NC-DNAs and cccDNAs from both days ( Figure 6A and B ) . The suppression occurred in a deaminase-dependent manner , given that the mutant A3G did not reduce the DHBV DNA levels . Similar to the HBV result in Figure 4 , UNG inhibition by UGI did not affect the A3G-mediated reduction of NC-DNA and cccDNA at day 3 . However , at day 6 , UGI expression enhanced the NC-DNA reduction of A3G-expressing cells , while the cccDNA level of the same transfectants was slightly increased ( Figure 6B , lanes 3 and 6 ) . We also performed qPCR to analyze the secreted virion DNA levels from day 2 to 6 after transfection of wild-type DHBV , A3G , and UGI vectors ( Figure 6C and D ) . Culture supernatant was collected daily and viral particles were precipitated by polyethylene glycol ( PEG ) precipitation . Purified DNA from the precipitants was treated with DpnI to digest any contaminating plasmids . In comparison with A3G alone , secreted viral particle-associated DNA levels in cells cotransfected with A3G and UGI were not significantly different at days 2 and 3 . However , consistent with the cytoplasmic Southern blotting data , at day 5 , the level of secreted viral DNA in A3G and UGI cotransfectants was lower than that of A3G transfectants ( Figure 6C and D ) . We also analyzed pre-C mRNA expression levels in day 5 samples . Since pre-C mRNA is transcribed from cccDNA but not from the replicon plasmid , pre-C mRNA expression reflects functional activity of the upstream viral intermediate , cccDNA . As shown in Figure 6E , qRT-PCR showed a consistent reduction with result of Figure 6D . To evaluate the outcome of cccDNA hypermutation by A3G expression and UNG inhibition , we used rolling circle amplification ( RCA ) , a method widely used to prove the presence of covalently closed circular DNA , including cccDNA of HBV and episomes of human papillomavirus [57] , [58] ( Figure 7A and S5 ) . cccDNAs were purified from cells 7 days after transfection and then treated with DpnI . The cccDNA was amplified by RCA and digested with EcoRI to cleave the concatemer into individual full-length viral genomes ( 3 . 0 kb , Figure 7A ) . The samples containing cccDNA show 3 . 0-kb bands ( Figure 7A , lanes 1–4 ) , whereas a control RCA reaction of the DHBV plasmid ( lane 5 ) shows both 3 . 0-kb and 4 . 7-kb bands . Amplification of the 3 . 0-kb band without the 4 . 7-kb DNA ( Figure 7A , lanes 1–4 ) indicates specific amplification of cccDNA but not of the replicon plasmid . The 3 . 0-kb DNAs were cloned into the replicon plasmid backbone to reconstruct the DHBV plasmids . Reconstructed 20 clones from each sample were pooled and transfected into LMH cells to measure viral replication activity . Importantly , this secondary transfection was performed without A3G and UGI expression vectors . At day 3 after transfection , cytoplasmic NC-DNA was quantified by qPCR ( Figure 7B ) . The reconstructed plasmid with cccDNA from A3G-UGI transfectants showed a significant decrease in NC-DNA production . Consistent with this data , sequence analysis of the P gene in the reconstructed DHBV plasmid revealed that extensive G-to-A hypermutation had accumulated in the RCA products from A3G-UGI cotransfectants ( Figure 7C ) . The data shown in Figure 7B and C may underscore the real impact of hypermutation , given that only 23 . 8% ( 720 bp ) of the full viral genome per clone was sequenced and that any destructive mutations on the DHBV promoter region would have been rescued by the CMV promoter provided by the backbone in the reconstructed plasmids ( Figure S5 ) . Considering all the data , we concluded that nuclear UNG activity repaired uracil bases in cccDNA that were generated by the action of A3s .
To avoid the mutagenic impact of dUTP misincorporation or cytosine deamination , organisms have dUTPase and uracil DNA glycosylases , including UNG . Some viruses such as poxviruses ( vaccinia viruses ) or herpesviruses ( HSV-1 and cytomegaloviruses ) also encode UNG homologs , and primate lentiviruses incorporate host UNG into the virion through interaction with the viral Vpr protein [29] , [32] , [59] . However , the involvement of uracil excision activity during replication and infection of these viruses has not been fully investigated . The effect of Vpr-bound UNG on the deaminated HIV-1 DNA is apparently controversial , although it is thought to be involved in DNA repair [31] , [32] or DNA degradation [33] or to have no role [34] , [35] , [36] . Thus , it seems that there is no unified view whether UNG plays a protective or suppressive role in viral replication . In this study , using in vitro models of HBV and DHBV , we investigated the role of UNG activity in hypermutation and viral replication in the presence of A3G . We found that UNG inhibition resulted in the enhancement of A3G-induced NC-DNA hypermutation . This study for the first time also found that the A3G protein induced cccDNA hypermutation ( Figures 5 and 7 ) . The cccDNA hypermutation was enhanced on UNG inhibition and subsequently resulted in significant decrease in viral production ( Figure 7 ) , suggesting a protective role of UNG for viral replication against cccDNA hypermutation . It has been difficult to determine which HBV intermediates are catalyzed by UNG because HBV shuttles between the cytoplasm and the nucleus during its life cycle . Plasmids can also act as a target molecule . Stenglein et al . showed that APOBEC3A induces C-to-U hypermutation in transfected plasmids and that uracilated plasmids were processed by UNG activity using UGI-expressing 293T cells [60] . In the present study , APOBEC3A was not expressed in HepG2 and LMH cells ( Figure 2B ) . The contaminated DHBV plasmid does not contribute to the results of cccDNA sequencing ( Figures 7A and S3C–E ) . Comparison of hypermutation frequencies of NC-DNA , cccDNA , and pre-C mRNA from the same cell source revealed that the mutation frequency of cccDNA was higher than that of NC-DNA . G-to-A mutations were predominant in cccDNA clones ( Figure 5 ) , whereas C-to-T mutation is a characteristic feature of plasmid hypermutation [60] . The transfected vector backbone did not accumulate hypermutation ( Figure S3E ) . Accordingly , we concluded that A3G does not target the replicon plasmid and that the mechanism of cccDNA hypermutation differs from that of plasmid hypermutation [60] . The observed G-to-A hypermutation is one of the prominent features of cccDNA hypermutation . Predominant G-to-A hypermutation in NC-DNA ( Figures 5 and S3B ) , a precursor of cccDNA , may partly account for the G-to-A cccDNA hypermutation . However , it does not explain the higher mutation frequency of exclusive G-to-A hypermutation in cccDNA than that in NC-DNA . Further study is required to clarify the mechanism of cccDNA hypermutation . A3G is known to be encapsidated into NC and deaminate NC-DNA . We propose that an additional deamination event by A3G may occur during or after cccDNA formation in the nucleus . Although the nuclear localization of GFP-A3G was not obvious in DHBV-replicating LMH cells ( Figure S4 ) , the encapsidated A3G may be able to enter the nucleus in the same manner as viral rcDNA . Indeed , a recent HIV-1 study showed that infection with Vif-deficient HIV-1 led to uracil accumulation in the host genome , implying that Vif-sensitive A3s , including A3G , can also deaminate nuclear DNA [61] . We found that A3G and UGI coexpression caused extensive nuclear cccDNA hypermutation ( Figures 5 and 7 ) . Neither a UNG nor a Vpr homolog has been identified in HBV , unlike the HIV-1 genome , and we failed to detect UNG proteins in HBV NC ( Figure S2 ) . It is reasonable to suggest that UNG excises uracils in the nucleus , resulting in extensive nuclear cccDNA hypermutation in A3G-UGI transfectants . A proposed antiviral role for UNG is DNA degradation following DNA cleavage of an abasic site generated by uracil excision [13] , although UNG is primarily considered a DNA repair factor . In the present study , we compared viral production between UGI presence and absence in A3G-expressing hepatocytes . Those experiments did not show obvious rescue of A3G-mediated suppression of viral replication by UNG inhibition ( Figures 4 and 6 ) . The A3G-mediated reduction of NC-DNA was enhanced by UNG inhibition in prolonged culture ( Figure 6B–E ) , coincidentally with extensive hypermutation of cccDNA ( Figure 5 ) . In this experimental setting , UNG-dependent DNA degradation was not obvious . We accordingly propose that UNG removes uracil residues on/after cccDNA formation in the nucleus and that these uracil excisions contribute to reducing dysfunctional mutagenesis induced by APOBEC deaminases ( Figure S6 ) . Observation of hypermutated pre-C mRNA ( Figure 5D ) suggests that hypermutated pgRNA is also transcribed from the cccDNA and may contribute to the enhanced hypermutation in NC-DNA by UNG inhibition ( Figures 3 and S3B ) . Previous in vivo studies showed no evidence of cytidine deamination in the DHBV genome during chronic infection [62]; however , cccDNA sequences were not analyzed . In the present study , several G-to-A/C-to-T mutations in DHBV cccDNA were observed in LMH cells even without A3G overexpression ( Figure 5F ) , suggesting the contribution of endogenous deaminase activity . The chicken genome possesses only 3 AID/APOBEC members: AID , APOBEC2 ( A2 ) , and APOBEC4 ( A4 ) [13] , [56] , [63] . In mammals , A2 is considered to play a role in muscle development , and A4 may not have deaminase activity [54] , [56] . There is a functional association of AID and UNG with immunoglobulin gene diversification in human , mouse , and chicken B cells [12] , [29] , [30] , [51] . Recent reports showed that human AID endogenous expression was detected in hepatocytes after TGF-β stimulation [64] , and overexpression of AID caused HBV hypermutation of NC-DNA [24] . In future studies , we plan to assess the contribution of AID to antiviral activity against HBV and DHBV . APOBEC3G-mediated hypermutation of Vif-deficient HIV-1 caused deleterious effects on viral replication , whereas partial Vif activity accelerated viral diversification [65] , [66] . Similarly , we speculate that the balance between AID/APOBECs and UNG activities on mutation frequency decides the consequence to hepadnaviruses: deleterious mutations vs . diversification . DNA sequencing showed the existence of several cccDNA clones having a single G-to-A or C-to-T mutation ( Figure 5D ) , a finding that favors the concept of generation of clonal diversity by APOBEC and UNG proteins . This study suggests a possible role of APOBEC proteins as a mutator of HBV cccDNA . Because mutation in cccDNA is direct resource of viral variants , we are currently exploring the possibility that APOBECs and BER factors are involved in the emergence of drug-resistant mutants of HBV and DHBV .
The HBV replicon plasmids pHBV1 . 5 and pPB express all the HBV viral gene products necessary for viral replication and are under the control of the HBV and CMV promoters , respectively [40] , [41] , [44] . The DHBV replicon plasmid pCSD3 . 5 was generated by insertion of DHBV viral genomic DNA ( equivalent to 1 . 17-mer ) into pCMV-script ( Stratagene ) ( kindly provided by Dr . K . Kuroki , Kanazawa University ) [49] . The pCSD3 . 5ΔS plasmid was generated by introduction of in-frame stop codons at positions 1327 , 1346 , and 1349 in the surface protein ORF without affecting the P gene ORF [49] . Other expression vectors are listed in Table S1 . Cells ( HepG2 , Huh7 , 293T , and LMH ) and subsequent transfectants were grown and maintained in Dulbecco's modified Eagle medium ( DMEM; Sigma ) containing 10% fetal bovine serum , 100 U/mL penicillin , and 100 µg/mL streptomycin . IFNγ ( recombinant IFNγ-1a ) was purchased from Shionogi . Tamoxifen ( Wako ) was dissolved in ethanol ( EtOH ) . A stable line of HBV-expressing HepG2 cells was established using a standard method [11] . In brief , HepG2 cells were transfected with linearized pPB , and G418-drug selection with limiting dilution was performed . Of the resulting transfectants , a cell line with a high level of HBV NC-DNA was chosen among G418-resistant clones . Retrovirus-mediated gene transduction was performed as described previously [38] , [67] . In brief , retroviral vectors were transfected into packaging platA cells ( Cell Biolabs ) , and virus-containing culture supernatants were used for infection of HepG2 cells . Two days after infection , 1 µg/mL puromycin ( Wako ) was added to eliminate uninfected cells . Plasmid transfection was performed using Fugene 6 ( Roche ) according to the manufacturers' instructions . The total DNA amount ( 6 µg for 60-mm dish ) for each transfection was kept constant by adding GFP vector . The UNG assay was performed as described previously , with minor modification [39] . Cells were resuspended in HE buffer [25 mM Hepes-KOH ( pH 7 . 8 ) , 1 mM EDTA , 1 mM DTT , 10% glycerol] and fractured by freezing in liquid nitrogen and thawing . A fluorescein isothiocyanate ( FITC ) -labeled 31-mer oligonucleotide containing a central dU residue ( 5′-AGCTTGGCTGCAGGTUGACGGATCCCCGGGA-3′ ) was synthesized as a substrate . An FITC-labeled 15-mer oligonucleotide ( 5′-AGCTTGGCTGCAGGT-3′ ) was also synthesized for use as a molecular size marker . Approximately 10 pmol of substrate was incubated with cell extracts for 2 h , and the resulting abasic sites were cleaved with alkali and heat treatment . The reaction products were separated by 6 M urea/20% polyacrylamide gel electrophoresis . An FITC signal was visualized in an LAS imager system ( FujiFilm ) and quantified by densitometry using ImageJ software . Cytoplasmic HBV NC-DNA was purified as reported by Gunther et al . , with minor modifications [68] . In brief , the cells were lysed with buffer [10 mM Tris-HCl ( pH 8 . 0 ) , 1 mM EDTA , 1% NP-40 , 8% sucrose , proteinase inhibitor cocktail ( Roche ) ] . After centrifugation , cytoplasmic supernatants were collected and further treated with DNase I and RNase A . NCs were PEG-precipitated and digested with proteinase K and sodium dodecyl sulfate ( SDS ) . Secreted DHBV particles were also precipitated with PEG8000 , followed by DNase I treatment and digestion with proteinase K and SDS to extract viral DNA . For purification of DHBV DNAs , LMH cells were lysed in 0 . 5% NP40 lysis buffer , and the nuclei were collected by centrifugation to separate cytoplasmic and nuclear fractions . The cccDNA extraction from the nuclear fraction was performed using a modified Hirt extraction procedure [69] . The nuclear pellet was lysed in 50 mM Tris–HCl ( pH 7 . 5 ) , 10 mM EDTA , and 2% SDS . After 20 min incubation at room temperature , 0 . 5 M KCl was added to the lysate and incubated at 4°C overnight . From the supernatant after centrifugation , DNA was purified by phenol∶chloroform extraction and ethanol precipitation . All purified DNA solutions were treated with DpnI restriction enzyme to digest any contaminating plasmid DNA . The 3D-PCR procedure was performed as described previously , with minor modifications [18] , [28] . Primers used for 3D-PCR are shown in Table S2 . For 3D-PCR of HBV , the first PCR was performed as follows: 94°C for 5 min , 35 cycles of 94°C for 30 s , 50°C for 30 s , and 72°C for 30 s , and a final elongation step at 72°C for 3 min . The nested PCR was performed as follows: 94–83°C for 5 min , 35 cycles of 94–83°C for 60 s , 45°C for 30 s , and 72°C for 30 s , and a final elongation step of 72°C for 3 min . For DHBV NC-DNA , 1 round of 3D-PCR was performed using primers indicated in Figure S3A [23] . Initial denaturation was for 5 min at 94–83°C , followed by 35 cycles of 30 s at 94–83°C , 30 s at 55°C , and 2 min at 72°C , with a final elongation for 7 min at 72°C . For standard ( 94°C ) PCR of cccDNA , a cccDNA-selective primer set was used ( Figure S3A and Table S2 ) [52] , [53] . Specific amplification of cccDNA is shown in Figures S3C and S3D . To determine the hypermutation frequency , PCR fragments from 3D-PCR or standard PCR were cloned into T vectors ( Promega ) , and the indicated number of successful recombinant clones was randomly selected and sequenced using ABI PRISM 3130 ( Applied Biosystems ) . Total RNA was extracted using TRIsure ( Bioline ) , treated by amplification grade DNase I ( Invitrogen ) , and reverse-transcribed using an oligo-dT primer and the SuperScript III kit ( Invitrogen ) . For DHBV pre-C mRNA , a DHBV-specific primer was used in the reverse transcription reaction . qPCR analysis was performed using SYBR Premix Ex Taq ( Takara ) on an MX3000 thermocycler ( Stratagene ) following the PCR protocol . Human AICDA , A3B , A3C , A3DE , A3F , A3G , A3H , HPRT , DHBV pre-C , and chicken HPRT expression levels were measured using PCR conditions of 95°C for 1 min; 40 cycles of 95°C for 15 s , 55°C for 30 s , and 70°C for 30 s; and 1 cycle of 95°C for 1 min , 55°C for 30 s , and 95°C for 30 s . For A3A amplification , an annealing temperature of 60°C was used . For analysis of purified viral DNAs , qPCR was performed using the following conditions: HBV , 40 cycles of 95°C for 15 s , 52°C for 30 s , and 70°C for 30 s; DHBV , 40 cycles of 95°C for 5 s and 60°C for 20 s . HBV and DHBV DNA copy numbers were determined using a pPB or pCSD3 . 5 plasmid standard curve , respectively . Amplified fragments were designed to contain at least 2 DpnI sites to avoid amplification from contaminated plasmids . Primer sequences are listed in Table S2 . Cells were lysed in SDS sample buffer , sonicated , boiled , separated by 12% SDS-PAGE , and then transferred to a Hybond ECL membrane ( Amersham ) . The membrane was incubated in a blocking buffer of 5% skim milk in phosphate-buffered saline containing 0 . 1% Tween 20 . Signals were detected using the LAS1000 imager system . The antibodies used in this study were as follows: rabbit anti-A3G ( Sigma; raised against A3G peptide [CQDLSGRLRAILQNQEN] ) , rabbit anti-GAPDH ( G9545 , Sigma ) , mouse anti-FLAG ( M2 , Sigma ) , rabbit anti-ER ( HC-20 , Santa Cruz Biotechnology ) , rabbit anti-UNG ( ab23926 , Abcam ) , rabbit anti-HBc ( B0586 , Dako ) , anti-rabbit Igs-horseradish peroxidase ( HRP ) ( ALI3404 , eBiosource ) , and rabbit and mouse IgG Trueblot ( eBioscience ) . NAGE analysis was performed as described previously [21] , [22] , [48] . In brief , crude cytoplasmic extracts containing HBV NC particles were loaded into a 1% agarose gel for electrophoresis to separate intact capsid particles . After electrophoresis , the NC particles were denatured with NaOH and transferred onto a nylon membrane . HBV and DHBV DNAs were detected using a double-stranded HBV and DHBV DNA probes spanning the entire viral genome , respectively . Probe labeling and signal development were performed using the AlkPhos direct labeling system ( Amersham ) , and the signals were detected using the LAS1000 imager system . Margeridon et al . previously demonstrated that the RCA method specifically amplifies cccDNA with high fidelity but does not amplify any other intermediate DNAs [57] . We followed their method with minor modifications . In brief , DHBV cccDNA purified by Hirt extraction ( DpnI treated ) was mixed with 8 DHBV-specific primers ( Table S2 ) ; denatured at 95°C for 3 min; cooled sequentially at 50°C for 15 s , 37°C for 15 s , and room temperature; and reacted with the phi29 DNA polymerase and buffers ( New England Biolabs ) at 37°C for 16 h . Note that the DNA polymerase ( phi-29 DNA pol ) used in the RCA reaction can polymerize DNA even with a uracil-containing DNA template [70] . The RCA concatemerized product was converted to the monomeric full-length DHBV genome by digestion with EcoRI , where the DHBV sequence contains a single site , and was cloned into the replicon vector backbone at the EcoRI site; thus , the full-length DHBV genome in the original vector was replaced with the corresponding fragment from the purified cccDNA . These reconstructed plasmids were cloned , sequenced , and transfected into LMH cells in order to analyze their replication activities ( see Figure S5 for experimental design ) . Two UNG-specific siRNAs , two A3G-specific siRNAs and control siRNA ( Stealth Select grade ) were purchased from Invitrogen and were used to transfect using Lipofectamine 2000 , according to the manufacturer's instructions . Cells and viruses were analyzed 48 h after transfection . Cells were lysed with IP lysis buffer [50 mM Tris-HCl ( pH 7 . 1 ) , 20 mM NaCl , 1% NP-40 , 1 mM EDTA , 2% glycerol , a proteinase inhibitor cocktail ( Roche ) ] . After centrifugation , the supernatants ( cytoplasmic fraction ) were incubated with anti-HBc antibody ( DAKO ) and protein G sepharose ( Amersham ) , and passed through a micro BioSpin chromatography column ( BioRad ) . After the column was washed with the lysis buffer , the coprecipitated proteins were used for Western blotting . Statistical analyses were performed using GraphPad Prism ( GraphPad Software ) . ANOVA analysis was used for qPCR data . The Kruskal–Wallis test with Dunn's post test or Pearson's chi-square test were used for mutation analyses . P values less than 0 . 05 between experimental groups were considered statistically significant . For all graphs in this study , error bars indicate standard error of the mean from triplicate samples .
|
Human cytidine deaminases , AID/APOBECs , are restriction factors against various types of viruses . These proteins have the ability to introduce a cytidine-to-uridine ( C-to-U ) hypermutation in the viral DNAs of the hepadnaviruses hepatitis B virus ( HBV ) and duck HBV ( DHBV ) models . It is well known that uracil residues in human genomic DNA are removed by uracil-DNA glycosylase ( UNG ) , resulting in the creation of abasic sites that are repaired by downstream repair factors . However , the consequence of uracil removal from the viral genomic DNA remains controversial , given that it may be possible for abasic sites to trigger DNA degradation with strand breakage . We investigated the role of UNG in viral hypermutation and hepadnaviruses replication using in vitro cell culture systems . We found that UNG inhibition enhanced APOBEC3G-induced hypermutation of hepadnaviral DNAs , especially DHBV cccDNA , a template used for viral replication in the nucleus . We measured the replication ability of purified cccDNA and found that recloned cccDNA from cells expressed by both APOBEC3G and UNG inhibitor protein replicated less efficiently due to higher hypermutation rates . These results suggest that hepadnaviruses usurp the repair system of host cells to compete with AID/APOBEC mutators .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"hepatitis",
"hepatitis",
"b",
"dna",
"modification",
"nucleic",
"acids",
"dna",
"viruses",
"viral",
"classification",
"virology",
"dna",
"dna",
"repair",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"viral",
"diseases"
] |
2013
|
Uracil DNA Glycosylase Counteracts APOBEC3G-Induced Hypermutation of Hepatitis B Viral Genomes: Excision Repair of Covalently Closed Circular DNA
|
Visual estimation of the material and shape of an object from a single image includes a hard ill-posed computational problem . However , in our daily life we feel we can estimate both reasonably well . The neural computation underlying this ability remains poorly understood . Here we propose that the human visual system uses different aspects of object images to separately estimate the contributions of the material and shape . Specifically , material perception relies mainly on the intensity gradient magnitude information , while shape perception relies mainly on the intensity gradient order information . A clue to this hypothesis was provided by the observation that luminance-histogram manipulation , which changes luminance gradient magnitudes but not the luminance-order map , effectively alters the material appearance but not the shape of an object . In agreement with this observation , we found that the simulated physical material changes do not significantly affect the intensity order information . A series of psychophysical experiments further indicate that human surface shape perception is robust against intensity manipulations provided they do not disturb the intensity order information . In addition , we show that the two types of gradient information can be utilized for the discrimination of albedo changes from highlights . These findings suggest that the visual system relies on these diagnostic image features to estimate physical properties in a distal world .
The physical parameters that affect a retinal image are extremely complex . In addition , the same retinal image can be produced from an infinite number of combinations of materials , shapes , and illuminations in the distal world . Therefore , it is a hard ill-posed problem to estimate what exists in the distal world from a single retinal image . This appears to be a chicken-and-egg problem as material estimation requires knowledge about shape ( and illumination ) , while shape estimation requires knowledge about material ( and illumination ) . Nevertheless , ( we believe ) we can estimate the physical parameters that produce a retinal image . For instance , from a single photograph wherein a metal teapot is placed on a table , we can simultaneously judge the material and shape of the object . This paper concerns the visual processing underlying such simultaneous estimation . We found a clue for solving this problem in the image-based material editing methods developed in the computer graphics community [1–3] . By changing image parameters , not physical ones in the distal world , these methods can alter the material appearance of an object without significantly affecting its apparent shape or illumination . Among them , a simple yet effective method is to modulate the luminance histogram of an image . For instance , when the histogram of the original image in Fig 1 ( A ) is matched with that of the reference image in Fig 1 ( B ) , the material appearance of the original image becomes very similar to that of the reference image ( Fig 1 ( C ) ) . Another example is the use of monotonic nonlinear tone-remapping for print or screen display devices to transform the intensity histogram of an input image and modify its qualitative appearance [4 , 5] . Successful manipulation of material appearance by histogram transformation suggests that luminance histograms contain critical information for material perception [6–12] . Specifically , Motoyoshi et al . [8] found that a surface tends to look glossier when the luminance histogram of the surface’s image is positively skewed . Although histogram skewness can be affected by various image parameters , it can be a very good predictor of apparent gloss when the images are nearly the same in other respects , as is the case for histogram-transformed images . Motoyoshi et al . [8] also showed that adaptation to textures with skewed statistics alters the perceived glossiness of surfaces subsequently viewed . These findings led them to conclude that human observers may use histogram skewness , or some image features correlated with it , in making judgments about glossiness . However , when the spatial structure of an image is inconsistent with a natural glossy surface ( e . g . , a pixel- or phase-scrambled image ) , the image does not look glossy regardless of histogram manipulation [8] . Kim , Marlow and Anderson [13] further investigated spatial conditions of gloss perception and found that when specular highlights of an object image are inconsistent in position and/or orientation with the diffuse shading component , they look more like white blobs produced by surface reflectance changes ( see also , [14–17] ) . Marlow , Todorovic and Anderson [18] have demonstrated that three-dimensional shape perception of a surface affects gloss perception of the surface . These findings suggest that the visual system has to simultaneously solve at least three mutually dependent problems: it has to estimate surface material , surface shape ( surface orientation ) , and reflectance changes . As mentioned above , we believe that material editing by histogram manipulation suggests a clue to this complex computation . The histogram-matching method , as shown in Fig 1 , successfully changes the material appearance of a surface image , while it seems to have a negligible effect on surface shape . This suggests that the image properties changed by the histogram transformation affect material processing , while those unchanged by the histogram transformation affect shape processing . In what follows , we will show which components in the image are changed and unchanged by histogram manipulation and then consider the effects of each component on the perception of material , shape and reflectance change . The analysis will lead us to a computational strategy the visual system may follow to simultaneously and nearly independently estimate material , shape and reflectance change from a single image of an object . To anticipate the conclusion , we here propose that the human visual system may use orthogonal features about image intensity gradient to estimate material and shape: the intensity gradient magnitude for material perception , and the intensity gradient order for shape perception . We also suggest that the intensity order structure provides the critical information for discrimination of highlights from albedo changes [13–17] . The intensity gradient magnitude is related to the intensity histogram statistics , which some have suggested are related to material perception [8] , while the intensity gradient order is related to the isophote and orientation flow that have been suggested to be important for robust shape estimations [19–27] . Combining thoughtful insights originating from past theories with new image analyses and psychophysical experiments , we attempt to comprehensively understand how the human visual system simultaneously estimates many interdependent object properties from a single picture .
To explore image constraints for discriminating material changes from other property changes , we focused on a histogram-transformation method that has been widely used to edit the material appearance of a surface image [1 , 3 , 6] . In this method , to adjust the luminance histogram of an original image to that of a reference image , each histogram of the original and reference images is converted into a cumulative histogram ( Fig 1 ( D ) and 1 ( E ) ) . Then , the bin values of the original histogram are transformed into those of the reference histogram so that each cumulative value of the original histogram is matched to that of the reference one . Consequently , histogram matching does not change the intensity order of the image . When the pixel intensity of the output image is plotted as a function of that of the original image ( Fig 1F ) , the tone-remapping function monotonically increases , or at least does not decrease . Similar features are observed in general tone mapping techniques [5] . These observations suggest that retaining the intensity order of the original image may be the key feature for editing material while keeping other physical properties constant . When we consider the image generation processing of an object image , there are good reasons to believe that retaining the intensity order information is critical for material editing . A ( monochromatic ) surface image can be decomposed into albedo , shading , and specular images ( Fig 2A ) . The albedo image of a surface indicates how much illumination is diffusely reflected at each surface point . It is irrelevant to the surface normal and thus independent of the shading and specular images . The shading image of a surface is the interaction map of the surface normal and the illumination . With diffuse Lambertian shading , the shading intensity is a function of the incident angle of light . The specular component is the direct mirror-like reflection of the incident light . The specular intensity of a surface is a function of the incident and viewing angles of light . Since both specular and shading intensities depend on the incident angle of light , the specular image is dependent on the shading image . Under a collimated illumination , the surface normal directions of highlight regions are nearly uniform , and the intensity of the matte component behind the highlight regions is also nearly uniform ( Fig 2B ) . If the highlight regions are uniformly painted with ( or replaced by ) the hidden matte intensity , the image becomes something akin to a matte surface image . Although the hidden matte intensity is not known , the specular highlights tend to be produced near the highest intensity of the diffuse shading , but the position could shift slightly depending upon the difference between the incident angle and viewing angle [24] . Because of these constraints , adding specular gloss on a matte image , unlike adding an albedo change , has relatively little effect on the intensity order of the image . If the luminance order structure is the same , so is the isophote structure ( an isophote is a contour of equal intensity in an image ) . The direction of the luminance gradient is orthogonal to the isophote . Hence , if the intensity order information of an image is kept constant , the direction of the intensity gradient is too . Fig 2C shows how adding gloss affects the intensity gradient structures . To make an intensity gradient map , we computed the horizontal and vertical derivatives of the intensity distribution , and then converted them to the polar coordinate . In Fig 2C , the magnitude and direction of the intensity gradient vector are indicated by the hue and saturation of a color map , respectively . The intensity gradient map shows that highlight regions have larger gradient magnitudes , which implies that adding specular highlights drastically changes the gradient magnitudes . However , when the gradient magnitudes are normalized and only the directional information of intensity gradients is preserved , the map of the gloss image is similar to that of the matte image . The results suggest that adding gloss to a matte image has a negligible effect on the direction map of the intensity gradient . To test the generality of the observations , we analyzed material images rendered using the MERL BRDF ( Bidirectional Reflectance Distribution Function ) database [28] , which is a set of measured BRDFs of 100 materials , including rubber , plastic , metal , and fabric . There were four illumination conditions: three single point light sources ( slant = 0 , 20 , and 40 degrees ) and one HDR ( High Dynamic Range ) environment map ( Fig 3 ) . Fig 4A shows the results of the analysis . Each cell of the panels indicates the correlation coefficient of the magnitude or the direction map of the intensity gradient between the images rendered with the BRDFs of the row and the column . When the point light source lit the objects from the viewing direction , the direction of the intensity gradient consistently showed quite a high correlation ( Fig 4 ( B ) , upper ) , whereas the magnitude of the intensity gradient showed correlations that are relatively low and highly variable depending on the comparison pair ( Fig 4 ( B ) , bottom ) . These findings can be confirmed from the probability density distribution of the correlation coefficients in Fig 4 ( B ) . When the incident angle of the light deviates from the viewing angle , the position of specular highlights tends to shift from the position of the highest intensity of the diffuse shading [24] . However , the displacements of the lighting direction do not significantly change the pattern of results ( Fig 5 ( A ) and 5 ( B ) ) . The analysis of the surface images rendered under point light sources suggests that material changes ( with no changes in shape and illumination ) have little effect on the direction of the intensity gradient . Under the HDR illumination environment , the correlation in the direction of the intensity gradient is reduced for some material combinations ( Fig 6 ( A ) ) . This is because the direction of the intensity gradient is disturbed by the spatially complex illumination that produces spatially non-uniform mirror reflections , especially when the BRDF has low specular roughness . Since we computed the intensity gradient on a small scale ( the kernel size was 5 x 5 pix for an image size of 256 x 256 pix ) , the fine structures of a mirror reflection of the environment affect the direction of the intensity gradient . However , a simple tone operation can reduce the effect of a mirror reflection of the environment . While tone remappings normally change the intensity histogram without changing the intensity order , strong compressive tone remappings in which the output intensity levels off beyond a certain input intensity can remove the intensity gradients in the high intensity range . Since the mirror reflection generally has higher intensities than those of the shading pattern , a strong compressive tone mapping can eliminate the variation caused by the spatially non-uniform mirror reflection . In one analysis ( Fig 6 ( B ) ) , when the magnitude became smaller than a very small threshold value , we excluded the gradient values from computation of gradient directions . The analysis showed that the correlation in the direction of the intensity gradient is markedly improved . In addition , it should be noted that the strong correlations in the direction of the intensity gradient across different materials can be obtained only under similar illumination conditions . When we compute the correlation across different illuminations , e . g . , between the lighting conditions of 0° and 40° , the direction information of the intensity gradient is markedly different ( Fig 7 ) . The present analysis suggests that the material change of a surface strongly modulates the magnitude of the intensity gradient but does not unduly disrupt the intensity order or the direction of the intensity gradient . This explains why the histogram-matching method , which affects the magnitude of the intensity gradient while preserving intensity order , effectively changes the material appearance . At the same time , our analysis suggests that the intensity order of an object image contains rich information about the surface shape and reflectance pattern . In the context of computer vision , intensity order information is widely utilized as a feature descriptor [31–34] . For instance , Dalal and Triggs [31] utilized the local histograms of image gradient orientation ( called histograms of oriented gradient , or HOG ) . They showed that the descriptor is robust against environmental changes . In addition , shape-from-shading studies suggest that the intensity order information or the directional information of the intensity gradient is useful for shape estimation [19–27] . Fig 8 shows a hypothetical processing scheme that the human visual system may use for simultaneous estimation of a variety of surface properties . The critical idea is that an input surface image is analyzed in two ways . One focuses on the information about the order of intensity . It could be in the form of isophote , gradient direction map , or orientation map . Shape processing mainly relies on this intensity order information . The other image analysis focuses on the information about the magnitude of the intensity gradient . Material processing mainly relies on this gradient magnitude information . To be precise , the important information for material estimation is likely to be the intensity gradient relative to the surface orientation change [18] , but we assume this is computed in subsequent stages . The estimation of the remaining properties , i . e . , surface albedo and illumination , relies both on the intensity order and gradient information , along with the absolute intensity level . According to this hypothesis , one can tell whether a bright spot is produced by an albedo change or by a specular highlight by checking how it affects the intensity order information . While previous studies have suggested that luminance histogram manipulation is an effective way to change material appearance [8] , as well as pointing out the importance of orientation field or isophote map in shape perception [19–27] , to our knowledge , one potential implication of these findings has not been recognized . That is , material perception and shape perception may be based on separate , independent , orthogonal features of the object image , and this is why the visual system can simultaneously estimate material and shape . Although visual estimation of the material and shape appears to include a hard chicken-and-egg problem ( material estimation requires shape information , while shape estimation requires material information ) , the brain may be able to solve it by computing the two attributes , at least initially , based on the independently measurable image features . Although our hypothesis includes an explanation as to how the visual system robustly estimates the shape for some materials , it does not cover every kind of material . This is because our basic intuition came from a critical observation that luminance histogram matching affects apparent material , but not shape . While luminance histogram matching realized by monotonic luminance re-mapping can produce a wide range of matte and glossy objects , it cannot easily make mirrored objects with a perfectly specular reflectance . Hence , we do not have a strong theoretical basis to assume that our theory is applicable to mirrored objects . Textured objects and line drawings are also outside our scope . Compared to the “orientation field” theory proposed by [23–25] , we consider a more specific problem of monocular shape perception ( see Discussion for details ) . In the five psychophysical experiments reported below , we empirically tested our hypothesis . The first three experiments measured the apparent shape ( surface orientation ) of object images to see whether it is affected by the intensity order information but not by the intensity gradient magnitude information . Experiment 1 changed the intensity distribution by means of histogram matching that preserved the intensity order information . Experiment 2 changed the intensity distribution by means of non-linear intensity remapping that disrupted the intensity order information under some conditions . Experiment 3 disrupted the intensity order information more naturally by using velvet-like surface reflectance . The last two experiments examined apparent surface gloss and reflectance uniformity to ascertain whether they are affected by the intensity gradient magnitude information but not by the intensity order information . By using objects with veridical and inconsistent highlights , we considered how the two types of intensity information are used to discriminate material features from reflectance changes . Experiment 4 changed the intensity distribution by means of histogram matching , while Experiment 5 changed it by means of compressive remapping . In the previous section , we showed that perceived shape is sensitive to the intensity order information but not to the intensity gradient magnitude information . In this section , we will consider the perception of materials and surface reflectance properties . Although we have seen effective modulations of perceived material by changing the intensity gradient magnitude information with no change in the intensity order information , it is hard for the visual system to estimate material only from the intensity gradient magnitude information . This is not only because of the effects of surface shape on material perception [18] but also because albedo/reflectance changes on the surface of the object affect the intensity gradient magnitude information . For example , a white patch with a steep intensity gradient on the object surface could be either a specular highlight or white paint . To distinguish between them , the visual system can use the intensity order information , since the addition of a reflectance change does , while that of a highlight does not , make the intensity order map significantly different from that of the shading pattern of an object with diffuse uniform reflectance . When a highlight is located and/or oriented in a manner inconsistent with the shading pattern , it is perceived as an albedo change ( e . g . , white blob ) and does not make the pattern look glossy [14–17] . While past studies proposed congruence in brightness and orientation as conditions for highlight consistency , we additionally suggest that if a white patch is a specular highlight , it does not disrupt the luminance order of the shading pattern . This means that when reducing the bright patch intensity by histogram matching to a less skewed intensity distribution or by applying a compressive tone remapping , one can smoothly erase the highlight and obtain a diffuse surface image . This should not happen if a white patch is an albedo change , since an albedo change disrupts the luminance order map . It should remain visible regardless of how the intensity gradient magnitude information is altered by the manipulation of the intensity distribution . If this hypothesis is correct , our predictions are as follows: For consistent highlights , apparent glossiness is reduced for negative skew or by compressive tone mapping , and the uniformity rating is always low . For inconsistent highlights , apparent glossiness is always low , and the uniformity rating is always high . These predictions were tested in the following two psychophysical experiments .
Numerous studies have utilized histogram-transformation methods as used in Experiment 1 to modulate the pattern of intensity histogram [1 , 6 , 8 , 40–42] . As shown in our image analysis , the transformation does not disturb the intensity order information of a surface image , but it does distort the magnitude information of the intensity gradient of the image . In addition to histogram matching , compressive nonlinear tone mapping is widely used for appearance control in printing or screen display devices [4 , 5] . The mapping also usually retains the intensity order information of an input image . These techniques are consistent with the present finding that the modulation of the gradient magnitude information can be a diagnostic for the material appearance of the surface . If the intensity order of the image histogram of a surface is kept constant , then so is the isophote structure or the direction of the intensity gradient of the surface image . The effect of the structure of isophotes on surface shape estimation has been traditionally recognized in the context of shape-from-shading [19–27] . For instance , Koenderink & van Doorn [19–20] showed the structural relationships between the pattern of isophotes across a diffuse surface and the geometric structure of the surface . Specifically , they focused on the “Gauss map” , which is a spherical image where a surface in Euclidean space is mapped to the unit sphere . Since in simple stimulus situations ( e . g . , Lambertian materials under a collimated illumination ) the radiance of a point in a surface only depends on the surface normal , each radiance of the surface image with an identical normal can be mapped to the same point on the sphere image . Koenderink and van Doorn showed that when a specific region of a surface image , such as a convex , concave , or saddle-shaped region , is extracted based on the local extrema of the image , its spherical image corresponds to the surface geometry in a one-to-one fashion . Then the isophotes of the Gauss map can have invariant structures related to the surface geometry , irrespective of illumination directions . Similarly , Breton & Zucker [21] showed that under a diffuse surface illuminated by a point light source , the orientation of the intensity gradient field of the surface only depends on the geometric properties of the surface irrespective of the irradiance and the diffuse reflectance . They computed the “shading flow field” based on the orientation information and showed that the flow field can be used for shape estimation and edge classification . For instance , an attached shadow cast on a corrugated surface produces discontinuity in the continuous shading field of the surface and thus the discontinuity can be a cue for edge classification . More recently , Zucker and his colleagues introduced the idea of constructing a set of local surfaces based on the shading flow field for diffuse surfaces under any point light source [26] . Although the elegant analyses by Koenderink and van Doorn ( 1980 ) and Zucker et al . on potential shape information in intensity gradient maps assume Lambertian objects , the present findings indicate that their theories are also helpful for understanding shape perception for non-Lambertian materials . In this regard , the contribution of the present study is to show that the effect of the intensity order information is considerably robust against material changes . Our image analysis showed that the changes in natural BRDFs ( 100 types ) did not strongly affect the intensity order information of object images . In addition , we showed in Experiment 3 that when a specific material change distorted the intensity order information of an object image , the perceived shape was changed with the distortion . This finding is consistent with previous studies showing that shape constancy across specific materials could not be obtained [38 , 39 , 43 , 44] . The findings suggest that human shape processing strongly relies on the intensity order information and that distortion of the information tends to cause the perceived shape’s modulation even when actual material changes produce the distortion . Our psychophysical experiments show that keeping the intensity order constant makes the perceived shape constant ( Experiment 1 and 2 ) . In addition , when we disturb the intensity order information , the perceived shape changed with the distortion ( Experiments 2 and 3 ) . However , we emphasize that the same shape can have different intensity order maps and that different intensity order maps do not always produce different perceived shapes , due to , say , the effect of illumination differences . Our study mainly investigated material and shape perception under the conditions where an object is placed in a specific illumination environment , but as shown in the Image Analysis section ( Fig 7 ) , illumination changes can produce large distortions of the intensity gradient information . Nevertheless , identical objects under different illuminations can be perceived as similar in shape even when their intensity order information is markedly different , as shown in Experiment 2b ( Fig 14 ) . The previous studies also reported that the perception of shape and material is quite robust across different illumination contexts [45] . Hence , to recover the perceived shape from the intensity order information , the visual system has to discount the influence of the illumination field . While how it does this remains an open question , one possibility is that the occluding contour of a surface image may normalize the mid-level representation of an object obtained from the intensity order information [18] . Another possibility is that the visual system may extract some illuminant-invariant higher order differential structures from the intensity order information ( cf . , [26] ) . A luminance-order map is far from sufficient to recover the geometrical ground truth of an object even when material information is given . Shape estimation solely from luminance-order information must introduce many ambiguities . It is obvious that the shape-from-intensity-order-map problem suffers from bas-relief ambiguity [46] , since our theory concerns perception of matte ( diffuse ) and gloss ( diffuse+specular ) objects seen without light source information . In addition , in many cases , luminance-order maps must be equated between completely different shapes by adjusting material ( BRDF , BSSRDF , BTF ) and/or illumination parameters . Although we have not theoretically analyzed this ambiguity , this would seem to be a hard analysis , since it should consider not only geometrical optics , but also natural statistics of reflectance and illumination parameters . Furthermore , in order to understand human vision , the important issue is not only ambiguity in estimation of the ground-truth 3D structure , but also ambiguity in estimation of the perceived shape . The perceptual representation of shape is degenerated in the sense that it does not contain full detailed information about the ground truth structure , though what is perceptually represented about shape remains controversial [47] . According to our experiments , provided we preserved the luminance order map , the observers reported similar shapes . Despite enormous physical ambiguity , we found little evidence of perceptual ambiguity like that observed in the Necker cube . We think this provides an important hint about perceptual shape representations in the human brain . Consider next the relationships of luminance-order information to orientation information that Fleming and his colleagues proposed were influential in shape estimation [23–25] . Like Zucker and his colleagues , they constructed the orientation field of a surface image . The dominant orientation of the field was determined according to the relative powers of oriented-linear filters’ outputs . They showed that the distortion of the orientation field of a surface image corresponds to the distortion of the perceived shape of the surface . In computer graphics also , the orientation field has been utilized for apparent shape editing [27] . Specifically , Vergne et al . [27] used structure tensors of a surface image to construct the orientation field and showed that the modulation of the field drastically changed the apparent shape . In addition , they showed in their statistical analysis that the orientation information of identical shapes with different materials ( four types ) or illuminations ( four types ) can be similar to each other , as in our image analysis . Fleming and his colleagues have constructed a general framework of shape perception from image orientation information . Their investigation started from perfectly specular ( mirrored ) objects , and then generalized their theory to shape perception from diffuse shading , texture or contours . In contrast , our theory was based on a critical observation that luminance histogram matching affects apparent material . Since luminance histogram matching realized by monotonic luminance re-mapping can control the material appearance of an object in the range between pure matte ( diffuse ) and gloss ( diffuse+specular ) , but cannot easily make perfectly specular appearances ( we need non-monotonic luminance re-mapping [41] ) , we do not have a strong theoretical basis to assume that our theory is applicable to mirrored objects . Textured objects and line drawings are also outside our scope . Despite having a scope narrower than that of Fleming et al . , our theory has more specific predictions about material and shape perception of objects within the scope of our analyses . A critical question is what kind of directional information , i . e . , a vector map modulo 180 or 360 degrees , or both , the visual system relies on . The orientation field of Fleming and his colleagues is based on a vector map modulo 180 degrees , while the present study used a vector map modulo 360 degrees to explain the perceived shape . In Experiments 2 and 3 , we found that the modulation of the tone-mapping curves of a surface image changed the perceived shape of the surface , even though it only distorted the vector map modulo 360 degrees while keeping constant the vector map modulo 180 degrees . The finding suggests that at least for the class of materials we considered , shape perception is different when the orientation map is similar , but the luminance order is different , as predicted by our theory . Fleming et al . proposed a ground theory for a wide range of monocular shape perception including cases where shape perception is similar even when luminance order is not preserved [48–50] . We also recognize that non-linear tone re-mappings that do not preserve luminance order information are able to change glossy objects into mirrored or translucent objects of similar shapes ( e . g . , [41] ) . We speculate the perceived shape distortion may depend on the spatial flow structure where the flow distortion emerges . For instance , in Experiment 3 ( Fig 17 ) a perceived shape distortion for the asperity ( a = 0 . 2 ) condition was obtained in gauge position 6 where the shading flow produces a cusp . The finding suggests that the effectiveness of the vector map modulo 360 degrees may depend on the diagnostic flow structure . Although we show the importance of the signed intensity gradient in shape estimation , we agree that unsigned orientation measurements are also useful in shape processing . For instance , the computation based on 180 degrees would be beneficial in the estimation of specular-only images because the processing modulo 180 degrees is tolerant to the first-order modulation due to mirror reflections . Thus , the processing based on the vector map modulo either 180 or 360 degrees has benefits in some situations . This suggests a possibility that two types of processing are adopted by the visual system , and thus further studies are necessary to elucidate how human shape processing codes directional information . While Fleming et al . [23–25] measured unsigned orientations at multiple scales , the luminance gradient computation normally has a single value at each location . Simultaneous gradient measurements at multiple spatial scales might be more beneficial , but we have not investigated this possibility yet . In sum , although the relationship of our theory with the theory of Fleming et al . is still to be clarified , we can at least say that a luminance-order map contains richer information than that of an orientation map , and that human shape processing does not refrain from using the extra information when it is available and useful . As for the effect of gradient magnitude , although we showed that the perceived shape is little affected by modulating the magnitude of intensity gradients , we understand that in some cases the perceived shape , especially perceived volume , can be affected by position/scale specific modulation . When intensity magnitudes are small , the surface tends to appear to have less curvature ( i . e . , they appear flatter ) than when the magnitudes are large , even if the intensity ordering is constant . In particular , Giesel & Zaidi [51] showed that enhancing the amplitude of specific spatial frequency components increases perceived volume , although the modulation does not change the perceived tilt . It has been shown that the perceived volume ( slant ) of an object image is unstable , compared with its perceived tilt [52] , and therefore it might be affected by several factors depending on the context in which the object is placed . This study suggests that the computation based on two types of intensity gradient information may facilitate a comprehensive understanding of material and shape processing . In addition , we showed that this computation may also be used for the perception of reflectance changes . That is , the present study revealed that the specular-shading consistency could be judged in the shading processing as a problem of discrimination of smooth shadings from reflectance changes . It is noteworthy that in Experiments 4 and 5 the perception of albedo-uniformity for an object image with inconsistent highlights was not changed by histogram modulations ( Figs 20 and 23 ) . This finding suggests that processing based on the intensity order information may be sufficient for discriminating an object image with veridical highlights from inconsistent ones . Although the algorithms of intrinsic image decomposition in the field of computer vision can discriminate smooth shadings from reflectance changes [53–59] , it is not easy for many of them to discriminate veridical specular highlights from reflectance changes such as white blobs . For instance , when one of the cutting-edge algorithms [58] is applied to object images with veridical and inconsistent highlights , even for a uniform albedo image with veridical highlights , it incorrectly detects the highlights as regions with different albedos . However , when the same algorithm is applied to a slope-normalized image ( Fig 24B , right ) , it correctly predicts the image with veridical highlights to have a uniform albedo , while the image with inconsistent highlights to have non-uniform albedos . This observation suggests that , if the visual system has a pre-processing stage to extract a luminance-order ( slope-independent ) image , it can easily discriminate smooth shadings from reflectance changes , and correctly solve the highlight consistency problem . We do not intend to dispute a previous hypothesis that the position and orientation congruence of specular highlights relative to diffuse surface shading could be critical for discrimination [13] . In terms of our hypothesis , position and orientation incongruences imply non-smooth luminance-order maps , and thus are likely to arise from albedo changes . The cortical processing of intensity order information , as well as that of intensity magnitude information , remains unclear . One plausible hypothesis is that the brain decodes the direction and magnitude of the local intensity gradient from the outputs of orientation-selective filters . These filters should be located at early stages where local phase information is preserved , not at later stages where local orientation energy is represented . However , since little attention has been paid to intensity order information , we can only speculate on its cortical mechanism at present . Our brain may adopt a completely different strategy to process luminance-order information . We hope the present psychophysical findings will motivate future neurological investigations into the mechanisms of cortical processing of the intensity order and magnitude information . Specifically , it would be interesting to see which cortical areas are more sensitive to intensity order than to intensity magnitude , and vice versa . Some neurological studies have found gloss-selective neurons in the ventral stream of monkeys and common marmosets [60–62] . For instance , Nishio et al . [61] found neurons in the inferior temporal ( IT ) cortex of the monkey that selectively and parametrically respond to physical gloss enhancements . These neurons are likely to be more sensitive to intensity gradient magnitude information . On the other hand , object-selective neurons in the other parts of the IT cortex may be more sensitive to intensity order information . While investigating the effects of histogram transformation methods on material perception , we showed that material processing depends on detailed gradient information rather than intensity order information , such as the direction of the intensity gradient . These findings also revealed the image constraints produced by other physical properties such as albedo and shape . The present study suggested that specular-shading consistency could be judged from intensity order information , with which a specular consistency problem becomes a general shading-reflectance separation problem . In addition , our study suggests that human perception of shape from shading is sensitive to the intensity order information of an object image but not sensitive to the detailed intensity gradient information .
All the psychophysical experiments were approved by the Ethical Committees at NTT Communication Science Laboratories and were conducted in accordance with the Declaration of Helsinki . We applied a variety of nonlinear remappings to three object images ( Fig 11 ) . The remapping function was defined as follows . f4 ( Lo ) =s1 ( Lo−m ) +s2sin ( ω ( Lo−m ) +ϕ ) +m , 4 ) where m is the mean intensity , s1 is the slope of the remapping function , s2 is the amplitude of a sinusoidal modulation , ω is the angular velocity of the modulation , andφis its phase . Specifically , we made fifteen remapping curves by adding a sinusoidal modulation with one of five different amplitudes ( s2 = 0 , 0 . 015 , 0 . 065 , 0 . 115 , or 0 . 165 ) to a linear tone remapping function with one of three different slopes ( s1 = 0 . 5 , 1 , or 2 ) . The ω and φ in the experiment were the constant values of 2 . 857π and π , respectively . When the slope was the steepest ( 2 ) , the remapping curve always monotonically increased . This implies that the intensity order of the original image was not disrupted even by the largest modulation . When the slope was midway between steep and gentle ( 1 ) , the remapping curve was non-monotonic , and the intensity order of the original image was disrupted when a1 was 0 . 115 or 0 . 165 . When the slope was gentle ( 0 . 5 ) , the intensity order was disrupted when a1 was 0 . 065 , 0 . 115 or 0 . 165 . It should be noted that our manipulation did not change the orientation ( modulo 180 degrees ) map of the image . The geometric models we used in the experiment were two bumpy spheres ( the displacement in the normal direction of the surface of each sphere was given by a coarse or fine Gaussian band-pass noise; Fig 11 , bottom left and bottom center ) , and a cylinder ( Fig 11 , bottom right ) . Each model was lit by a point light source from the camera direction . Eight observers were asked to estimate the perceived shape of the objects by setting a gauge probe with the matching apparent surface slant/tilt . The position of the nine gauge probes is shown in Fig 11 . Each of the 45 stimuli ( 3 objects x 15 tone-mapping types including the original image ) was tested three times for each observer . In addition , to confirm the performance stability of the gauge task , the gauge matching for the original images of the three objects was again conducted three times in a different session . These data were used for the baseline and showed as no shape changes in Fig 12 . The other methods were the same as in Experiment 1 . In the experiment , we applied a variety of non-linear remappings , as in Experiment 2a , on several object images rendered under a point light source place in the same direction as the viewing one or in the upper-right direction for the object where the illumination slant was 45° and the illumination tilt was 30° . Ten observers were asked to estimate the perceived shape of the objects by setting a gauge probe with the matching apparent surface slant/tilt . The position of the six gauge probes is shown in Fig 13 . Each of the 10 stimuli ( 2 illumination direction x 5 tone-mapping types including the original image ) was tested ten times for each observer . The other methods were the same as those used in Experiment 2a . The geometric model was a bumpy object with low spatial frequency bumps ( Object 4 ) ( Fig 16 ) . The model was lit by a point light source from the camera direction . The reflection model used in the experiment was the Lambertian or asperity material [38 , 39] . The output intensities on the models were determined as follows: Ll=ρdπll ( I⋅N ) , 5 ) La=aπ[a+ ( I⋅N ) ( J⋅N ) ]la ( I⋅N ) , 6 ) where Ll and La are the intensity of the Lambertian and asperity materials , respectively . I , and J are the incident and reflected angles , respectively . N is the surface normal , l is the intensity of the light source , and a is the asperity parameter known as the edge brightening factor . When the incident and reflected angles are the same , which is true under the lighting condition we used , the intensity of the asperity material , La , can be described as a function of the intensity of the Lambertian material as follows: La=alaLlaρdll+π2Ll2ρdll 7 ) This could be regarded as an intensity remapping function from Lambertian to asperity materials , the shape of which is dependent on parameter a ( Fig 15 ) . As the Ll increases , the remapping function first rises and then falls after a transition point . The transition point ( peak ) shifts towards the lower range with a decrease in a . When a is moderately small ( 0 . 2 ) , the intensity order is preserved in the lower range of Lambertian pixel intensity , while reversed in the higher range . When a is even smaller ( 0 . 02 ) , the intensity order is reversed in most of the intensity range . We rendered asperity objects using these two values of a ( Fig 16 ) . In the experiment , the parameters ll , and ρd were π and 0 . 6 , respectively . The parameter la was π under the asperity ( a = 0 . 2 ) condition , and 4π under the asperity ( a = 0 . 02 ) condition . In addition to a Lambertian object and two asperity objects , we used an object the intensity of which was completely reversed from that of the Lambertian object ( Fig 16 ) . Ten observers participated in Experiment 3 . The observers were asked to estimate the perceived shape of the objects by setting a gauge probe . The other methods were the same as those used in Experiment 2 . The same ten observers in Experiment 3 participated in one additional experiment . In the experiment , the geometric model was lit by a point light source in the upper-right direction for the object where the illumination slant was 45° and the illumination tilt was 30° ( Fig 18 ) . The same four reflection models and nine gauge probes as those in Experiment 3 were used . The observers were asked to estimate the perceived shape of the objects by setting a gauge probe .
|
Objects in our visual world contain a variety of material information . Although such information enables us to experience rich material impressions , it can be a distraction for the estimation of other physical properties such as shapes , albedos , and illuminations . The coupled estimation of these properties we humans perform in daily situations is one of the fundamental mysteries in visual neuroscience . Here , we show that material and shape perception depend on two different types of intensity gradient information . Specifically , our image analyses and psychophysical experiments show that a human’s material perception relies mainly on the intensity gradient magnitude information , while shape perception relies mainly on the intensity gradient order information . In addition , we show that the intensity order information can be utilized for discriminating albedo changes on an object surface from other physical properties including specular highlights .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"classical",
"mechanics",
"atmospheric",
"science",
"astronomical",
"sciences",
"social",
"sciences",
"light",
"neuroscience",
"electromagnetic",
"radiation",
"probability",
"distribution",
"luminance",
"mathematics",
"reflection",
"visible",
"light",
"skewness",
"research",
"and",
"analysis",
"methods",
"albedo",
"sensory",
"physiology",
"imaging",
"techniques",
"probability",
"theory",
"physics",
"visual",
"system",
"psychology",
"climatology",
"planetary",
"sciences",
"earth",
"sciences",
"physiology",
"biology",
"and",
"life",
"sciences",
"sensory",
"systems",
"physical",
"sciences",
"sensory",
"perception",
"image",
"analysis"
] |
2018
|
Material and shape perception based on two types of intensity gradient information
|
The chronic phase of HIV infection is marked by pathological activation of the immune system , the extent of which better predicts disease progression than either plasma viral load or CD4+ T cell count . Recently , translocation of microbial products from the gastrointestinal tract has been proposed as an underlying cause of this immune activation , based on indirect evidence including the detection of microbial products and specific immune responses in the plasma of chronically HIV-infected humans or SIV-infected Asian macaques . We analyzed tissues from SIV-infected rhesus macaques ( RMs ) to provide direct in situ evidence for translocation of microbial constituents from the lumen of the intestine into the lamina propria and to draining and peripheral lymph nodes and liver , accompanied by local immune responses in affected tissues . In chronically SIV-infected RMs this translocation is associated with breakdown of the integrity of the epithelial barrier of the gastrointestinal ( GI ) tract and apparent inability of lamina propria macrophages to effectively phagocytose translocated microbial constituents . By contrast , in the chronic phase of SIV infection in sooty mangabeys , we found no evidence of epithelial barrier breakdown , no increased microbial translocation and no pathological immune activation . Because immune activation is characteristic of the chronic phase of progressive HIV/SIV infections , these findings suggest that increased microbial translocation from the GI tract , in excess of capacity to clear the translocated microbial constituents , helps drive pathological immune activation . Novel therapeutic approaches to inhibit microbial translocation and/or attenuate chronic immune activation in HIV-infected individuals may complement treatments aimed at direct suppression of viral replication .
Persistently elevated immune activation characterized by polyclonal B cell activation [1] , increased T-cell turnover [2] , increased frequencies of T cells with an activated phenotype [3] , and increased levels of pro-inflammatory molecules [4] is a hallmark of disease progression in pathogenic HIV/SIV primate lentiviral infections and is a stronger predictor of disease progression than either CD4+ T-cell count or plasma viral load [5] . The importance of immune activation to disease progression in HIV/SIV infections is highlighted by the low levels of immune activation measured during the chronic phase of infection in natural hosts of SIV such as African green monkeys ( AGMs ) and Sooty mangabeys ( SMs ) , which do not progress to AIDS [6] . While the consequences of immune activation in HIV/SIV infection are numerous and include increased numbers of activated CD4+ T-cell targets for the virus , attrition of the memory CD4+ T-cell pool and accumulation of high frequencies of terminally differentiated and exhausted memory T and B cells , the underlying mechanisms and sources of immune activation during infection are not well understood [7] , [8] . Given accumulating evidence that persistent immune activation is at the heart of disease progression , understanding the mechanisms driving immune activation in chronic HIV disease will be important for the development of new adjunctive treatment strategies targeting this process . Although many factors may contribute to immune activation during chronic HIV/SIV infection , recent evidence has indicated that translocation of microbial products from the lumen of the intestine into the periphery may contribute importantly to this process [9] , [10] , [11] , [12] , [13] , [14] . These microbial products can stimulate immune cells directly via pattern recognition receptors such as toll-like receptors . Indeed , immune activation related to microbial translocation occurs in other settings and has been implicated in many other pathological conditions . For example , the preconditioning chemotherapy and radiation prior to progenitor stem cell transplantation in individuals with hematological malignancies leads to damage of the tight epithelial barrier of the gastrointestinal ( GI ) tract resulting in microbial translocation [15] . These translocated microbial products can then stimulate the immune system , exacerbating graft versus host disease [16] , [17] , [18] . Microbial translocation leading to immune activation also occurs in inflammatory bowl disease [19] , after invasive surgery [20] , [21] , and in pancreatitis [22] . While microbial translocation has been indirectly implicated in driving immune activation in chronically HIV-infected humans and SIV-infected rhesus macaques ( RMs ) , the mechanisms underlying this phenomenon remain unclear , with enterocyte apoptosis [23] , massive loss of GI tract CD4+ T cells [24] and/or preferential loss of GI tract Th17 cells [25] , [26] all proposed as important contributing factors . Moreover , the timing of the onset of microbial translocation relative to infection has remained obscure , and direct evidence of translocation at the tissue level has been lacking . Here , using a quantitative image analysis approach to study large segments of tissue , we provide direct immunohistochemical evidence of translocation , define the timing of microbial translocation in pathogenic SIV infection of RMs and identify loss of the integrity of the intestinal epithelial barrier as a plausible mechanistic correlate of microbial translocation . The absence of translocation or associated immune activation in chronic SIV infection of SMs , which does not result in progressive disease , underscores the critical role this process plays in the pathogenesis of primate lentiviral infections and the potential value of limiting it as an approach to adjunctive therapy .
Initially , we sought evidence of microbial translocation by staining with a monoclonal antibody against LPS-core antigen in paraffin-embedded colon tissue sections obtained at necropsy from SIV-uninfected RMs ( n = 6 ) , RMs euthanized during early acute ( n = 10 ) or late acute ( n = 3 ) SIV infection , and chronically SIV-infected RMs euthanized at protocol specified endpoints ( “Non-AIDS”; n = 8 ) or clinical endpoints ( AIDS defining conditions , “AIDS”; n = 5 , Table 1 ) . Prior to and during early acute SIV infection , the LPS-core antigen-specific mAb stained only rare cells within the lamina propria ( LP ) , but in dramatic contrast , in chronically infected animals both numerous LPS+ cells and abundant extracellular core LPS antigen were observed ( Figures 1 and S1 ) . Figures show a broad spectrum of microbial tanslocation and discontinuities to the structural barrier of the GI tract ( discussed in detail below ) that was seen in our animal cohort , from negligible ( SIV- ) to most severe ( AIDS ) . Although the amount of apparently extracellular bacterial constituents within the LP varied among our chronically SIV-infected RMs; all chronically SIV-infected RMs showed readily demonstrable evidence of microbial translocation in multifocial lesions along the GI tract , findings that were absent in the SIV-uninfected animals and in early acute SIV infection ( i . e . 1–10 dpi ) . Importantly , microbial translocation was evident in the large bowel of RMs chronically-infected with different pathogenic strains of virus ( i . e . SIVmac239 , SIVmac251 and SIVsmE660; Table 1 ) , suggesting that intestinal damage leading to microbial translocation is a common feature of pathogenic SIV infections . The specificity of our immunohistochemical staining directly documenting microbial translocation in gut tissue sections is supported by: ( i ) the remarkably low frequency of LPS+ cells within the LP in SIV-uninfected or RMs with early acute infection; ( ii ) the absence of staining with isotype-matched control antibodies in SIV-uninfected , acutely and chronically SIV-infected animals ( data not shown ) ; ( iii ) the lack of any evidence of non-cell associated LPS in the LP of our SIV-uninfected and early acute SIV-infected RMs , despite the abundant LPS staining in the residual luminal content of these same samples; ( iv ) the detection of microbial products within the LP of the colon of chronically SIV-infected RMs that were rarely seen in SIV-uninfected and early acute infected RMs , using a rabbit polyclonal antibody against Escherichia coli that recognizes many E . coli proteins ( Figures 2 and S2 ) , or using peptide nucleic acid fluorescent in situ hybridization to detect bacterial 16S RNA ( data not shown ) . To relate microbial translocation in the large bowel of chronically SIV-infected RMs to systemic microbial translocation , we stained sections from lymph nodes , identifying large numbers of LPS+ cells within both local draining lymph nodes ( mesenteric; MesLN ) ( Figures 3 and S3 ) , and remote peripheral lymph nodes ( axillary; AxLN ) ( Figures 4 and S4 ) of chronically SIV-infected RMs , suggesting systemic dissemination of translocated microbial constituents originating from the GI tract . Consistent with what might be expected for the anatomic site of inductive T-cell immune responses , we found immunoreactive LPS within the medullary cords and paracortex of draining MesLN as well as peripheral AxLN during the chronic stage of infection . In contrast , in SIV-uninfected or early/acute SIV-infected RMs we found only low levels of LPS+ cells in the gut-draining MesLN and virtually no LPS+ cells in the peripheral AxLN of ( Figures 3 , 4 S3 and S4 ) . The primary localization of LPS within the medullary cords and sinuses , and to a lesser extent the paracortex and germinal centers , in SIV+ Non-AIDS RMs is consistent with microbial products traversing from the damaged intestine via the lymphatics . The presence of LPS within the paracortex and germinal centers suggests a ) antigen presenting cells which have bound microbial antigens migrate into the T cell inductive site of the LN and b ) possibly microbial product-immune complex deposition on follicular dendritic cells may be occurring . However , the biological relevance of finding microbial constituents in these anatomical sites , at this point , remains unclear . Moreover , consistent with the liver's important function as a “gatekeeper” between the intestine and peripheral circulation , we also found multifocal evidence of microbial products within the liver , in regions surrounding the hepatic portal veins and tracts in chronically SIV-infected Non-AIDS RMs with more extensive staining into the lobules of the liver in chronically SIV-infected AIDS RMs , consistent with increased intestinal damage correlating with more microbial dissemination ( Figure 5 and Figure S5 ) . Using these approaches , we identified unequivocal direct evidence of microbial constituents within the LP of the large bowel , and in the liver and lymph nodes of all chronically SIV-infected RMs studied , but not uninfected RMs . This finding was consistent across all 32 RMs studied , including SIV-uninfected RMs ( n = 6 ) ; early acute SIV-infected RMs ( 1–10 dpi , n = 10 ) ; late acute RMs ( 14–28 dpi infection , n = 3 ) ; and chronically SIV-infected RMs ( 56–397 dpi , n = 13 ( Non-AIDS and AIDS ) ; Table 1 ) and indicate that microbial translocation involves infiltration of microbial products into the LP of the GI tract during the chronic phase of SIV infection of RMs . After demonstrating the qualitative presence of increased microbial products in the LP of large bowel , liver and lymph nodes of SIV infected RMs , we used quantitative image analysis techniques [27] to quantify the extent of microbial translocation demonstrable in high power ( 400× ) digital scans , of tissue section whole mounts ( ScanScope CS System , Aperio ) . Sections of colonic mucosa analyzed for each animal represented , on average , a total of 350 distinct 400× image fields per scanned tissue , providing an in-depth , systematic , assessment of segments of the GI tract that ranged from 43 to 101 linear mm of colonic epithelial lining and 16 to 39 mm2 of intestinal mucosal area . This comprehensive analysis approach , applied to randomly selected tissue sections , provided us with an unbiased and detailed evaluation of microbial translocation in representative sections of the GI tract that otherwise may not have been possible using conventional established tissue analysis methods . Using this approach , we found that the percent area of LP of the colonic mucosa containing LPS was significantly higher in chronically-infected RMs compared to uninfected and acutely ( early ) SIV-infected RMs ( Figure 6A ) . We used the same quantitative image analysis approach to evaluate the relationship between the microbial product burden within the LP of the colon and in draining ( MesLN ) and distant LN ( AxLN ) from the same animals , calculating the percent area of each tissue that contained LPS . We found a significant positive correlation between the amount of LPS within the LP of the colon and the amount of LPS within the corresponding draining MesLN ( r = 0 . 69 , P = 0 . 0065 , Figure 6B ) . Moreover , we also found a significant positive correlation between LPS staining within the MesLN and within the matched AxLN ( r = 0 . 59 , P = 0 . 027 , Figure 6C ) . Taken together , these data indicate that the presence of microbial products in the peripheral lymphatic tissues is intimately linked to microbial translocation from the gut . Microbial products can directly stimulate the innate immune system via interactions with Toll-like receptors ( TLR ) that lead to an inflammatory cascade . To evaluate the possible relationship between translocation of microbial products and inflammation we performed double-label immunohistochemical staining for microbial products and the innate proinflammatory cytokine IFNα . Reflecting the immune activation of chronic SIV infection in RMs , IFNα expression was widespread and we consistently demonstrated co-localization of IFNα and microbial products within the LP of the colon ( Figure S6 ) , and in AxLN , and MesLN ( data not shown ) . The majority of such co-localization occurred in the absence of detectable local viral replication , as during the chronic phase of infection productively infected cells within the LP are only rarely demonstrable by in situ hybridization in most SIV-infected RMs that have not progressed to AIDS ( data not shown ) . Indeed , in extensive double label studies using immunohistochemical staining for IFNα and in situ hybridization for SIV RNA in tissues from chronically SIV-infected Non-AIDS RMs , there was only very limited co-localization between IFNα expression and in situ hybridization for SIV RNA in the colon or MesLN ( data not shown ) . The overwhelming majority of the abundant IFNα immunostaining was found relatively distant from local viral replication , but in close proximity to microbial products ( data not shown ) . Moreover , the levels of IFNα immunostaining were significantly greater than the tissue level of SIV RNA . To evaluate further the role of translocated microbial products in driving immune activation in chronic SIV infection , we assessed the distribution and co-localization of interleukin-18 ( IL-18 ) and LPS in the gut-draining MesLN in SIV-uninfected and chronically SIV-infected Non-AIDS RMs ( Figure S7 ) . IL-18 is produced by activated macrophages and dendritic cells in response to microbial product stimulation [28] . We found that before SIV infection there was a basal , low-level , expression of IL-18 in all structural compartments of MesLN ( i . e . B cell follicles , T-cell zone and medullary cords ) and that this expression was dramatically increased in SIV infection ( data not shown ) . Although IL-18 was up-regulated in some regions where there were no visualized microbial products , high levels of IL-18+ cells were always found in close proximity to LPS in MesLN ( Figure S7 ) . Consistent with a role for translocated microbial products inducing immune activation and IL-18 expression , Ahmad and colleagues recently described significantly elevated levels of IL-18 in the serum of HIV-infected/AIDS patients compared to those of HIV-seronegative healthy individuals [29] . Collectively , these data strongly suggest that microbial products , which infiltrate the LP of the GI tract , and then spread systemically , can directly stimulate the immune system and contribute to chronic immune activation . To determine how early during infection microbial translocation occurs , we compared microbial translocation into the LP of the colon of RMs throughout the acute stage of SIV infection ( 1 to 28 dpi ) in vaginally-challenged animals . With the exception of one animal at 8 dpi , we found only very low levels of LPS+ cells within the LP of RMs between 1–10 dpi , observations that were indistinguishable from SIV-uninfected animals ( Figures 6 and data not shown ) . There was a statistically significant increase in LPS seen within the LP of animals infected for between 14 and 28 days compared to uninfected or early/acute animals and evidence of microbial translocation was detected at small foci associated with breaks in the epithelial lining ( Figures 6 and data not shown ) . Importantly , during the acute phase of infection , areas of discontinuity in the epithelial barrier were relatively infrequent , while LPS staining in the LP appeared to increase into the late acute stage of SIV infection ( 14–28 dpi ) . Interestingly the extent of lesions and discontinuities were lower than might have been expected relative to the massive enterocyte apoptosis previously described as peaking at 14 dpi in these same animals [23] . We comment below on the possible mechanisms responsible for this dissociation . Although enterocyte apoptosis subsequently decreased , even as microbial translocation increased in the late acute stage of infection , the abundant evidence of LPS+ within the lamina propria at 28 dpi suggested that early damage to the integrity of the epithelial lining could facilitate translocation of microbial products ( data not shown ) . Thus , we sought to determine directly if the compromise of the integrity of the epithelial barrier is a distinguishing feature of microbial translocation during chronic SIV infection , evaluating the integrity of the epithelial barrier by staining for the tight junction protein claudin-3 . A similar technique has been used in studying human samples to assess the integrity of the epithelial barrier in diseases associated with discontinuities in the GI tract [30] . Examination of the integrity of the epithelial barrier by staining for claudin-3 , revealed multifocal disruptions and epithelial loss of the normally continuous , epithelial barrier in tissues from the chronic stages of infection , but not in tissues from uninfected or early acute animals ( Figure 7A and Figure S8 ) . We recognize the possible contribution that undetected opportunistic enteric pathogens may play in the continuum of epithelial damage seen in our chronically SIV-infected ( AIDS ) RMs , and thus show in Figures 7 and S8 two examples of end-stage RMs demonstrating this dynamic spectrum of epithelial damage from multifocal epithelial loss to severe epithelial damage and ulceration . Importantly , confocal analysis showed that disruptions in the integrity of the epithelial barrier were directly associated with translocated microbial products ( Figure 7B ) . Furthermore , quantitative image analysis from high power entire colonic tissue section scans , confirmed that chronically-infected animals had significantly more damage to the integrity of the epithelial barrier compared to uninfected and acutely-infected RMs ( Figure 8A ) . Moreover , the degree to which the integrity of the epithelial barrier was compromised was significantly correlated with the amount of LPS within the LP ( Figure 8B; r = 0 . 57 , P = 0 . 032 ) . Loss of integrity of the epithelial barrier of the GI tract would be expected to result in several host responses including polymorphonuclear neutrophil ( PMN ) infiltration and increased local proliferation of enterocytes within colonic crypts in an attempt to restore the integrity of the GI tract . Importantly , we observed increased levels of myeloperoxidase+ PMNs within the lamina propria of chronically SIV-infected RMs associated with damage to the epithelial barrier ( Figure S9 ) , providing strong evidence for a tissue specific response to GI epithelial damage . To assess proliferation of GI tract enterocytes we immunohistochemically stained colon tissues with a monoclonal antibody against Ki67 , a cellular marker for proliferation , and performed quantitative image analysis measuring the fraction of enterocytes along the colonic crypt that were proliferating ( Ki67+ ) . We found increased levels of proliferating enterocytes ( Ki67+ ) in chronically SIV-infected animals compared to SIV uninfected and acutely infected RMs ( Figure 9A–B and Figure S10 , P = 0 . 016 ) . Moreover , there was a trend towards an increase of Ki67+ enterocytes in animals between 14 and 28 days post SIV infection compared to animals infected for ∼1 week ( Figure 9B , P = 0 . 057 ) . These data are consistent with early ( 14–28 dpi ) and progressive damage to the integrity of the epithelial barrier , and indicate that one mechanism underlying microbial translocation likely involves breakdown of the structural barrier of the GI tract at a rate that exceeds enterocyte proliferation and other repair mechanisms , consistent with previous reports of abnormalities within the GI tracts of chronically HIV or SIV-infected individuals [31] , [32] , [33] , [34] , [35] , [36] , [37] , [38] . Our findings of host responses to microbial-product infiltration into the lamina propria are consistent with previous findings of fibrosis within the lamina propria of the GI tract [39] . While our data and those of previous studies are consistent with increased discontinuity of the structural barrier of the GI tract during chronic SIV infection of RMs , we cannot exclude that the structural damage we observed by immunohistochemical analysis may be attributed to increased enterocyte turnover overall , leading to an apparent increase in structural damage to the GI tract . However , the increased levels of myeloperoxidase+ PMNs within the lamina propria of chronically SIV-infected RMs associated with observed damage to the epithelial barrier ( Figure S9 ) , strongly support our conclusion of GI epithelial damage . Regardless , these data suggest that the integrity of the structural barrier of the GI tract is significantly weakened in SIV-infected individuals leading to microbial translocation . When microbial products cross the epithelial barrier under normal , physiological conditions , they are generally phagocytosed by specialized intestinal macrophages [40] . The relatively abundant and apparently non-cell associated microbial constituents we saw in the LP of our chronically SIV-infected RMs , particularly in animals with AIDS , suggested that this might result from microbial translocation in excess of the phagocytic capacity of these macrophages . This could reflect a saturation of the maximum capacity of these macrophages to phagocytose translocated microbial constituents or alternatively , might reflect a compromise of macrophage phagocytic function resulting in a net relative defect in microbial clearance and the extracellular accumulation of microbial products . To assess this , we performed confocal microscopy of GI tract tissue from SIV-uninfected and acute and chronically SIV-infected RMs to assess the localization of microbial constituents relative to GI tract macrophages ( intracellular vs . extracellular ) . In the rare instances where microbial products were detected in the LP of SIV-uninfected RMs , they were virtually always within HAM56+ macrophages ( Figure 10 and Figure S11 ) . In addition , during the acute phase of infection ( until 28 dpi ) , most microbial products were mostly found within HAM56+ macrophages ( Figure S11 ) , perhaps helping to explain why LPS levels in plasma are not elevated during acute infection while sCD14 levels were moderately raised [9] . Moreover , microbial products crossing the epithelial barrier at 14 dpi were virtually always found within macrophages , whereas abundant macrophages were juxtaposed to damaged regions of the colon in late acute ( 28 dpi ) ( Figure S11 ) . Furthermore , as infection progressed into the chronic stage of disease , the frequency of macrophages negative for microbial products increased , even though abundant numbers of macrophages were present and adjacent to microbial products ( Figure 10 and Figure S11 ) . There was no apparent change in the overall frequency of HAM56+ macrophages at the tips of the colonic crypts between SIV-uninfected , and acutely or chronically SIV-infected RMs ( data not shown ) . The presence of high frequencies of macrophages without internalized microbial constituents , along with abundant extracellular microbial products suggested that GI tract macrophages in the later phase of acute infection and chronic stages of disease may become increasingly incapable of phagocytosing microbial products that translocate into the LP . A distinguishing feature of non-progressive infection in natural hosts of SIV is the absence of immune activation during the chronic phase of infection [6] , [41] , [42] , [43] , [44] , [45] . Because elevated plasma LPS levels are absent in both chronically SIVsmm-infected SMs and SIVagm-infected AGMs [9] , [45] , [46] , we evaluated the structural integrity of the intestinal epithelial barrier in chronically SIVsmm-infected SMs . In contrast to findings in chronically SIV-infected RMs , and consistent with the lack of LPS within the circulation of these SIV natural host animals [9] , we found no evidence of damage to the integrity of the epithelial barrier ( Figure 11A ) and no infiltration of microbial products into the LP of large bowel ( Figure 11B ) or peripheral lymph nodes ( Figure 11C ) . These data were consistent among 7 animals studied ( n = 2 , SIVsmm-uninfected SMs; n = 5 , chronically SIVsmm-infected SMs; Table 1 ) . Hence preservation of the tight epithelial barrier is associated with lack of microbial translocation and immune activation in non-progressive , natural SIV infection .
Indirect evidence has implicated microbial translocation from the gut as a factor contributing to pathological immune activation in chronic HIV/SIV infection , but direct evidence of translocation and demonstration of a plausible underlying mechanism have been lacking . Using an unbiased , comprehensive approach for quantitative and qualitative immunohistologic analysis of randomly selected tissue specimens obtained from non-human primates at various times relative to SIV infection , we have shown that: 1 ) microbial products can be found in the LP of the large bowel , in draining and distant lymph nodes , and in the liver of chronically SIV-infected RMs; 2 ) microbial translocation is associated with breakdown of the integrity of the epithelial barrier of SIV-infected RMs; 3 ) the extent of epithelial breakdown correlates with the extent of microbial translocation; 4 ) epithelial barrier breakdown and microbial translocation begin to be apparent during the late acute phase of infection ( 14 dpi ) ; 5 ) the presence of microbial products in multiple anatomical sites is associated with expression of IFN-α and IL-18 in the absence of detectable local viral replication in the LP , consistent with direct induction of immune activation; 6 ) macrophages in chronically SIV-infected RMs appear dysfunctional with respect to their ability to phagocytose translocated microbial products; and 7 ) neither epithelial barrier breakdown nor infiltration of microbial products into the LP occur during the chronic phase of SIV infection of SMs . We provide two lines of evidence linking microbial translocation to immune activation . First , we show that in pathogenic SIV infection of RMs , damage to the integrity of the epithelial barrier of the GI tract is associated with microbial translocation , and that microbial translocation is linked to local immune activation , based on co-localization of microbial products and production of the immunoinflammatory cytokines IFNα and IL-18 , including in lymph nodes anatomically distant from the GI tract . Second , in marked contrast , in SIV-infected SMs where immune activation is quickly resolved in the acute stage of infection [47] , and chronic infection is not accompanied by persistent immune activation , we found neither damage to the intestinal barrier nor microbial translocation . Recent in vitro studies with peripheral blood lymphocytes from SMs have been interpreted to suggest that a lack of type I IFN cytokine response to SIV RNA accounts for the typically nonprogressive nature of SIV infection in SMs [48] . Furthermore , these data have been used to suggest that the raised plasma LPS levels observed in chronically SIV-infected RMs and HIV-infected humans , are simply markers of damage to the GI tract and that the microbial translocation that is reflected in elevated plasma LPS levels does not contribute significantly to causing systemic immune activation [48] . However not only were LPS levels increased in chronically-infected individuals , but sCD14 and LPS binding protein levels were also increased [9] , [11] , [13] . These data strongly suggested that LPS was directly stimulating the immune system in vivo . In the present study , we show directly that damage to the integrity of the epithelial barrier of the GI tract allows microbial products to infiltrate into the LP and this infiltration is associated with local immune activation demonstrated by co-localization of microbial products within the LP with IFNα and IL-18 in MesLN . As only limited viral replication is demonstrable in the LP of the GI tract of chronically SIV-infected RMs , the damage to the GI epithelial lining , microbial translocation and local immune activation are unlikely to be caused by the direct effects of local viral replication . Rather , where rare infected cells are seen in the LP underlying damaged mucosa , it is more likely that the chronic immune activation , due to translocated microbial products , has provided activated CD4+ T cell targets for the virus . Taken together , these data suggest that microbial translocation , resulting from damage to the GI tract epithelial barrier and impaired macrophage-mediated phagocytosis , results in immune activation during the chronic phase of HIV/SIV infection of humans and RMs . Importantly , we found a significant correlation between the extent of damage to the epithelial barrier of the colon and the amount of LPS within the underlying mucosa and the extent of translocated microbial products in draining and remote lymph node tissues . The extent of microbial constituents present in lymph node tissues correlated with the extent of local evidence of immune activation , further substantiating the link . Moreover , while we find that microbial translocation begins during the acute phase of infection , our previous work had indicated that elevated levels of microbial products were not seen in plasma until the chronic phase [9] . Our data suggest that microbial products are localized within tissue macrophages during the acute phase thus limiting their circulation . The causes of damage to the integrity of the epithelial barrier of the GI tract are likely to be multifaceted , but in the chronic stages of SIV infection seem unlikely to be due to direct virotoxic effects , given the lack of association with very low levels of demonstrable local viral replication in the LP relative to the extensive epithelial damage . One possible mechanism may be related to the preferential loss of Th17 cells in the GI tract in progressive immunodeficiency lentiviral infections [25] , [26] , because Th17 cells produce cytokines important for enterocyte proliferation and antibacterial defensins [49] , [50] , [51] and IL-17 has recently been shown to suppress Th1-mediated damage to gut epithelium . Importantly , preservation of this T cell subset in the gut of chronically SIVsmm-infected SMs and SIVagm-infected AGMs [26] , [52] and the sparing of the epithelial barrier of SIVsmm infected SMs we show here supports this mechanism . We speculate that the association we found between immune activation , microbial translocation and chronic stages of SIV infection , and similarly , later stages of HIV-1 infection , reflects damage to the structural integrity of the GI tract and a potential “deficiency” of the GI tract macrophage-phagocytic system . Our observation that intestinal macrophages from SIV-infected RMs , which are generally not proinflammatory [40] , are unable to clear translocated microbial products , within the LP , and could lead to increased proinflammatory responses locally are supported by several groups findings that showed: i ) impaired monocyte phagocytosis in HIV-infected individuals [53]; ii ) reduced LPS-mediated enhancement of phagocytosis in monocytes HIV-infected individuals compared to healthy donors [54]; and iii ) significantly higher colonic mucosa proinflammatory mRNA expression levels ( e . g . TNF-α , IFN-γ , and IL-6 ) in HIV-infected patients than in control patients [55] . These data certainly warrant further investigation into the functional properties of tissue macrophages from HIV/SIV-infected individuals and the mechanisms underlying their apparent dysfunction . While increased microbial translocation begins in the late acute stage of SIV infection , it was not until the later stages of infection that the capacity of macrophages for clearance was apparently affected , suggesting that microbial translocation has an increasing major contribution to immune activation as the host progresses towards disease . The evidence for this model comes from images of MesLN and AxLN stained for bacterial products in the chronic stage that showed dramatically increased extracellular bacterial constituents in the late AIDS stage of SIV infection , versus the mainly cellular staining of LPS at earlier stages . Taken together , these data strongly suggest that in SIV infection of RMs , and by extension , HIV infection of humans , damage to the epithelial barrier of the GI tract leads to levels of microbial translocation that exceed the capacity of host defense mechanisms to sequester away microbial constituents from secondary lymphatic tissues , resulting in persistent immune activation that contributes importantly to pathogenesis during the chronic phase of infection . Understanding the factors underlying damage to the integrity of the epithelial barrier and macrophage deficiencies that we report may lead to novel therapeutic interventions that aim to reduce microbial translocation and the deleterious effects of the consequent immune activation .
To characterize the extent of microbial translocation in the gastrointestinal tract in SIV infection , we studied tissues from an assembled cohort of SIV-uninfected and infected RMs and SIVsmm-infected and uninfected SMs originally involved in separate studies ( summarized in Table 1 ) . Tissues ( colon and LNs ) were obtained at necropsy from 13 rhesus macaques ( Macaca mulatta ) of Indian origin euthanized 1–28 days after atraumatic intravaginal infection with SIVmac251 or SIVmac239 as described elsewhere [56] . Six additional SIV-negative RMs were used as controls . In a separate study , tissues were obtained at necropsy from 13 adult RMs chronically infected with SIVmac239 or SIVsmE660 that were sacrificed either because of end-stage disease ( AIDS; defined by opportunistic infections , lymphomas , or a diagnosis of wasting , based on greater than 15% body mass weight loss , n = 5 ) or protocol specified experimental end point ( Non-AIDS; n = 8 ) . For immunohistochemistry studies , samples were very quickly processed into fixative to avoid potential artifacts associated with post-mortem tissue changes . The post mortem interval , the time from euthanasia until collection of lymph nodes and GI tract segments were placed into fixative , ranged from ∼10–30 minutes and was consistent over four independent primate facilities contributing tissues to the present study . The GI tract segments sampled at necropsy were representative and were not selected with regard to any visually apparent lesions or other pathology . In a third study , LNs and rectal biopsies were obtained from 2 SIVsmm-negative SMs ( Cercocebus atys ) and 5 SMs that were naturally infected with SIVsmm as previously described [47] . Animals were housed and cared in accordance with American Association for Accreditation of Laboratory Animal Care standards in AAALAC accredited facilties , and all animal procedures were performed according to protocols approved by the Institutional Animal Care and Use Committees of the National Cancer Institute , California National Primate Research Center or Yerkes National Primate Research Center ( Table 1 ) . Unfortunately , paraffin-embedded tissue from colon , mesLN , and axLN samples were not available from all animals . Plasma samples were analyzed for SIV vRNA by using a quantitative branched DNA ( bDNA ) assay [53] or using a fluorescent resonance energy transfer probe-based real-time RT-PCR ( TaqMan ) assay that provides a threshold sensitivity of 125 copy Eq/ml , as previously described [57] . All PCR reactions were run on ABI Prism 7700 Sequence Detection System and the fluorescent signal-based quantitation of viral RNA copy numbers in test samples was determined by ABI sequence detection software ( Applied Biosystems , Foster City , CA ) . Immunohistochemical staining and SIV in situ hybridization were performed as previously described [58] . In brief , unselected specimens of tissues of interest were obtained at necropsy , fixed , and paraffin embedded . Immunohistochemistry was performed using a biotin-free polymer approach ( MACH-3; Biocare Medical ) on 5-µm tissue sections mounted on glass slides , which were dewaxed and rehydrated with double-distilled water . Antigen retrieval was performed by heating sections in 1× DIVA Decloaker reagent ( Biocare Medical ) in a pressure cooker ( Biocare Medical ) followed by cooling to room temperature . All slides were stained using the intelliPATH FLX autostaining system ( Biocare Medical ) according to experimentally determined optimal conditions . This included blocking tissues with Blocking Reagent ( Biocare Medical ) for 10 min followed by an additional blocking step with TNB ( 0 . 1 M Tris-HCL ( pH 7 . 5 ) , 0 . 15 M NaCl , and 0 . 5% Blocking Reagent ( NEN ) ) containing 2% Blocking Reagent and 100 µg/mL goat ChromePure IgG ( Jackson Immunoresearch ) for 10 minutes at room temperature . Endogenous peroxidase was blocked with 1 . 5% ( v/v ) H2O2 in TBS ( pH 7 . 4 ) . Primary antibodies were diluted in TNB containing 2% Blocking Reagent and 100 µg/mL goat ChromePure IgG for 1 h at room temperature . Mouse or rabbit MACH-3 secondary polymer systems ( Biocare Medical ) were applied for 20 minutes each . Double immunohistochemical staining was performed on colon and lymph node sections with either mouse monoclonal anti-LPS-core and rabbit polyclonal anti-IL18 antibodies or mouse monoclonal anti-IFNα and rabbit polyclonal anti-E . coli using the MACH-2 multiplex staining system ( Biocare Medical ) according to manufacturer's instructions . Sections were developed with ImmPACT DAB ( Vector Laboratories ) and/or Vulcan Fast Red chromogen ( Biocare Medical ) , counterstained with hematoxylin , and mounted in Permount ( Fisher Scientific ) . All stained slides were scanned at high magnification ( 400× ) using the ScanScope CS System ( Aperio Technologies , Inc . ) yielding high-resolution data from the entire tissue section . Bacterial PNA FISH was performed using the universal bacterial 16 s ribosomal RNA specific ( UniBac ) FITC conjugated PNA probe ( AdvanDx , Inc . ) according to manufacturer's instructions , with the exception that a heat induced epitope retrieval pretreatment step was performed in 1× Diva retrieval buffer ( Biocare Medical ) for 20 minutes in a 95°C water bath prior to hybridization . FISH samples were examined and imaged using a Nikon 80i upright fluorescent microscope ( Nikon Instruments , Inc ) equipped with a BrightLine multiband bandpass FITC/Texas Red filter ( Semrock; data not shown ) . Primary antibodies used were: mouse anti-human Interferon-α ( clone MMHA-2; PBL InterferonSource ) , mouse anti-LPS core ( clone WN1 222-5; Hycult or provided by Dr . Robin Barclay ) , mouse anti-macrophage ( clone HAM56; Dako ) , mouse anti-cytokeratin ( clone MNF116; Dako ) , polyclonal rabbit anti-E . coli ( Dako ) , polyclonal rabbit anti-IL-18 ( Sigma Prestige Antibodies by Atlas Antibodies ) and polyclonal rabbit anti-Claudin-3 ( Labvision ) . Immunofluorescent confocal microscopy was performed on treated slides as above , but stained overnight with primary antibodies at 4°C , washed , stained with fluorescently conjugated secondary antibodies for 1 h in the dark , counterstained with DAPI ( Molecular Probes ) , mounted in AquaPoly mount ( Polysciences , Inc . ) and imaged using a Olympus FluoView FV1000 . Z-stack images were taken for each high power field that spanned the entire 5 µm tissue section and representative 3-D projections from z-stack images were generated using Imaris 7 . 0 . 0 software ( Bitplane Inc . ) . Secondary antibodies used were donkey anti-mouse IgG Alexa Fluor 488 and donkey anti-rabbit IgG Alexa Fluor 555 ( all from Molecular Probes ) . To quantify microbial translocation ( LPS ) into the LP and the extent of epithelial barrier damage ( claudin-3 ) , 5-µm thick sections were cut from paraffin blocks of unselected tissue sections obtained at necropsy and stained with either monoclonal antibody against LPS-core or polyclonal antibody for claudin-3 and counter stained with heamotoxalin . High power ( 400× ) whole tissue scans were obtained using an Aperio ScanScope as described above and imported into Photoshop CS3 ( Adobe Systems Inc . , Mountain View , California , USA ) . Images were manually trimmed to remove the submucosae , muscularis and residual luminal content , leaving only the LP mucosae to analyze . The percent area of the LP staining for LPS was determined essentially as previously described using Photoshop CS3 tools with plug-ins from Reindeer Graphics [39] , [47] . The percent area of LN staining for LPS was determined from whole LN scans as above but without the need to trim the image . The proportion of the epithelial barrier that was damaged during SIV infection was first determined by manually tracing ( in red ) the area of the lumen/GI epithelial tract interface that had no claudin-3 staining epithelial cells , using the brush tool in Photoshop CS3 . The remaining claudin-3 staining intact epithelial cell regions were then manually traced ( in black ) . The percent damage was calculated by determining the proportion of the image that was red ( lack of claudin-3 stain ) compared to the total epithelial surface area ( red+black ) using plug-in tools from Reindeer Graphics . Spearman's rank correlation and Mann-Whitney tests were performed using Prism 4 . 0 software ( Prism , San Diego , CA ) .
|
Persistent activation of the immune system is a hallmark of chronic HIV/SIV infections and predicts disease progression better than either plasma viral load or CD4+ T cell count . While the causes of immune activation during chronic infection are likely multifactorial , recent work has shown that microbial translocation is associated with immune activation . However , direct , tissue level in vivo evidence of translocation and the underlying mechanisms remain unclear . Here , we sought direct in vivo evidence of translocation , and an understanding of the timing and the underlying mechanisms . We found that in RMs , microbial translocation begins during the late acute phase of SIV infection and increases progressively during chronic infection and is associated with structural damage of the GI tract . We further discovered that immune activation is temporally and causally related to microbial translocation and by the relative inability of intestinal macrophages to bind/phagocytose translocated microbial products . In SIV-infected sooty mangabeys , however , no evidence of epithelial barrier breakdown , nor increased microbial translocation or chronic immune activation were observed . Our results provide direct evidence for microbial translocation in vivo , coupled with early and progressive intestinal epithelial damage , and eventual impairment of macrophage clearance associated with dissemination of microbial products and systemic immune activation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/hiv",
"infection",
"and",
"aids"
] |
2010
|
Damaged Intestinal Epithelial Integrity Linked to Microbial Translocation in Pathogenic Simian Immunodeficiency Virus Infections
|
The highly variable tprK gene of Treponema pallidum has been acknowledged to be one of the mechanisms that causes persistent infection . Previous studies have mainly focused on the heterogeneity in tprK in propagated strains using a clone-based Sanger approach . Few studies have investigated tprK directly from clinical samples using deep sequencing . We conducted a comprehensive analysis of 14 primary syphilis clinical isolates of T . pallidum via next-generation sequencing to gain better insight into the profile of tprK in primary syphilis patients . Our results showed that there was a mixture of distinct sequences within each V region of tprK . Except for the predominant sequence for each V region as previously reported using the clone-based Sanger approach , there were many minor variants of all strains that were mainly observed at a frequency of 1–5% . Interestingly , the identified distinct sequences within the regions were variable in length and differed by only 3 bp or multiples of 3 bp . In addition , amino acid sequence consistency within each V region was found among the 14 strains . Among the regions , the sequence IASDGGAIKH in V1 and the sequence DVGHKKENAANVNGTVGA in V4 showed a high stability of inter-strain redundancy . The seven V regions of the tprK gene in primary syphilis infection demonstrated high diversity; they generally contained a high proportion sequence and numerous low-frequency minor variants , most of which are far below the detection limit of Sanger sequencing . The rampant variation in each V region was regulated by a strict gene conversion mechanism that maintained the length difference to 3 bp or multiples of 3 bp . The highly stable sequence of inter-strain redundancy may indicate that the sequences play a critical role in T . pallidum virulence . These highly stable peptides are also likely to be potential targets for vaccine development .
Syphilis , caused by Treponema pallidum subsp . pallidum , is an ancient sexually transmitted disease that was initially recognized in the 15th century and is a public health threat that cannot be neglected [1 , 2] . The completion of the first whole-genome sequencing of the Nichols strain of T . pallidum provided a wealth of information about the characteristics of this pathogen [3] , since then the sequence of other laboratory treponemal strains has also been released [4–12] . These particular achievements have revealed slight variations among different strains in a small genome ( ∼1 . 1 Mb ) , and most of the genetic diversity occurs in six genomic regions , including a polymorphic multigene family encoding 12 paralogous proteins ( tpr A through tprL ) , highlighting most likely a factor in the pathogenesis of T . pallidum [2 , 6 , 13] . Within the tpr family , the antigen-coding tprK has been found to be the direct target of the human immune response [14] , although its surface exposure has been challenged and remains to be fully confirmed [15–17] . Several remarkable studies performed in the rabbit model have demonstrated that the tprK gene possesses high genetic diversity at both the intra- and inter-strain levels , and the genetic variation in tprK is localized to seven variable regions ( V1-V7 ) flanked by highly conserved domains [18–20] . Theoretically , through gene conversion , variations in the V regions would generate millions of chimeric tprK variants , resulting in a constant alteration in the T . pallidum antigenic profile [21] . Therefore , the tprK gene is acknowledged to have a pivotal role in immune evasion and pathogen persistence [22 , 23] . Previous studies focusing on the genetic variability of tprK were mainly based on the clone-based Sanger approach; when using this approach , it would inevitably encounter a bottleneck in clone selection where minor variants , especially at low frequencies , are lost; consequently , the complete mutation profile of tprK is not fully understood . Furthermore , except for one recent publication that reported on whole-genome sequencing directly from clinical samples of T . pallidum to investigate how tprK diversifies in the context of human infection [24] , other tprK-related studies were conducted based on rabbit-derived samples [18 , 19 , 25 , 26] . In the present study , we seek to systematically reveal the profile of tprK in T . pallidum directly from patients with primary syphilis by employing next-generation sequencing ( NGS ) , thus providing important insights into the understanding of the diversity of tprK directly from primary syphilis patients and contributing to further explorations of the mechanisms of long-term T . pallidum infection .
All participants in this study were adults and written consent was obtained with signatures from all patients in accordance with institutional guidelines prior to the study . The study was approved by the Ethics Committee of Zhongshan Hospital , Xiamen University , after a formal hearing and was in conformance with the Declaration of Helsinki . Swab samples were obtained from the skin lesions of 14 patients ( X-1~14 ) with primary syphilis . The clinical diagnosis of syphilis was based on the US Centers for Disease Control and Prevention ( CDC ) [27] and the European CDC ( ECDC ) guidelines [28] . DNA was extracted from the swab samples using the QIAamp DNA Mini Kit ( Qiagen , Inc . , Valencia , CA , USA ) according to the manufacturer’s instructions , and careful precautions were implemented to avoid DNA cross-contamination between isolates [11] . Each sample was quantified by targeting tp0574 through qPCR using a 96-well reaction plate with a ViiA 7 Real-Time PCR System ( Applied Biosystems , USA ) . For the absolute quantification of treponemal copies , a standard curve was constructed using 10-fold serial dilutions of cloned plasmids ( for tp0574 ) generated through TOPO TA technology ( Invitrogen , Carlsbad , CA , USA ) and transformation of DH5α Escherichia coli cells [29] . The DNA samples that tested positive were used to amplify tp0136 to determine whether these 14 clinical stains belong to the Nichols-like group or SS14-like group [30] . First , the extracted DNA was directly used in the amplification of the tprK full open reading frame ( ORF ) . The primers used for the amplification are listed in S1 Table . For amplification , KOD FX Neo polymerase ( Toyobo , Osaka , Japan ) was used . The reaction mixture contained 25 μL of 2× PCR buffer , 0 . 4 mM deoxynucleotide triphosphates , 0 . 3 μM of each primer , 1 U of KOD FX Neo polymerase , and 5 μL of genomic DNA in a final volume of 50 μL . The cycling conditions were as follows: 94°C for 2 min , followed by 40 cycles of 98°C for 10 s , 60°C for 30 s , and 68°C for 30 s . Then , the amplicons were gel purified and stored at -20°C for further processing as the template for segmented amplification described below . Second , partial amplification of four fragments of 400–500 bp , overlapping by at least 20 bp , covered tprK ORF . The primers are listed in S1 Table . The purified full length tprK amplicons were diluted 1000-fold and used as a segmented amplification template . The amplification mixture was the same as described above except that the primers were 0 . 15 μM . The cycling conditions were denaturation at 94°C for 2 min , followed by 30 cycles of 98°C for 10 s , 55°C for 30 s , and 68°C for 30 s . The size of all the products was verified by 2% agarose gel electrophoresis , and the products were gel purified . The four subfragments corresponding to each sample were mixed in equimolar amounts into one pool to produce a separate library using a barcode to distinguish each sample . Library construction and sequencing were performed by the Sangon Biotech Company ( Shanghai , China ) on the MiSeq platform ( Illumina , San Diego , CA , USA ) in paired-end bi-directional sequencing ( 2×300 bp ) mode . FastQC ( http://www . bioinformatics . babraham . ac . uk/project/fatsqc/ ) and FASTX ( http://hannonlab . cshl . edy/fastx_toolkit ) tools were applied to check and improve the quality of the raw sequence data , respectively . The final reads collected from 14 patients were compared with the tprK of the Seattle Nichols strain ( GenBank accession number AF194369 . 1 ) using Bowtie 2 ( version 2 . 1 . 0 ) . Based on the previously published principle that was used to extract sequence [24] , an in-house Perl script was developed and applied to specifically capture DNA sequences within seven regions of the tprK gene across 14 strains from raw data , both forward and reverse . Briefly , the user-defined strings that matched the conserved sequence flanking the variable regions were used to catch the variable sequences . The defined strings referred to the mapping result of the reference and should be as long as necessary to ensure specificity ( approximately 12–16 bp ) . Thus , the exact number of distinct sequences within seven regions of the tprK gene from each sample was acquired . The intrastrain heterogeneous sequences were valid if the following conditions were simultaneously verified for any variant sequence: 1 ) being supported by at least fifty reads and 2 ) displaying a frequency above 1% . Then , the relative frequency of the sequences within each variable region was calculated . The raw data sequences of these 14 primary syphilis samples were deposited in the SRA database ( BioProject ID: PRJNA498982 ) under following BioSample accession numbers: SAMN10340238- SAMN10340251 for X-1-X-14 , respectively .
The samples ( N = 14 ) were collected from patients diagnosed with primary syphilis at Zhongshan Hospital , Xiamen University . The clinical data of patients are shown in Table 1 . The qPCR data of tp0574 showed that the number of treponemal copies in each clinical sample was eligible for the amplification of the tprK full ORF . And based on the sequencing data of tp0136 , most of them belonged to SS14-like group and only two belonged to the Nichols-like group . The median sequencing depth of the tprK segment samples ranged from 10568 . 99 to 56676 . 38 , and the coverage ranged from 99 . 34% to 99 . 61% , showing high identity with the tprK gene of the Seattle Nichols strain . Nucleotide sequences found in variable regions were translated into amino acid sequences in silico . In eight cases , two or more amino acid sequences were found to be identical in one sample although they were coded by different nucleotide sequences ( S3 and S4 Tables ) . No sequence yielded a tprK frame shift or premature termination . When distinct sequences within each V region from each strain were compared , a scenario of sequence consistency was found . As Fig 3 shows , V1 and V4 presented a strong shared sequence capacity . The sequence IASDGGAIKH in V1 was observed in five strains ( 5/14 ) and DVGHKKENAANVNGTVGA in V4 was shared across seven strains ( 7/14 ) . However , the parallel sequences in V3 and V6 did not seem as significant as in other V regions , especially in V6 . To further explore whether the shared scenario was usually displayed by the predominant sequence across all the strains , we involved only the predominant sequence in the V region of each sample , which was represented by the bold arc in Fig 3 and found that V1 and V4 still presented similar shared sequence abilities despite the decreased redundant sequences . The occurrence of the consistent sequence in V1 and V4 could reach five strains and six strains , respectively ( Table 2 ) . For the V3 and V6 regions , which were rarely consistent with sequences , the shared sequence in V3 occurred only between two strains , and there was no consistent sequence found in V6 . Meanwhile , there was also no redundant sequence observed in V7 .
Although a recent landmark study has reported the successful long-term in vitro propagation of T . pallidum [31] , research on this pathogen has been greatly hindered by the lack a system for genetic manipulations in past decades [19 , 32] . The whole genome sequencing of the Nichols strain of T . pallidum provided a new perspective for the study of treponemal genes and proteins . Among these genes , tprK has been extensively studied because of its highly variable antigenic profile . In the present study , we performed NGS , a more sensitive and reliable approach , to gain better insight into the profile of tprK in primary syphilis patients . Overall , there was a sequence mixture concentrated on seven variable regions of tprK in primary syphilis samples . Among the seven V regions , V1 and V6 were found to have the lowest and highest variability , respectively ( Figs 1 and 2A ) , which was consistent with the findings of previous studies [24 , 33] . Although tprK was previously revealed to have rampant genetic diversity within each strain , the exact proportion of these variant sequences within one strain would not be clearly known by using previous clone-based Sanger approach [18 , 19 , 25] . In fact , we also applied the clone-based Sanger approach to analyse the tprK gene in this research . As described in Pinto et al . ’ study [24] , it generally displayed the predominant sequence within each V region ( consistent with the sequence found by NGS ) but could not identify all the minor variants ( S1 Fig ) . However , it is an advantage of NGS to fully discover the variants [34 , 35] . Combined with the use of an in-house Perl script , we were able to retrieve the variants within the regions of each strain and calculate the relative frequency of the variants , thus disclosing the proportion of these variant sequences in primary syphilis patients . As shown in Fig 2 , the distribution of variants within the V regions of tprK from primary syphilis patients reveals that the vast majority of them have a high proportion of predominant sequences ( frequency above 80% ) and numerous minor variants ( frequency below 20% ) , but very few sequences have a frequency between 20% and 80% . Moreover , these minor variants were found to be mostly distributed at a frequency of 1–5% ( Fig 2B ) , which was extremely below the detection limit for Sanger sequencing [36] . This feature may represent a logical fitness-based evolution where high-frequency sequences are better fitted to avoid immune recognition and numerous low-frequency minor variants may simply emerge and most of them would likely disappear if they were not advantageous for syphilis developing [37] . It is worth noting that the sequences appearing between the frequency of 20–80% were mainly concentrated in the V2 , V5 , V6 and V7 regions mostly from X-6 , 8 , 10 , 13 ( Fig 2 ) . The distribution pattern of these variants from these samples may suggest that with disease progression or increasing immunity , the balance of the original sequence distribution was broken and some V regions ( e . g . , V2 , V5 , V6 and V7 ) began to change . As a result , a minor variant ( or a new variant ) became advantageous and its frequency gradually increased , ultimately replacing the original predominant sequence , which further promoted the antigenic diversity of TprK for T . pallidum to escape immune clearance and potentially leading to the development of late syphilis , neurosyphilis or serofast status [15 , 21 , 38 , 39] . Additionally , among these four V regions , the frequency of the predominant sequence in V6 was particularly low ( Fig 2A ) , suggesting that V6 may be the first affected region and is involved in immune evasion during the course of infection [21 , 24] . In this study , besides the distinct variations in tprK sequences , we also found length heterogeneity in this gene ( Fig 1 ) . The size range of the captured sequences was the largest for V3 , V6 and V7 , which was similar to the findings of Pinto et al . [24] , demonstrating that the variations in these three regions could more easily cause changes in length . Nevertheless , the diversity of length forms was much lower than the diversity of variants within each V region . Especially in the V5 region , there were many different variants observed , but only two lengths ( 84 and 90bp ) were present , which was also observed in the previous study [21] . Additionally , it was interesting that the length of all distinct sequences differed by only 3 bp or multiples of 3 bp , and previous research data also supported this pattern change [21 , 24] . A change pattern characterized multiple of 3 bp matched the triplet codon in protein coding , which has made us think that this feature probably explains why it is rare to uncover a tprK frame shift . In fact , no frameshifts were detected in our research and only one was detected in the study of Pinto et al . [24] . Additionally , synonymous nucleotide sequence of tprK was rare and was found only in the V2 and V5 regions ( S3 and S4 Tables ) , in accordance with the study by Pinto et al . [24] . These phenomena suggest that the rampant diversity of tprK could be regulated by a strict gene conversion mechanism to avoid yielding an abnormal detrimental antigen for T . pallidum . A dominant amino acid sequence for a specific V region in one patient depends on the immune response of that specific patient . For this reason , it may be difficult to find out several syphilitic patients for which the amino acid sequences for some V regions are exactly the same . Actually , despite the significant polymorphic characteristic of tprK , at least half of the strains had sequences shared by other strains ( Fig 3 ) in our study , which was similar to previous findings [24] . And tprK inter-population redundancy was maintained at a high level in V1 and V4 in contrast to other regions , especially when only the predominant sequence within each V region was analysed ( Table 2 ) . Interestingly , the most stable amino acid sequence ( IASDGGAIKH ) of inter-population redundancy in V1 among 14 primary syphilis patients was also found to be the most frequent sequence in the 24 syphilis patients in Pinto’s study [24] . And the sequence ( DVGHKKENAANVNGTVGA ) in V4 was also found at a moderate proportion in share among the 24 clinical samples . The similar findings that were observed between the two studies using different approaches to investigate the adaptive traits of the pathogen during different human infection were exciting and clearly suggest the existence of better fitted antigenic profiles to address the immune response of the host . In previous studies [15 , 38 , 40] , the molecular localization in the N-terminal region of tprK was conformed to displayed promising partial protection in a rabbit model . Therefore , the highly stable shared peptide of V1 and V4 across all the strains would likely be a potential target for vaccine development . Finally , the limitations of our research should be discussed . First , the findings reported above were based on amplicons of tprK . The possible introduction of errors by polymerases used for the amplification of templates for NGS could not be ignored , although the data showed that the error was minimal . Second , this study provides information on individual V regions instead of information on a single tprK ORF . It would not be correct to assume that certain nucleotide sequences within the V regions are derived from a same single tprK ORF , as this would result in artificial sequences . In summary , the characteristic profile of tprK in primary syphilis patients was unveiled to generally contain a high proportion sequence and many low-frequency minor variants within each V region . The variations in V regions were regulated by a strict gene conversion mechanism to keep the length differences to 3 bp or multiples of 3 bp . The findings could provide important information for further exploration of the role of tprK in immune evasion and persistent infection with syphilis . Furthermore , the peptides in each V region , especially the highly conserved peptides found in this study , could serve as a database of B cell epitopes of TprK for human immunological studies in the future .
|
Variations in tprK have been acknowledged to be the major contributors to persistent Treponema pallidum infections . Previous studies were based on the clone-based Sanger approach , and most of them were performed in propagated strains using rabbits , which could not reflect the actual heterogeneous characteristics of tprK in the context of human infection . In the present study , we employed next-generation sequencing ( NGS ) to explore the profile of tprK directly from 14 patients with primary syphilis . Our results showed a mixture of distinct sequences within each V region of tprK in these clinical samples . First , the length of identified distinct sequences within the region was variable , which differed by only 3 bp or multiples of 3 bp . Then , among the mixtures , a predominant sequence was usually observed for each V region , and the remaining minor variants were mainly observed at a frequency of 1–5% . In addition , there was a scenario of amino acid sequence consistency within the regions among the 14 primary syphilis strains . The identification of the profile of tprK in the context of human primary syphilis infection contributes to further exploration of the pathogenesis of syphilis .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"sequencing",
"techniques",
"urology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"population",
"genetics",
"tropical",
"diseases",
"microbiology",
"treponematoses",
"bacterial",
"diseases",
"next-generation",
"sequencing",
"sexually",
"transmitted",
"diseases",
"genome",
"analysis",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"dna",
"dna",
"recombination",
"gene",
"conversion",
"population",
"biology",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"infectious",
"diseases",
"genomics",
"genetic",
"polymorphism",
"bioinformatics",
"medical",
"microbiology",
"microbial",
"pathogens",
"biological",
"databases",
"molecular",
"biology",
"nucleotide",
"sequencing",
"biochemistry",
"treponema",
"pallidum",
"sequence",
"databases",
"nucleic",
"acids",
"genitourinary",
"infections",
"database",
"and",
"informatics",
"methods",
"transcriptome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"evolutionary",
"biology",
"dna",
"sequencing",
"protein",
"sequencing",
"syphilis"
] |
2019
|
Profile of the tprK gene in primary syphilis patients based on next-generation sequencing
|
Gut microbiota are shaped by a combination of ecological and evolutionary forces . While the ecological dynamics have been extensively studied , much less is known about how species of gut bacteria evolve over time . Here , we introduce a model-based framework for quantifying evolutionary dynamics within and across hosts using a panel of metagenomic samples . We use this approach to study evolution in approximately 40 prevalent species in the human gut . Although the patterns of between-host diversity are consistent with quasi-sexual evolution and purifying selection on long timescales , we identify new genealogical signatures that challenge standard population genetic models of these processes . Within hosts , we find that genetic differences that accumulate over 6-month timescales are only rarely attributable to replacement by distantly related strains . Instead , the resident strains more commonly acquire a smaller number of putative evolutionary changes , in which nucleotide variants or gene gains or losses rapidly sweep to high frequency . By comparing these mutations with the typical between-host differences , we find evidence that some sweeps may be seeded by recombination , in addition to new mutations . However , comparisons of adult twins suggest that replacement eventually overwhelms evolution over multi-decade timescales , hinting at fundamental limits to the extent of local adaptation . Together , our results suggest that gut bacteria can evolve on human-relevant timescales , and they highlight the connections between these short-term evolutionary dynamics and longer-term evolution across hosts .
The gut microbiome is a complex ecosystem comprised of a diverse array of microbial organisms . The abundances of different species and strains can vary dramatically based on diet [1] , host species [2] , and the identities of other co-colonizing taxa [3] . These rapid shifts in community composition suggest that individual gut microbes may be adapted to specific environmental conditions , with strong selection pressures between competing species or strains . Yet , while these ecological responses have been extensively studied , much less is known about the evolutionary forces that operate within populations of gut bacteria , both within individual hosts and across the larger host-associated population . This makes it difficult to predict how rapidly strains of gut microbes will evolve new ecological preferences when faced with environmental challenges , such as drugs or diet , and how the genetic composition of the community will change as a result . The answers to these questions depend on two different types of information . At a mechanistic level , one must understand the functional traits that are under selection in the gut and how they may be modified genetically . Recent work has started to address this question , leveraging techniques from comparative genomics [4–6] , evolution in model organisms [7–9] , and high-throughput genetic screens [10 , 11] . Yet , in addition to the targets of selection , evolution also depends on population genetic processes that describe how mutations spread through a population of gut bacteria , both within individual hosts and across the larger population . These dynamical processes can strongly influence which mutations are likely to fix within a population , and the levels of genetic diversity that such populations can maintain . Understanding these processes is the goal of our present work . Previous studies of pathogens [12] , laboratory evolution experiments [13] , and some environmental communities [14–17] have shown that microbial evolutionary dynamics are often dominated by rapid adaptation , with new variants accumulating within months or years [7 , 14 , 18–25] . However , it is not clear how this existing picture of microbial evolution extends to a more complex and established ecosystem like the healthy gut microbiome . On the one hand , hominid gut bacteria have had many generations to adapt to their host environment [26] , and they may not be subjected to the same immune pressures as pathogens . The large number of potential competitors in the gut ecosystem may also provide fewer opportunities for a strain to adapt to new conditions before an existing strain expands to fill the niche [27 , 28] or a new strain invades from outside the host . On the other hand , it is also possible that small-scale environmental fluctuations , either driven directly by the host or through interactions with other resident strains , might increase the opportunities for local adaptation [29] . If immigration is restricted , the large census population size of gut bacteria could allow residents to produce and fix adaptive variants rapidly before a new strain is able to invade . In this case , one could observe rapid adaptation on short timescales , which is eventually arrested on longer timescales as strains are exposed to the full range of host environments . Additional opportunities for adaptation can occur if the range of host environments also shifts over time ( e . g . , due to urbanization , antibiotic usage , etc . ) . Determining which of these scenarios apply to gut communities is critical for efforts to study and manipulate the microbiome . While traditional amplicon sequencing provides limited resolution to detect within-species evolution [30] , whole-genome shotgun metagenomic sequencing is starting to provide the raw polymorphism data necessary to address these questions [31] . In particular , several reference-based approaches have been developed to detect genetic variants within individual species in larger metagenomic samples [31–36] . While these approaches enable strain-level comparisons between samples , they have also documented substantial within-species variation in individual metagenomes [31 , 35 , 37] . This makes it difficult to assign an evolutionary interpretation to the genetic differences between samples , because they arise from unobserved mixtures of different bacterial lineages . Several approaches have been developed to further resolve these mixed populations into individual haplotypes or "strains . " These range from simple consensus approximations [35 , 37 , 38] , to sophisticated clustering algorithms [39 , 40] and the incorporation of physical linkage information [41] . However , while these methods are useful for tracking well-defined strains across samples , it is not known how their assumptions and failure modes might bias inferences of evolutionary dynamics , particularly among closely related strains . As a result , the evolutionary processes that operate within species of gut bacteria remain poorly characterized . In this study , we take a different approach to the strain detection problem that is specifically designed for inferring evolutionary dynamics in a large panel of metagenomes . Building on earlier work by [4 , 35] , we show that many prevalent species have a subset of hosts for which a portion of the dominant lineage is much easier to identify . By focusing only on this subset of samples , we develop methods for resolving small differences between the dominant lineages with a high degree of confidence . We use this approach to analyze a large panel of publicly available human stool samples [42–46] , which allows us to quantify evolutionary dynamics within and across hosts in approximately 40 prevalent bacterial species . We find that the long-term evolutionary dynamics across hosts are broadly consistent with models of quasi-sexual evolution and purifying selection , with relatively weak geographic structure in many prevalent species . However , our quantitative approach also reveals interesting departures from standard population genetic models of these processes , suggesting that new models are required to fully understand the evolutionary dynamics that take place across the larger population . We also use our approach to detect examples of within-host adaptation , in which nucleotide variants or gene gains or losses rapidly sweep to high frequency on 6-month timescales . We find evidence that some within-host sweeps may be seeded by recombination , in addition to de novo mutations , as might be expected for a complex ecosystem with frequent horizontal exchange . However , by analyzing differences between adult twins , we find that short-term evolution can eventually be overwhelmed by the invasion of distantly related strains on multi-decade timescales . This suggests that resident strains are rarely able to become so well adapted to a particular host that they prevent future replacements . Together , these results show that the gut microbiome is a promising system for studying the dynamics of microbial evolution in a complex community setting . The framework we introduce may also be useful for characterizing evolution of microbial communities in other environments .
To investigate evolutionary dynamics within species in the gut microbiome , we analyzed shotgun metagenomic data from a panel of stool samples from 693 healthy individuals sequenced in previous work ( S1 Table ) . This panel includes 250 North American subjects sequenced by the Human Microbiome Project ( HMP ) [42 , 44] , a subset of which were sampled at 2 or 3 time points roughly 6–12 months apart . To probe within-host dynamics on longer timescales , we also included data from a cohort of 125 pairs of adult twins from the TwinsUK registry [45] , and 4 pairs of younger twins from [46] . As we describe below , the differences between these cohorts provide a proxy for the temporal changes that accumulate in adult twins over longer timescales . Finally , to further control for geographic structure , we also included samples from 185 Chinese subjects sequenced at a single time point [43] . We used a standard reference-based approach to measure single nucleotide variant ( SNV ) frequencies and gene copy number across a panel of prevalent species for each metagenomic sample ( see S1A Text for details on the bioinformatic pipeline , including mapping parameters and other filters ) . Descriptive summaries of this genetic variation have been reported elsewhere [31 , 33–35 , 37 , 44] . Here , we revisit these patterns to investigate how they emerge from the lineage structure set by the host colonization process . Using these results , we then show how certain aspects of this lineage structure can be inferred from the statistics of within-host polymorphism , which enable measurements of evolutionary dynamics across samples . As an illustrative example , we first focus on the patterns of polymorphism in Bacteroides vulgatus , which is among the most abundant and prevalent species in the human gut . These properties ensure that the B . vulgatus genome has high coverage in many samples , which enables more precise estimates of the allele frequencies in each sample ( Fig 1A–1D ) . The overall levels of within-host diversity for this species are summarized in Fig 1E , based on the fraction of synonymous sites in core genes with intermediate allele frequencies ( white region in Fig 1A–1D ) . This measure of within-host genetic variation varies widely across the samples: some metagenomes have only a few variants along the B . vulgatus genome , while others have mutations at more than 1% of all synonymous sites ( comparable to the differences between samples , S5 Fig ) . Similar patterns are observed in many other prevalent species ( S3 Fig ) . We first asked whether these patterns are consistent with a model in which each host is colonized by a single B . vulgatus clone , so that the intermediate frequency variants represent mutations that have arisen since colonization . Using conservatively high estimates for per-site mutation rates ( μ~10−9 [47] ) , generation times ( approximately 10 per day [48] ) , and time since colonization ( <100 years ) , this model predicts that the neutral polymorphism rate at synonymous sites should be no greater than 0 . 1% ( S1B Text , part ii ) . This is at odds with the higher levels of diversity observed in many samples ( Fig 1E and S3 Fig ) . Instead , we conclude that the samples with higher synonymous diversity have been colonized by multiple divergent bacterial lineages that accumulated mutations for many generations before coming together in the same gut community . As a plausible alternative , we next asked whether the data are consistent with a large number of colonizing lineages ( nc≫1 ) drawn at random from the broader population . However , this process is expected to produce fairly consistent polymorphism rates and allele frequency distributions in different samples , which is at odds with the variability we observe even among the high-diversity samples ( e . g . , Fig 1A , 1B , S1 Fig and S2 Fig ) . Instead , we hypothesize that many of the high-diversity hosts have been colonized by just a few diverged lineages [i . e . , ( nc−1 ) ∼O ( 1 ) ] . Consistent with this hypothesis , the distribution of allele frequencies in each host is often strongly peaked around a few characteristic frequencies , suggesting a mixture of several distinct lineages ( Fig 1A–1C , S1 Fig and S2 Fig ) . Similar findings have recently been reported in a number of other host-associated microbes , including several species of gut bacteria [4 , 35 , 49 , 50] . Fig 1A–1C shows that hosts can vary both in the apparent number of colonizing lineages and the frequencies at which they are mixed together . As a result , we cannot exclude the possibility that even the low-diversity samples ( e . g . , Fig 1D ) are colonized by multiple lineages that happen to fall below the detection threshold set by the depth of sequencing . Compared with the extreme cases of single-colonization ( nc = 1 ) or colonization by many strains ( nc≫1 ) , it is more difficult to identify evolutionary changes between lineages when there are only few strains at intermediate frequency . In this scenario , within-host populations are not clonal , but the corresponding allele frequencies derive from idiosyncratic colonization processes rather than a large random sample from the population ( as , e . g . , in [16] ) . To disentangle genetic changes between lineages from these host-specific factors , we must estimate phased haplotypes ( or "strains" ) from the distribution of allele frequencies within individual hosts . This is a complicated inverse problem , and we will not attempt to solve the general case here . Instead , we adopt an approach similar to [35] and others , and leverage the fact that the lineage structure in certain hosts is sufficiently simple that we can assign alleles to the dominant lineage with a high degree of confidence . Our approach is based on the simple observation that two high-frequency variants must co-occur in an appreciable fraction of cells ( S1C Text , part i ) . This "pigeonhole principle" suggests that we can estimate the genotype of one of the lineages in a mixed sample by taking the major alleles present above some threshold frequency , f*≫50% , and treating the remaining sites as missing data . Although the potential errors increase with the length of the inferred haplotype , we will not actually require genome-length haplotypes for our analysis here . Instead , we leverage the fact that significant evolutionary information is already encoded in the marginal distributions of one- and two-site haplotypes , so that these "quasi-phased" lineages will be sufficient for our present purposes . The major challenge with this approach is that we do not observe the true allele frequency directly but must instead estimate it from a noisy sample of sequencing reads . This can lead to phasing errors when the true major allele is sampled at low frequency by chance and is assigned to the opposite lineage ( S4 Fig ) . We will refer to these as "polarization errors , " because they stem from an incorrect inference of the major allele . The probability of a polarization error will vary dramatically depending on the sequencing coverage and the true frequency of the major allele ( S1C Text , part ii ) . Previous approaches based on consensus alleles [35 , 37] can therefore induce an unknown number of errors that make it difficult to confidently detect a small number of evolutionary changes between samples . In S1C Text , we show that by explicitly modeling the sampling error process , the expected probability of a polarization error in our cohort can be bounded to be sufficiently low if we take f* = 80% , and if we restrict our attention to samples with sufficiently high coverage and sufficiently low rates of intermediate-frequency polymorphism . We will refer to these as quasi-phaseable ( QP ) samples . In the B . vulgatus example above , Fig 1C and 1D are classified as QP , while Fig 1A and 1B are not . Note that quasi-phaseability is separately defined for each species in a metagenomic sample , rather than for the sample as a whole . For simplicity , we will still refer to these species-sample combinations as QP samples , with the implicit understanding that they refer to a particular focal species . In Fig 1F , we plot the distribution of QP samples across the most prevalent gut bacterial species in our panel . The fraction of QP samples varies between species , ranging from about 50% in the case of Prevotella copri to nearly 100% for B . fragilis [4] , and it accounts for much of the variation in the average polymorphism rate between species ( S6 Fig ) . Most individuals carry a mixture of QP and non-QP species ( S7 Fig ) , suggesting that quasi-phaseability arises independently for each species in a sample , rather than for the sample as a whole . Thus , although many species-sample combinations are not QP , our approximately 500-sample cohort still contains tens to hundreds of QP samples in many prevalent species , yielding about 3 , 000 quasi-phased haplotypes in total . Consistent with previous studies of the stability of personal microbiomes [31 , 35 , 51] , a majority of the longitudinally sampled species maintain their QP classification at both time points , although this pattern is not universal ( S8 Fig ) . We will revisit the peculiar properties of this within-host lineage distribution in “Discussion . ” For the remainder of the analysis , we will take the distribution in Fig 1F as given and focus on leveraging the QP samples to quantify the evolutionary changes that accumulate between lineages in different samples . We investigate two types of evolutionary changes between lineages in different QP samples . The first class consists of single nucleotide differences , which are defined as SNVs that segregate at frequencies ≤1−f* in one sample and ≥f* in another , with f*≈80% as above ( S4 Fig ) . These thresholds are chosen to ensure a low genome-wide false positive rate given the typical coverage and allele frequency distributions among the QP samples in our panel ( S1C Text , part iv ) . The second class consists of differences in gene presence or absence , in which the relative copy number of a gene , c , is below the threshold of detection ( c<0 . 05 ) in one sample and is consistent with a single-copy gene ( 0 . 6<c<1 . 2 , see S9 Fig ) in the other sample . These thresholds are chosen to ensure a low genome-wide false positive rate across the QP samples , given the typical variation in sequencing coverage along the genome ( S1C Text , part v ) , and to minimize mapping artifacts ( S1A Text , part ii ) . Note that these SNV and gene changes represent only a subset of the potential differences between lineages . We neglect other evolutionary changes ( e . g . , indels , genome rearrangements , or changes in high copy number genes ) that are more difficult to quantify in a metagenomic sample , as well as more subtle changes in allele frequency and gene copy number that do not reach our stringent detection thresholds . We will revisit these and other limitations in more detail in “Discussion” .
By focusing on the QP samples for each species , we can measure genetic differences between lineages in different hosts , as well as within hosts over short time periods . Descriptive summaries of this variation have been reported elsewhere [31 , 33–35 , 37 , 44] . Here , we aim to leverage these patterns ( and the increased resolution of the QP samples ) to quantify the evolutionary dynamics that operate within species of gut bacteria , both within and across hosts . To interpret within-host changes in an evolutionary context , it will be useful to first understand the structure of genetic variation between lineages in different hosts . This variation reflects the long-term population genetic forces that operate within each species , presumably integrating over many rounds of colonization , growth , and dispersal . To investigate these forces , we first analyzed the average nucleotide divergence between strains of a given species in different pairs of QP hosts ( Fig 2A ) . In the case of twins , we included only a single host from each pair , to better approximate a random sample from the population . Fig 2B shows the distribution of pairwise divergence , averaged across the core genome , for about 40 of the most prevalent bacterial species in our cohort . In a panmictic , neutrally evolving population , we would expect these distances to be clustered around their average value , d≈2μTc , where Tc is the coalescent timescale for the across-host population [52] . By contrast , Fig 2 shows striking differences in the degree of relatedness for strains in different hosts . Even at this coarse , core-genome-wide level , the genetic distances vary over several orders of magnitude . Some species show multiple peaks of divergence for high values of d , consistent with the presence of subspecies [36] , ecotypes [53 , 54] , or other strong forms of population structure . These coarse groupings have been observed previously and are not our primary focus here . Rather , we seek to understand the population genetic forces that operate at finer levels of taxonomic resolution . From this perspective , the more surprising parts of Fig 2 are the thousands of pairs of lineages with extremely low between-host divergence ( e . g . , d≲0 . 01% ) , more than an order of magnitude below the median values in most species . Similar observations have recently been reported by [35] and are often interpreted as strain sharing across hosts . However , the evolutionary interpretation of these closely related strains remains unclear . The simplest explanation for a long tail of closely related strains is cryptic relatedness [55] , arising from a breakdown of random sampling . For microbes , this can occur when two cells are sampled from the same clonal expansion , e . g . , when strains are transferred between mothers and infants [33 , 56] , between cohabitating individuals [46] , or within a hospital outbreak [57] . While these transmission events have been observed in other studies , they are unlikely to account for the patterns here . All of the lineages in Fig 2 are sampled from individuals in different households , and more than a third of the closely related pairs derive from individuals on different continents ( Fig 2B ) . Of course , there could still be some other geographic variable , beyond household or continent of origin , that could explain an elevated probability of transmission between two individuals . Fortunately , our metagenomic approach allows us to rule out these additional sources of cryptic host relatedness by leveraging multiple species comparisons for the same pair of hosts . If there were a hidden geographic variable , then we would expect that individuals with closely related strains in one species would be much more likely to share closely related strains in other species as well . However , we observe only a small fraction of hosts that share multiple closely related strains ( Fig 2C ) , consistent with a null model in which these strains are randomly and independently distributed across hosts . This suggests that host-wide sampling biases are not the primary driver of the closely related strains in Fig 2 . Although the rates of nucleotide divergence are low , the vast majority of these strains are still genetically distinguishable from each other . The absolute number of SNV differences typically exceeds our estimated false positive rate ( S10A Fig , S1C Text , part iv ) , and these SNV differences are often accompanied by ≳10 differences in gene content ( S10B Fig ) . Furthermore , we found that closely related strains frequently differed in their collections of private marker SNVs ( S11 Fig ) , which are often used to track strain transmission events [33 , 46] . Together , these lines of evidence suggest that closely related strains are often genetically distinct and do not arise from a simple clonal expansion . Instead , the data suggest that there are additional population genetic timescales beyond Tc that are relevant for microbial evolution . This hypothesis is bolstered by the large number of species , particularly in the Bacteroides genus , with anomalously low divergence rates between some pairs of hosts . However , we note that this pattern is not universal: some genera , like Alistipes or Eubacterium , show more uniform rates of divergence between hosts . Apart from these phylogenetic correlations , we cannot yet explain why some species have low-divergence host pairs and others do not . Natural candidates such as sample size , abundance , vertical transmissibility [33] , or sporulation score [58] struggle to explain the differences between Bacteroides and Alistipes species . We next examined how natural selection influences the genetic diversity observed between hosts . Previous work has suggested that genetic diversity in many species of gut bacteria is strongly constrained by purifying selection , which purges deleterious mutations that accumulate between hosts [31] . However , the temporal dynamics of this process remain poorly understood . We do not know whether purifying selection acts quickly enough to prevent deleterious mutations from spreading to other hosts , or if deleterious mutations typically spread across multiple hosts before they are purged . In addition , it is plausible that the dominant mode of natural selection could be different for the closely related strains above ( e . g . , if they reflect recent ecological diversification [15] ) . To address these questions , we analyzed the relative contribution of synonymous and nonsynonymous mutations that comprise the overall divergence rates in Fig 2A . We focused on the ratio between the per-site divergence at nonsynonymous sites ( dN ) and the corresponding value at synonymous sites ( dS ) . Under the assumption that synonymous mutations are effectively neutral , the ratio dN/dS measures the average action of natural selection on mutations at nonsynonymous sites . In Fig 3 , we plot these dN/dS estimates across every pair of QP hosts for each of the prevalent species in Fig 2A . The values of dN/dS are plotted as a function of dS , which serves as a proxy for the average divergence time across the genome . We observe a consistent negative relationship between these two quantities across the prevalent species in Fig 2 . For large divergence times ( dS~1% ) , we observe only a small fraction of nonsynonymous mutations ( dN/dS~0 . 1 ) , indicating widespread purifying selection on amino acid replacements [31] . Yet , among more closely related strains , we observe a much higher fraction of nonsynonymous changes , with dN/dS approaching unity when dS~0 . 01% ( we observe a similar trend if we restrict our attention to singleton SNVs , S12 Fig ) . Moreover , this negative relationship between dN/dS and dS is much more pronounced than the between-species variation in the typical values of dN/dS ( black crosses in Fig 3 ) . While between-species variation may be driven by mutational biases , the strong within-species signal indicates that there are consistent differences in the action of natural selection as a function of time . In principle , the dN/dS increases in the recent past could be driven by interesting biological processes , such as enhanced adaptation or ecological diversification on short timescales , or a recent global shift in selection pressures caused by host-specific factors ( e . g . , the introduction of agriculture ) . However , the data in Fig 3 appear to be well explained by an even simpler null model of purifying selection , in which deleterious mutations are purged over a timescale inversely proportional to their cost ( S1D Text ) . This dynamical model can explain the varying signatures of natural selection without requiring that the selective pressures themselves vary over time . We find reasonable quantitative agreement for a simple distribution of fitness effects , in which 10% of nonsynonymous sites are neutral and the remaining 90% have fitness costs on the order of s/μ~105 . Although the true model is likely more complicated , we argue that this simple null model should be excluded before more elaborate explanations are considered . For example , unambiguous proof of recent adaptation could be observed if dN/dS consistently exceeded 1 among the most closely related strains ( because this can only occur by chance under purifying selection ) . While a few of the individual comparisons in Fig 3A have dN/dS>1 , the cumulative version in Fig 3B shows that dN/dS does not significantly exceed 1 , even for the lowest values of dS . This suggests that , if positive selection is present , it is not sufficiently widespread to overpower the signal of purifying selection in these global dN/dS measurements . However , there is also substantial variation around the average trend in Fig 3 , which could hide important biological variation among species ( or among different genomic regions in the same species ) . Resolving the signatures of natural selection at these finer scales remains an important avenue for future work . So far , we have focused on evolutionary changes that accumulate over many host colonization cycles . In principle , evolutionary changes can also accumulate within hosts over time . Longitudinal studies have shown that strains and metagenomes sampled from the same host are more similar to each other on average than to samples from different hosts [31 , 33 , 35 , 44 , 64 , 65] . This suggests that resident populations of bacteria persist within hosts for at least a year ( approximately 300 to 3 , 000 generations ) , which is potentially enough time for evolutionary adaptation to occur [7] . However , the limited resolution of previous polymorphism- [31] or consensus-based comparisons [35 , 44] has made it difficult to quantify the individual changes that accumulate within hosts and to interpret these changes in an evolutionary context .
Evolutionary processes can play an important role in many microbial communities . Yet , despite increasing amounts of sequence data , our understanding of these processes is often limited by our ability to resolve evolutionary changes in populations from complex communities . In this work , we quantify the evolutionary forces that operate within bacteria in the human gut microbiome by characterizing in detail the lineage structure of approximately 40 species in metagenomic samples from individual hosts . Building on previous work [35] and others , we found that many resident populations from a variety of prevalent species are best described by an "oligo-colonization" model , in which a few distinct strains from the larger population are present at intermediate frequencies , with the identities and frequencies of these strains varying from person to person ( Fig 1 ) . The distribution of strain frequencies in this oligo-colonization model is itself quite interesting: in the absence of fine tuning , it is not clear what mechanisms would allow for a second or third strain to reach intermediate frequency , while preventing a large number of other lineages from entering and growing to detectable levels at the same time . A better understanding of the colonization process and how it might vary among the species in Fig 1F is an important avenue for future work . Given the wide variation among species and hosts , we chose to focus on a subset of samples with particularly simple strain mixtures for a given species , in which we can resolve evolutionary changes in the dominant lineage with a high degree of confidence . Our quasi-phasing approach can be viewed as a refinement of the consensus approximation employed in earlier studies [4 , 35 , 37 , 38] but with more quantitative estimates of the errors associated with detecting genetic differences between lineages in different samples . By analyzing genetic differences between lineages in separate hosts , we found that long-term evolutionary dynamics in many gut bacteria are consistent with quasi-sexual evolution and purifying selection , with relatively weak geographic structure . Earlier work had documented extensive horizontal transfer between distantly related species in the gut [71 , 72] , but our ability to estimate rates of recombination within species was previously limited by the small number of sequenced isolates for many species of gut bacteria [73] . The high rates of homologous recombination we observed with our quasi-phasing approach are qualitatively consistent with previous observations in other bacterial species [16 , 73–77]; although the rates of recombination are high relative to the typical divergence time , we note that they may still allow for genome-wide sweeps or divergence between nascent ecotypes given sufficiently strong selection pressures . Beyond the overall rates , our quantitative characterization of LD also revealed interesting departures from the standard neutral prediction that cannot be captured by any choice of recombination rate . Understanding the origin of this discrepancy is an interesting topic for future work . It is also interesting to ask how these long-term rates of recombination could emerge from the oligo-colonization model above , because it would seem to limit opportunities for genetic exchange among strains of the same species . In a complex community like the gut , a key advantage of our metagenomic approach is that it can jointly measure genetic differences in multiple species for the same pair of hosts . By leveraging this feature , we found that previous observations of highly similar strains in different hosts [35 , 44] are not driven by cryptic host relatedness . Instead , the presence of these closely related strains and the genetic differences that accumulate between them may be driven by more general population genetic processes in bacteria that operate on timescales much shorter than the typical coalescent time across hosts . It is difficult to produce such closely related strains in traditional population genetic models of loosely linked loci [78] ( or "bags of genes" [79] ) , although recent hybrid models of vertical and horizontal inheritance [77 , 80] or fine-scale ecotype structure [62] could potentially provide an explanation for this effect . Further characterization of these short-term evolutionary processes will be vital for current efforts to quantify strain sharing across hosts [33 , 46 , 56] , which often require implicit assumptions about how genetic changes accumulate on short timescales . Our results suggest that these short-term dynamics of across-host evolution may not be easily extrapolated by comparing average pairs of strains . The other main advantage of our quasi-phasing approach is its ability to resolve a small number of evolutionary changes that could accumulate within hosts over short timescales . Previous work has shown that on average , longitudinally sampled metagenomes from the same host are more similar to each other than metagenomes from different hosts [31 , 33 , 64 , 65] , and that some within-host changes can be ascribed to replacement by distantly related strains [35 , 44] . However , the limited resolution of previous polymorphism- [31] or consensus-based comparisons [35 , 44] had made it difficult to determine whether resident strains also evolve over time . Our quasi-phasing approach overcomes this limitation , enabling finely resolved estimates of temporal change within individual species in individual hosts . This increased resolution revealed an additional category of within-host variation , which we have termed modification , in which resident strains acquire modest numbers of SNV and gene changes over time . This broad range of outcomes shows why it is essential to understand the distribution of temporal variation across hosts: even though modification events were about 3 times more common than replacements in our cohort , their contributions to the total genetic differences are quickly diluted as soon as a single replacement is included ( S19 Fig ) . As a result , we expect that previous metagenome-wide [31] or species-averaged [44] estimates of longitudinal variation largely reflect the rates and genetic differences associated with replacement events , rather than evolutionary changes . Although we have interpreted modifications as evolutionary events ( i . e . , mutations to an existing genome ) , it is possible that they could also reflect replacement by extremely closely related strains , as in Fig 2 . The present data seem to argue against this scenario: modifications are not only associated with different patterns of SNV sharing ( S11 Fig ) , but we also observe significant asymmetries in the prevalence distributions in Fig 5C and 5D that depend on the temporal ordering of the 2 samples ( see Fig 5 ) . This temporal directionality arises naturally in certain evolutionary models ( e . g . , the de novo mutation model in Fig 5C ) , but it is less likely to emerge from steady-state competition between a fixed set of strains . Unambiguous proof of evolution could also be observed in a longer time course , because subsequent evolutionary changes should eventually accumulate in the background of earlier substitutions . Further investigation of these nested substitutions remains an interesting topic for future work . The signatures of the sweeping SNVs , along with the presence of gene gain events , suggest that some of the within-host sweeps we observed were seeded by recombination , rather than de novo mutation . In particular , many of the alleles that swept within hosts were also present in many other hosts , yet their dN/dS values indicated strong purifying selection , consistent with an ancient polymorphism ( Fig 3 ) . Sweeps of private SNVs , by contrast , were associated with a much higher fraction of nonsynonymous mutations , consistent with adaptive de novo evolution . Interestingly , we also observe a slight excess of private nonsynonymous mutations between closely related strains in different hosts ( S12 Fig ) . This suggests that some of the differences observed between hosts may reflect a record of recent within-host adaptation . Recombination-seeded sweeps would stand in contrast to the de novo mutations observed in microbial evolution experiments [13] and some within-host pathogens [21 , 22] . Yet in hindsight , it is easy to see why recombination could be a more efficient route to adaptation in a complex ecosystem like the gut microbiome , given the large strain diversity [42] , the high rates of DNA exchange [71 , 72] , and the potentially larger selective advantage of importing an existing functional unit that has already been optimized by natural selection [11] . Consistent with this hypothesis , adaptive introgression events have also been observed on slightly longer timescales in bacterial biofilms from an acid mine drainage system [14] , and they are an important force in the evolution of virulence and antibiotic resistance in clinical settings [81] . While the data suggest that some within-host changes may be seeded by a recombination event , it is less clear whether ongoing recombination is relevant during the sweep itself . Given the short timescales involved , we would expect many of the observed sweeps to proceed in an essentially clonal fashion , because recombination would have little time to break up a megabase-sized genome given the typical rates inferred in S17 Fig . If this were the case , it would provide many opportunities for substantially deleterious mutations ( with fitness costs of order Sd~1% per day ) to hitchhike to high frequencies within hosts [70] , thereby limiting the ability of bacteria to optimize to their local environment . The typical fitness costs inferred from Fig 2D lie far below this threshold and would therefore be difficult to purge within individual hosts . In this scenario , the low values of dN/dS observed between hosts ( as well as the putative introgression events ) would crucially rely on the competition process across hosts [82] . Although the baseline recombination rates suggest clonal sweeps , there are also other vectors of exchange ( e . g . , transposons , prophage , etc . ) with much higher rates of recombination . Such mechanisms could allow within-host sweeps to behave in a quasi-sexual fashion , preserving genetic diversity elsewhere in the genome . These sweeps of local genomic regions are predicted in certain theoretical models [83 , 84] and have been observed in a few other bacterial systems [15 , 17 , 85] . If sweeps of local genomic regions were also a common mode of adaptation in the gut microbiome , they would allow bacteria to purge deleterious mutations more efficiently than in the clonal scenario above . Although evolution was more common than replacement on 6-month timescales , our analysis of adult twins suggests that the rare replacement events eventually dominate on multi-decade timescales . This suggests that resident strains are limited in their ability to evolve to become hyper-adapted to their host , because most strains were eventually susceptible to replacement . Such behavior would be consistent with theoretical models in which strains of the same species only partially overlap in their ecological niches [27 , 54] . Although our results indicate that the long-term probability of replacement is largely uniform across hosts , it remains an open question whether these events occur more or less uniformly in time or whether they occur in punctuated bursts during major ecosystem perturbations ( e . g . , antibiotic treatment ) . This would be an interesting question to address with denser and longer time series data . Finally , while we have identified many interesting signatures of within-host adaptation , there are several important limitations to our analysis . One class concerns the events that we cannot observe with our approach ( i . e . , false negatives ) . These are particularly relevant here , because we have discarded substantial amounts of data in an attempt to overcome the traditional problems of metagenomic inference ( S24 Fig ) . For example , our reference-based method only tracks SNVs and gene copy numbers in the genomes of previously sequenced isolates of a given species . Within this subset , we have also imposed a number of stringent bioinformatic filters , further limiting the sequence space that we consider . Thus , it is likely that we are missing many of the true targets of selection , which might be expected to be concentrated in the host-specific portion of the microbiome , multi-copy gene families , or in genes that are shared across multiple prevalent species . A further limitation is that we can only analyze the evolutionary dynamics of QP samples ( although the consistency of our results for species with different QP fractions suggests that this might not be a major issue ) . Finally , a potentially more important false negative is that our current method can only identify complete or nearly complete sweeps within individual hosts . While we observed many within-host changes that matched this criterion , we may be missing many other examples of within-host adaptation in which variants do not completely fix . Given the large population sizes involved , such sweeps can naturally arise from phenotypically identical mutations at multiple genetic loci [69 , 86] , or through additional ecotype partitioning between the lineages of a given species [23 , 25] . Both mechanisms have been observed in experimental populations of Escherichia coli adapting to a model mouse microbiome [7] . In addition to these false negatives , the other limitation of our approach concerns potential false positives inherent in any metagenomic analysis . With short-read data , it is difficult to truly know whether a paticular DNA fragment is linked to a particular species or whether it resides in the genome of another species ( perhaps an uncultured one ) that is fluctuating in abundance . False SNV and gene changes can therefore occur because of these read donating effects . The temporally asymmetric prevalence distributions in Fig 5C and 5D suggest that our filters were successful in eliminating many of these events ( S1H Text , part iii ) . However , isolate or long-read sequences are required to unambiguously prove that these variants are linked to the population of interest . Fortunately , two concurrent studies have also documented short-term evolution of gut bacteria within healthy human hosts using an isolate-based approach [87 , 88] . Each study focused on a single bacterial species , E . coli in [87] and B . fragilis in [88] . Although E . coli was not sufficiently abundant in our cohort to be included in our within-host analysis , the observations in B . fragilis are largely consistent with our findings that within-host evolution can be rapid and that it can be mediated by recombination in addition to new mutations . Crucially , because these observations were obtained using an isolate-based approach , they are not subject to the same methodological limitations described above , and they therefore serve as an independent verification of our results . However , because our statistical approach provides simultaneous observations across more than 40 prevalent species , our results show that these general patterns of within-host evolution are shared across many species of gut bacteria , and they demonstrate a general approach for investigating these forces in widely available metagenomic data . Future efforts to combine metagenomic- and isolate-based approaches , e . g . , by incorporating long-range linkage information [41 , 89 , 90] , will be crucial for building a more detailed understanding of these evolutionary processes .
|
The human gut harbors a diverse microbial community whose composition is shaped by a variety of ecological forces . Given the high rates of turnover , the residents of this community might also have the opportunity to evolve over time by acquiring heritable changes to their genomes . Yet , despite the potential importance of these effects , we currently know very little about the evolutionary dynamics that occur within species in this complex community . Here , we introduce a new approach for extracting evolutionary signals from a large panel of human gut metagenomes and for interpreting these signals using simple null models from population genetics . We use this approach to quantify the evolutionary dynamics of approximately 40 prevalent species of gut bacteria , both within individual hosts and across the larger population . We find that resident populations of gut bacteria can evolve within their hosts on short timescales , but after many years , the resident populations are typically replaced by distantly related strains . The patterns of variation across hosts indicate widespread recombination within species , but the quantitative signals suggest interesting departures from traditional population genetic models . Together , these results show that short-term evolution in the gut microbiome may be more complex and widespread than is often assumed .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"bacteriology",
"organismal",
"evolution",
"gut",
"bacteria",
"population",
"genetics",
"microbiology",
"twins",
"developmental",
"biology",
"metagenomics",
"microbial",
"evolution",
"evolutionary",
"adaptation",
"population",
"biology",
"bacteria",
"evolutionary",
"genetics",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"evolutionary",
"biology",
"bacterial",
"evolution",
"evolutionary",
"processes",
"organisms"
] |
2019
|
Evolutionary dynamics of bacteria in the gut microbiome within and across hosts
|
Studies of nervous system connectivity , in a wide variety of species and at different scales of resolution , have identified several highly conserved motifs of network organization . One such motif is a heterogeneous distribution of connectivity across neural elements , such that some elements act as highly connected and functionally important network hubs . These brain network hubs are also densely interconnected , forming a so-called rich club . Recent work in mouse has identified a distinctive transcriptional signature of neural hubs , characterized by tightly coupled expression of oxidative metabolism genes , with similar genes characterizing macroscale inter-modular hub regions of the human cortex . Here , we sought to determine whether hubs of the neuronal C . elegans connectome also show tightly coupled gene expression . Using open data on the chemical and electrical connectivity of 279 C . elegans neurons , and binary gene expression data for each neuron across 948 genes , we computed a correlated gene expression score for each pair of neurons , providing a measure of their gene expression similarity . We demonstrate that connections between hub neurons are the most similar in their gene expression while connections between nonhubs are the least similar . Genes with the greatest contribution to this effect are involved in glutamatergic and cholinergic signaling , and other communication processes . We further show that coupled expression between hub neurons cannot be explained by their neuronal subtype ( i . e . , sensory , motor , or interneuron ) , separation distance , chemically secreted neurotransmitter , birth time , pairwise lineage distance , or their topological module affiliation . Instead , this coupling is intrinsically linked to the identity of most hubs as command interneurons , a specific class of interneurons that regulates locomotion . Our results suggest that neural hubs may possess a distinctive transcriptional signature , preserved across scales and species , that is related to the involvement of hubs in regulating the higher-order behaviors of a given organism .
Neuronal connectivity provides the substrate for integrated brain function . Recent years have seen several systematic and large-scale attempts to generate comprehensive wiring diagrams , or connectomes , of nervous systems [1] in species as diverse as the Caenorhabditis elegans [2 , 3] , Drosophila melanogaster [4 , 5] , zebrafish [6 , 7] , mouse [8 , 9] , rat [10] , cat [11] , macaque [12 , 13] , and human [14 , 15] . One of the most striking findings to emerge from analyses of these diverse data , acquired using different measurement techniques and at resolution scales ranging from nm to mm , is of a strong conservation of certain topological properties of network organization ( reviewed in [1 , 16–19] , see also [20] ) . These properties include a modular organization , such that subsets of functionally related ( and usually spatially adjacent ) elements are densely interconnected with each other; a hierarchy of modules , such that modules contain nested sub-modules and so on over multiple scales [21 , 22]; economical connectivity , such that wiring costs ( typically measured in terms of wiring length ) are near-minimal given the topological complexity of the system [22 , 23]; a heterogeneous distribution of connections across network nodes , such that most nodes possess relatively few connections and a small proportion of nodes have a very high degree of connectivity [3 , 24]; and stronger interconnectivity between hub nodes than expected by chance , leading to a so-called rich-club organization ( i . e . , the hubs are rich because they are highly connected and form a club because they are densely interconnected ) [5 , 25–28] . The rich-club organization of hub connectivity is thought to play a central role in integrating functionally diverse and anatomically disparate neuronal systems [26 , 29–32] . Consistent with this view , experimental data and computational modeling indicates that hub nodes , and the connections between them , are topologically positioned to mediate a high volume of signal traffic [33–36] . This integrative role comes at a cost , with connections between hubs extending over longer distances , on average , than other types of connections , a finding reported in the human [33] , macaque [34] , rat [37] , mouse [38] , and nematode [28] . Human positron emission tomography also suggests that hub nodes consume greater metabolic resources and have higher levels of blood flow than other areas [39–41] . This high metabolic cost may underlie the involvement of hub regions in a broad array of human diseases [17 , 29 , 31] . Recent studies in mice and humans suggest that the high cost of hub connectivity is associated with a distinct molecular signature , as inferred from brain-wide transcriptomic data . This work has focused on how patterns of expression across many thousands of genes covary between pairs of brain regions . Such patterns of covariation are variously referred to as transcriptional coupling [38] , gene coexpression [42] or correlated gene expression [43–45] . The goal of this work is to understand how pair-wise coupling of gene expression corresponds to the pairwise connection topology of the brain , thus drawing a link between molecular function and large-scale network organization . For example , Fulcher and Fornito [38] combined viral tract tracing data on connectivity between 213 regions of the right hemisphere of the mouse brain [8] with in situ hybridization measures of the expression of 17 642 genes in each of those regions . They found that transcriptional coupling , across all genes , is stronger for connected compared to unconnected brain regions and that , in general , this coupling decays with increasing separation distance between brain regions . Countering this general trend , however , connected pairs of hubs ( i . e . , the “rich club” of the brain ) show the highest levels of transcriptional coupling ( compared to hub-nonhub and nonhub-nonhub pairs ) , despite being separated by larger distances , on average , and being distributed across diverse anatomical brain divisions . Moreover , this coupling is driven predominantly by genes regulating the oxidative synthesis and metabolism of adenosine triphosphate ( ATP ) , the primary energetic currency of neuronal signaling [46 , 47] . Vértes et al . [48] later combined gene expression data of 20 737 genes through 285 cortical areas of the human brain and found evidence that inter-modular hubs in resting state fMRI connectivity networks also have local transcriptional profiles enriched in oxidative metabolism and mitochondria . Together , the analyses of Fulcher and Fornito [38] and Vértes et al . [48] , which were performed using measures of mesoscale ( μm to mm ) and macroscale ( mm to cm ) connectivity , respectively , suggest that the molecular signature of hub connectivity , characterized by elevated coupling of genes regulating oxidative metabolism , may be conserved across species and resolution scales . Here , we sought to test this hypothesis by examining the link between gene expression and microscale connectivity in the nematode C . elegans . C . elegans is the only species to have its connectome mapped almost completely at the level of individual neurons and synapses using electron microscopy [2 , 3] . It comprises 302 neurons and around 5600 chemical and electrical synapses [2] . We combined these data on neuronal connectivity with curated information on the binary expression patterns of 948 genes across neurons to examine the relationship between gene expression and the large-scale topological organization of the nematode nervous system . We also used detailed information on neuron spatial positions , birth times , neuronal lineage as well as the functional and chemical composition of each neuron to understand the mechanisms through which gene expression might influence network topology . Paralleling findings in humans and mouse , we find that hub neurons of C . elegans are genomically distinct , with connected hub neurons showing the most similar patterns of gene expression . Genes that contribute most to this effect are involved in regulating glutamate and acetylcholine function , and neuronal communication . We demonstrate that this effect cannot be explained by factors such as neuronal birth time , lineage , neurotransmitter system or spatial position , but may instead be related to the functional specialization of hub neurons in mediating higher order behaviours of the organism .
The nervous system of C . elegans comprises 302 neurons , divided into the pharyngeal nervous system ( 20 neurons ) and the somatic nervous system ( 282 neurons ) . While research detailing the genetic underpinnings that guide the formation of C . elegans nervous system is still ongoing , the spatial positions of neurons , and their interconnections , are known to be stereotypical across organisms [49] . Neuronal connectivity data for the adult hermaphrodite C . elegans was first mapped by White et al . [2] through detailed electron microscopy , and then revised by Chen et al . [50] and Varshney et al . [3] . Here we analyze the larger somatic nervous system using data from Varshney et al . [3] , who mapped connectivity between 279 neurons ( 282 somatic neurons , i . e . , excluding CANL/R , and VC6 , which do not form synapses with other neurons ) , which we obtained from WormAtlas [51] ( www . wormatlas . org/neuronalwiring . html#NeuronalconnectivityII ) . Connectivity data are available for both electrical gap junctions and chemical synapses . The chemical synapse network is both directed ( i . e . , the pre-synaptic and post-synaptic neurons are identified ) and weighted ( as the number of synapses from one neuron to another ) , while gap junctions are conventionally represented as weighted ( as the number of electrical synapses connecting two neurons ) , undirected connections . Previous investigations of C . elegans neuronal connectivity have used differently processed versions of these data , including: ( i ) only chemical synapses [52]; ( ii ) a combination of chemical and electrical synapses as a directed network ( electrical synapses represented as reciprocal connections ) [53 , 54]; ( iii ) a combination of chemical and electrical synapses as an undirected network ( representing unidirectional and reciprocal chemical connections equivalently ) [28 , 55–57]; or ( iv ) comparing multiple connectome representations [58] . Our analysis here focuses on the combined directed , binary network , treating gap junctions as bidirectional connections . We chose to focus on a binarized network to follow previous studies on C . elegans connectome data [28 , 56 , 59–61] and to enable a more direct comparison to our previous analysis of the relationship between ( binary ) connectivity and gene expression in mouse [38] . The resulting connectome contains 279 neurons , with 2 990 unique connections linking 2 287 pairs of neurons . Note that the qualitative results presented here are not highly sensitive to the types of connections included in the connectome . For example , neuron degree is highly correlated between networks generated using: ( i ) combined chemical and electrical synapses and ( ii ) chemical synapses only ( Spearman’s ρ = 0 . 9 , p = 3 × 10−107 ) . We also obtained qualitatively similar results for rich-club organization and trends in CGE when excluding gap junctions from our analysis ( see S1 Fig ) . In addition , we assembled a range of data characterizing other properties of C . elegans neurons . To examine the effect of physical distance between pairs of neurons , we obtained two dimensional spatial coordinates for each neuron as celegans277 . mat from www . biological-networks . org/ ? page_id=25 [62] . Coordinates for three neurons ( AIBL , AIYL , SMDVL ) were missing in this dataset , and were reconstructed by assigning identical coordinates to the corresponding contralateral neurons ( AIBR , AIYR , SMDVR ) [59] . To examine the influence of anatomical location , each neuron was labeled by its anatomical location , as: ( i ) ‘head’ , ( ii ) ‘tail’ , or ( iii ) ‘body’ , using data from release WS256 of WormBase [63] , ( ftp://ftp . wormbase . org/pub/wormbase/releases/WS256/ONTOLOGY/anatomy_association . WS256 . wb ) . These annotations were assigned to individual neurons using the anatomical hierarchy defined in WormBase , which we retrieved using the WormBase API ( WormMine: intermine . wormbase . org ) [63] , propagating each term down the hierarchy to individual neurons . We manually corrected the assignment of twelve head neurons ( ALA , AVFL , AVFR , AVG , RIFL , RIFR , RIGL , RIGR , SABD , SABVL , SABVR , SMDVL ) , which were assigned as ‘head’ in WormAtlas [51] but not on WormBase . To examine the influence of neuronal subtype , all neurons were labeled to one ( or multiple ) of the following three categories: ( i ) ‘sensory’ ( have clear sensory specializations ) , ( ii ) ‘motor’ ( make synaptic contacts onto muscle cells ) , or ( iii ) ‘interneuron’ ( receive synapses from and send synapses onto other neurons ) [2] . A total of 79 neurons are annotated as sensory , 121 annotated as motor , and 97 annotated as interneurons ( including eighteen neurons assigned to two categories: five are both ‘interneuron’ and ‘sensory’ , seven are ‘interneuron’ and ‘motor’ , and six are both ‘sensory’ and ‘motor’ neurons ) . To examine the neurotransmitter systems used by each neuron , neurons were labeled by matching to Table 2 of Pereira et al . [64] . In order to determine the influence of neuron birth time , we obtained neuronal birth time information in minutes from the Dynamic Connectome Lab website ( https://www . dynamic-connectome . org/ ? page_id=25 ) , [59] . To assess the influence of lineage similarity , we obtained a measure of lineage distance for all pairs of neurons from previously published embryonic and post-embryonic lineage trees [65 , 66] , using data downloaded from WormAtlas ( http://www . wormatlas . org/neuronalwiring . html#Lineageanalysis ) [51] . In this dataset , the closest common ancestor neuron was identified for each pair of neurons , and then the lineage distance was calculated as the number of cell divisions from the closest common progenitor neuron . In this section we describe the network analysis methods used to characterize the C . elegans connectome . Gene expression is represented as a binary indicator of which genes are expressed in a given neuron using data from many individual experiments compiled into a unified data resource on WormBase [63] . We use release WS256 of this dataset ( ftp://ftp . wormbase . org/pub/wormbase/releases/WS256/ONTOLOGY/anatomy_association . WS256 . wb ) and analyze annotations made ‘directly’ to individual neurons , excluding ‘uncertain’ annotations ( see S1 Text ) . We denote the expression of a gene in a neuron as a ‘1’ , and other cases as a ‘0’ . Note that a value of ‘0’ indicates either: ( i ) ‘gene is not expressed’ or ( ii ) ‘there is no information on whether the gene is expressed’ . Expression data are sparse , in part due to incomplete annotations—an average of 30 genes are expressed in each neuron ( range: 3 to 138 genes ) , and each gene is expressed in an average of 9 neurons ( range: 1 to 148 neurons ) . A total of 948 genes are expressed in at least one neuron , allowing us to represent the full expression dataset as a binary 279 ( neurons ) × 948 ( genes ) matrix , shown in Fig 1C . Our primary aim in this work is to understand how pairwise patterns of neuronal connectivity ( shown in Fig 1A ) relate to coupled expression across 948 genes between pairs of neurons ( i . e . , pairs of rows of Fig 1C ) . To estimate the coupling between neuronal gene expression profiles , we required a similarity measure for pairs of neurons that captures their similarity of gene expression profiles , or correlated gene expression ( CGE ) . We used a binary analogue of the linear Pearson correlation coefficient , the mean square contingency coefficient [74]: r ϕ = n 11 n 00 - n 10 n 01 n 1 • n 0 • n • 0 n • 1 , ( 3 ) for two vectors , x and y , of length L ( = 948 ) , where nxy counts the number of observations of each of the four outcomes ( e . g . , n10 = ∑i δxi , 1δyi , 0 counts the number of times x = 1 and y = 0 ) , and the symbol • sums across a given variable ( e . g . , n•0 = ∑i δyi , 0 counts the number of times y = 0 ) . The coefficient assumes a maximum value rϕ = 1 when x and y are identical ( such that n11 + n00 = L ) , and a minimum value rϕ = −1 when x and y are always mismatched ( such that n10 + n01 = L ) . Note that we use the notation rϕ to denote the phi coefficient , Eq ( 3 ) ; this notation should not be confused with the rich-club coefficient , ϕ ( k ) . One concern about applying this measure to sparsely annotated data is that it may be biased by differences in the number of expressed genes in a neuron , which ranged from 3 ( 0 . 3% of 948 genes analyzed here ) to 138 ( 14 . 6% ) . To explore this further , we compared rϕ with several other commonly used metrics of association between binary vectors including Yule’s Q and the Jaccard index . While these other binary similarity metrics exhibited strong bias with the proportion of gene expression annotations made to a given neuron , rϕ was not biased ( see S2 Fig and S2 Text ) . The 92 bilateral pairs of neurons ( e . g . , AVAL/AVAR , CEPVL/CEPVR , etc . ) exhibit highly correlated gene expression patterns: all bilateral pairs of neurons have rϕ > 0 . 8 , and 96% of bilateral pairs have rϕ > 0 . 95 . Although including bilateral pairs of neurons do not change the main results of this paper , we excluded CGE values of bilateral pairs of neurons from all analyses to ensure that our results are not driven by high CGE between these pairs . Our CGE measure , rϕ , quantifies the similarity between the expression profiles of two neurons across all 948 genes . To further investigate the role of individual genes in producing different CGE patterns , we developed a method for scoring the contribution of each individual gene to the overall correlation coefficient , following prior work using continuous gene expression data [38] . Performing similar analyses with C . elegans data poses additional challenges due to: ( i ) binary expression data , making robust quantification difficult; ( ii ) sparse and incomplete data , posing statistical problems for quantifying a signal in genes with limited expression; and ( iii ) low genome coverage ( less than 5% of the protein coding genes in C . elegans ) , constraining our ability to perform a comprehensive enrichment analysis , e . g . , using the Gene Ontology ( GO ) [75] . Given that rϕ treats mutual gene expression ( i . e . , cases in which a gene is expressed in pairs of neurons , n11 ) the same as mutual absence of gene expression ( n00 ) , we started by developing a new analytic measure of the probability of mutual gene expression , given its clearer biological interpretation ( see S3 Text ) . This measure was not biased by differences in the relative number of expressed genes ( S2D Fig ) and yields qualitatively similar outputs to rϕ on our data . Thus , while we use rϕ throughout this work for its ease of interpretation ( as an analogue of the conventional correlation coefficient ) , our new probability-based CGE measure allowed us to motivate a method for quantifying the contribution of individual genes ( and functional groups of genes ) to patterns of CGE that addresses some of the above-mentioned challenges . Note that our main findings , obtained using rϕ , are replicated using our new CGE measure ( cf . S3 Fig ) . As a starting point , we quantified the contribution of individual genes to differences in CGE for different categories of neuron pairs , specifically for ( i ) increased CGE in connected compared to unconnected pairs of neurons , and ( ii ) increased CGE in rich and feeder compared to peripheral connections . Note that our method scores genes on their contribution to differences in CGE between categories of pairwise connections . We first scored each gene for whether it is more likely to be expressed in a given class of neuron pair over another as the probability of obtaining at least as many matches ( defined as expression in both neurons of a pair ) as observed under a random CGE null model using the binomial distribution: p ( a ) = 1 - ∑ i = 0 m - 1 ( n i ) p class i ( 1 - p class n - i ) , ( 4 ) where m is the number of matches ( a match indicates that a given gene was expressed in both neurons ) on the class of neuron pairs of interest , n is the total number of matches across all neuron pairs considered in the analysis , pclass = nclass/M is the probability of the given class of inter-region pairs , as the total number of neuron pairs of that class , nclass , divided by the maximum number of possible neuron pairs , M , for a given gene , indexed with a . This score , p ( a ) , can be interpreted as a p-value under the null hypothesis that the number of expression matches of gene a is consistent with a purely random pattern of matches/mismatches across edges , giving lower values to genes with more matches in the edge class of interest ( compared to an alternative set of edges ) than expected by chance . For reasons described earlier , bilateral pairs of neurons were excluded from all scoring procedures and , to ensure that each gene contributes a meaningful score , we imposed a data quality threshold on the number of possible matches , n ≥ 10 . Our first analysis compares two mutually exclusive types of neuron pairs: ( i ) all pairs of neurons that are connected by at least one chemical or electrical synapse , and ( ii ) all pairs of neurons that are unconnected . For this analysis , pclass = 0 . 059 is the proportion of neuron pairs that are connected , n is the total number of neuron pairs that both exhibit expression of gene a , and m is the number of neuron pairs that are structurally connected for which both neurons express gene a . A total of 414 ( /948 ) genes had n ≥ 10 for this analysis . Our second analysis compares pairs of connected neurons for which at least one is a hub ( i . e . , rich , feed-in , or feed-out connections ) , to pairs in which both neurons are nonhubs ( i . e . , peripheral connections ) . In this case , pclass = 0 . 35 is the proportion of connected pairs of neurons that involve hubs , n is the number of connected neuron pairs for which gene a is expressed in both , and m counts the number of connected neuron pairs involving hubs for which gene a is expressed . A total of 168 ( /948 ) genes had n ≥ 10 for this analysis . As well as interpreting the list of individual genes that were significant after correcting for multiple hypothesis comparison , we performed an over-representation analysis ( ORA ) using the genes that contribute most to a given connectivity pattern to assess whether any gene sets ( GO categories ) were statistically over-represented in this list . To obtain the gene list , we used the false discovery rate correction of Benjamini and Hochberg [76] on p-values computed using Eq ( 4 ) , which were thresholded at a stringent level of p = 10−4 ( corresponding to approximately the top 20% of genes in each analysis ) . Over-representation for each biological process GO category with 5 to 100 genes available was quantified as an FDR-corrected p-value ( across around 700 GO categories ) using version 3 . 0 . 2 of ErmineJ software [77] . Biological process GO annotations [75] were obtained from GEMMA [78] ( using Generic_worm_noParents . an . txt . gz downloaded on March 31 , 2017 ) . Gene Ontology terms and definitions were obtained in RDF XML file format downloaded from archive . geneontology . org/latest-termdb/go_daily-termdb . rdf-xml . gz on March 31 2017 .
Previous work has demonstrated the importance of spatial effects in driving patterns of gene expression , with more proximal brain areas [38 , 53 , 79–84] having both a increased connection probability and more similar gene expression profiles [38 , 42 , 85 , 86] than more distance brain areas . Unlike network analyses of mammalian brains , where all neurons are confined to a spatially contiguous organ , neurons of the C . elegans nervous system are distributed throughout the entire organism , forming a dense cluster of 147 neurons in the head ( all within 130 μm ) , 105 sparser neurons in the body ( spanning 1 . 02 mm ) , which are predominantly motor neurons ( 75% ) , and another dense cluster of 27 neurons in the tail ( all within 90 μm of each other ) , as plotted in Fig 2 . In order to examine the relationship between connectivity and CGE , we need to understand the spatial dependence of both connectivity and CGE to characterize the extent to which previously reported spatial dependencies of these measurements apply to the microscale nervous system of C . elegans . We first characterize the probability that two neurons will be connected given their source and target types , labeling each neuron as being in either the ‘head’ , ‘body’ , or ‘tail’ of C . elegans . Connection probability is plotted as a function of Euclidean separation distance in Fig 3 for each combination of source and target neuron labels , across 10 equiprobable distance bins ( with exponential fits added for visualization ) . Distinguishing connections by source and target neuron types uncovers clear spatial relationships ( that are obscured when all connections are grouped , as in [53] ) , that differ across connection classes . From the very short distance scale of ⪅ 100 μm of head→head and tail→tail connections to the very longest-range head→tail and tail→head connections ( ⪆ 1 mm ) , connection probability decreases with separation distance ( Fig 3A ) . For connections between pairs of neurons located in the body , ranging up to ≈ 1 mm , a near-exponential trend is exhibited , mirroring results in other species and across longer length scales [80] , including mouse [38 , 87] , and in rodents and primates [79] . Other connections do not exhibit strong spatial connectivity relationships , i . e . , connections between the body and head or between the body and tail , shown in Fig 3B . We next investigate the dependence of CGE , rϕ , on the separation distance between neuron pairs , shown in Fig 4 . CGE decreases slightly with separation distance for the spatially close neurons within the head ( Fig 4A ) and within the tail ( Fig 4B ) , but not for pairs of neurons involving the body ( Fig 4C ) . The decreasing trend in CGE with distance within the head and tail is primarily driven by a subset of nearby neurons with high rϕ . It may therefore represent a relationship specific to particular functionally related neurons , rather than a general , bulk spatial relationship seen in macroscopic mammalian brains [38] . Accordingly , attempting to correct for a bulk , non-specific trend by taking residuals from an exponential fitted to the relationship produced artifactual reductions in the CGE of many neuron pairs ( shown in S4 Fig ) . Thus , we found no evidence for bulk spatial relationships of rϕ in the neuronal connectome of C . elegans . Next we analyze the topological properties of the C . elegans connectome , represented as a directed , binary connectivity matrix between 279 neurons , combining directed chemical synapses and undirected electrical gap junctions ( Fig 1 ) . The degree distribution is shown in Fig 5A , where neurons are labeled as sensory neurons , interneurons , motor neurons , or neurons with multiple functional assignments . Consistent with the results of Towlson et al . [28] , who used an undirected version of the connectome ( by ignoring the directionality of chemical synapses ) , we found a positively-skewed degree distribution containing an extended tail of high-degree hub interneurons . Hub interneurons of C . elegans are mostly command interneurons and mediate behaviors like coordinated locomotion and foraging [88] . Using the normalized rich-club coefficient , Φnorm , to quantify the extent to which hubs are densely interconnected , we confirmed the results of Towlson et al . [28] , finding rich-club organization in the connectome , as shown in Fig 5B . The figure plots the variation of Φnorm across degree thresholds , k , at which hubs are defined ( as neurons with degree > k ) , with red circles indicating a significant increase in link density among hubs relative to 1000 degree-preserving nulls ( permutation test , p < 0 . 05 ) . The plot reveals rich-club organization ( Φnorm > 1 ) at the upper tail of the degree distribution , particularly across the range 44 < k < 63 , shaded gray in Fig 5B . Similar results were obtained using weighted representations of the connectome ( i . e . , using information about the number of synapses in the connectivity network ) for two different definitions of the weighted rich-club coefficient [89] , shown in S5 Fig . Throughout this work , we define a set of hubs as the sixteen neurons with k > 44 , which corresponds to the lowest degree threshold at which the network displays a contiguous region of significant rich-club organization at high k . Our list of hubs includes all of the 11 hub neurons of Towlson et al . [28] at 3σ ( see S1 Table ) , with five additional hubs identified in our analysis of the directed connectome . The rich-club connectivity of the C . elegans connectome is accompanied by an increase in mean hub-hub connection distance [28] , with a significant increase through the topological rich-club regime ( right-tailed Welch’s t-test , p < 0 . 05 ) , shown in Fig 5B . This can be attributed to a relative increase in long-distance hub-hub connections between the head and tail , shown in Fig 2 ( cf . S6 Fig ) . The high connection density and long mean anatomical distance between pairs of hub neurons counters the general trend in the C . elegans connectome , where the probability of connectivity between two neurons decays with their separation distance ( Fig 3 ) . These results are consistent with previous findings across diverse neural systems and suggest that the rich club may provide a central yet costly backbone for neuronal communication in C . elegans [28 , 33] . We next investigate how the network connectivity properties of C . elegans relate to patterns of CGE , using the mean square contingency coefficient , rϕ . To test whether CGE varies as a function of connectivity , we compared the distribution of rϕ between ( i ) all connected pairs of neurons , and ( ii ) all unconnected pairs of neurons . Connected pairs of neurons have more similar expression profiles than unconnected pairs ( Wilcoxon rank-sum test , p = 1 . 8 × 10−78 ) . Fig 6A ( left ) shows distributions of rϕ for: ( i ) all pairs of neurons that are connected via electrical gap junctions ( 474 pairs , after excluding bilateral pairs ) , ( ii ) all pairs of neurons that are connected via reciprocal ( 291 pairs ) and , ( iii ) unidirectional chemical synapses ( 1721 pairs ) as well as ( iv ) all pairs of neurons that have neither connection ( 36 450 pairs ) . Note that 175 pairs of neurons are connected by both chemical synapses and gap junctions , and are thus included in both chemical and electrical categories . Amongst connected pairs of neurons , those connected via gap junctions exhibit more similar gene expression profiles than those connected via chemical synapses ( Wilcoxon rank-sum test , p = 5 . 4 × 10−22 ) . We found no difference in CGE between pairs of neurons connected reciprocally by chemical synapses ( N1 ↔ N2 for two neurons N1 and N2 ) versus those connected unidirectionally ( N1 → N2 ) ( Wilcoxon rank-sum test , p = 0 . 99 ) . We next investigated whether CGE varies across different types of connections defined in terms of their hub connectivity . For a given hub threshold , k , we first labeled each neuron as either a ‘hub’ ( nodes with degree > k ) or a ‘nonhub’ ( degree ≤ k ) , and then labeled each connection as either ‘rich’ ( hub → hub ) , ‘feed-in’ ( nonhub → hub ) , ‘feed-out’ ( hub → nonhub ) , or ‘peripheral’ ( nonhub → nonhub ) . The median CGE , r ˜ ϕ , of each of these four connection types is plotted in Fig 6B , with circles indicating statistically significant increases of a given connection type relative to all other connections ( one-sided Wilcoxon rank-sum test , p < 0 . 05 ) . Correlated gene expression in rich connections increases with degree , k , particularly in the topological rich-club regime where hubs are densely interconnected ( shaded gray in Fig 6B ) . In this topological rich-club regime , both feed-in and feed-out connections exhibit increased CGE relative to peripheral connections , which show the lowest levels of CGE . Full distributions of rϕ for each edge type at a hub threshold of k > 44 are in Fig 6A ( right ) . This plot shows that , compared to all different types of pairs of neurons , connected pairs of hubs showed the highest CGE . In summary , our results reveal: ( i ) increased CGE in connected pairs of neurons; ( ii ) the highest CGE in rich connections; and ( iii ) lowest CGE in peripheral connections . These results , obtained using incomplete binary annotations of gene expression across 948 genes in a microscale neuronal connectome , are consistent with a prior analysis of the expression of over 17 000 genes across 213 regions of the mesoscale mouse connectome [38] . The sixteen hub neurons in C . elegans ( k > 44 ) : are all interneurons , are all located in either the head or tail , are mostly contained within a single topological module of the network , are mostly cholinergic ( 13/16 ) , and are all born prior to hatching . We therefore investigated whether the similarity of gene expression profiles between hubs is specific to their high levels of connectivity , or whether it could instead be driven by these other characteristics . Having characterized a robust relationship between CGE and ( i ) connectivity , and ( ii ) hub connectivity , we next investigated which specific genes contribute most to this relationship . Despite challenges with the incomplete binary expression measurements in a small proportion of the genome , we developed a method to score genes according to their contribution to a given CGE pattern ( see Methods ) . We characterized individual high-scoring genes , with pcorr < 10−4 ( approximately 20% of genes with the highest scores in each analysis ) , and attempted to summarize functional groups of genes as biological process categories of the gene ontology ( GO ) that were enriched in high scoring genes using overrepresentation analysis ( ORA ) [75 , 77] . We first investigated which genes drive increased CGE in connected pairs of neurons relative to unconnected pairs . Previous studies in mouse have indicated that genes driving an increase in CGE between connected pairs of brain regions are enriched in GO categories related to neuronal connectivity and communication [38 , 96–98] . First , we manually investigated individual high-scoring genes ( i . e . , those with pcorr < 10−4 , see Supplementary Data File ( S1 File ) ) . Given that glutamate is a prevalent neurotransmitter in C . elegans ( 26% of neurons with known neurotransmitter type are glutamatergic [64] ) , it is appropriate that many high scoring genes are related to glutamate receptors ( including glr-1 , glr-2 , glr-4 , glr-5 , nmr-1 , and nmr-2 ) . Consistent with the importance of innexins in forming electrical synapses [99] , our list contained the following innexin genes: unc-9 , unc-7 , inx-7 , inx-19 , inx-13 . Genes encoding cell adhesion molecules related to axon outgrowth and guidance , cell migration and locomotion ( sax-3 , cam-1 , unc-6 , rig-1 , unc-5 ) , learning ( casy-1 ) [63 , 100–103] , as well as genes involved in determining cell polarity ( vang-1 , prkl-1 ) [104 , 105] were also amongst the top scoring genes for connectivity . These genes have been implicated in neuronal connectivity in both flies and humans [106–108] , with our results predicting that they may play a similar role in C . elegans . In addition , transcription factors regulating neuronal development , fate specification ( lin-11 , unc-3 , unc-42 , ceh-14 , ast-1 , cfi-1 ) [109–114] , and locomotion ( unc-3 ) [110] ) were also implicated in driving increased CGE amongst connected pairs of neurons . Both adhesion molecules and transcription factors are candidates for facilitating signal transduction and communication . In order to summarize the above mentioned results and determine if any particular functional groups of genes drive this effect , we performed an enrichment analysis . Top scoring biological process GO categories from ORA analysis ( of 85 genes relative to the 414 genes with sufficient data for this analysis ) are listed in S2 Table . Although no GO categories are significant at a false discovery rate of 0 . 05 , the top categories are consistent with a connectivity profile , including ‘glutamate receptor signaling’ , ‘cell surface receptor signaling’ , and ‘ion transport’ , with other categories involved in regulation of growth rate and several related to catabolic processes . Thus , despite incomplete gene expression data that do not provide sufficient coverage to detect statistically significant effects , these results indicate that our data-driven gene scoring method is able to yield sensible , biologically relevant insights into the genetic basis of neuronal connectivity in C . elegans connectome . Having characterized genes that contribute to the increase in CGE between connected pairs of neurons , we next investigated whether particular functional groups of genes drive differences in CGE between connections involving hub neurons ( i . e . , in rich , feed-in , and feed-out connections ) relative to connections between pairs of nonhub neurons ( i . e . , peripheral connections ) . In order to investigate which specific genes contribute most to the increase in CGE for connections involving hubs , we first investigated the highest-scoring genes , with pcorr < 10−4 ( corresponding to approximately the top 20% of genes in the analysis ) . In addition to glutamate receptor genes ( glr-5 , nmr-1 , nmr-2 , glr-1 , glr-2 , grld-1 ) and acetylcholine related genes ( ace-2 , cho-1 , unc-17 , deg-3 ) , we again find a high number of genes regulating cell adhesion ( cam-1 , rig-1 , rig-6 , unc-6 , grld-1 , dbl-1 , ncam-1 ) and relevant transcription factors ( unc-3 , unc-42 , ast-1 ) . The implication of glutamate and acetylcholine may be attributable to the importance of glutamate in the regulation of locomotion in command interneurons [115 , 116] , with acetylcholine being the dominant neurotransmitter in hubs ( 13 out of 16 hubs are cholinergic ) . We also find a high overlap between adhesion molecule and transcription factor encoding genes found in the previous analysis and the implication of human orthologs ( rig-1 , ncam-1 , grld-1 , corresponding to human genes ROBO4 , NCAM2 , and RBM15 respectively [63] ) for genes regulating cell migration , differentiation and neuron cell adhesion . While previous work implicated genes regulating oxidative metabolism for connections involving hubs in mouse [38] , and for hub regions in human [48] , the gene expression dataset used here was not sufficiently comprehensive to investigate these processes . For example , only one of the 948 genes annotated to the GO categories related to hub connectivity in mouse is present in our gene expression dataset ( unc-32 is annotated to the GO category: ‘ATP hydrolysis coupled proton transport’ ) . Thus , although a direct test of the metabolic hypothesis for neural hubs is not possible from current data , we investigated whether other biological process GO categories were overrepresented in pairs of connected hubs using ORA ( of 30 genes relative to the 168 genes with sufficient data for this analysis ) , with results listed in S3 Table . Even though no categories are statistically significant at a false discovery rate of 0 . 05 , the list of top categories includes both ‘glutamate receptor signaling pathway’ as well as more general ‘cell surface receptor signaling pathway’ in addition to several ion transport related gene groups ( ’ion transport’ , ‘ion transmembrane transport’ , ‘transmembrane transport’ ) . Other top-ranked GO categories include regulation of locomotion , and various metabolism and biosynthesis related processes . Our gene scoring method again yields interpretable insights into the types of genes that contribute to differences in CGE between different classes of neuronal connections in C . elegans . While current data are limited , more comprehensive expression annotations in the future would allow more systematic and statistically powered inferences across GO categories .
Highly connected hubs of neural systems play an important role in brain function , with their dense rich-club interconnectivity integrating disparate neural networks [24 , 27 , 29 , 30] . Here , our analysis linking hub connectivity of the microscale connectome of C . elegans to patterns of neuron-specific gene expression has identified a transcriptional signature that appears to be highly conserved , given recent findings reported in a mesoscale investigation of the mouse [38] and a macroscale study of humans [48] . Specifically , we show that: ( i ) CGE is higher for connected pairs of neurons compared to unconnnected pairs; ( ii ) the neuron connection probability decays as a function of spatial separation , and; ( iii ) connected pairs of hub neurons , which are generally separated by longer anatomical distances , show the highest levels of CGE . This association between CGE and hub connectivity followed a gradient , such that CGE was lowest for connected nonhubs , intermediate for hub-nonhub pairs , and highest for connected hubs , consistent with results reported in the mouse brain [38] . Amongst the genes considered here , many of those with the greatest contribution to connectivity are biologically plausible genes related to receptors , neurotransmitters , and cell adhesion , and those with the greatest contribution to hub connectivity are related to glutamate receptors , acetylcholine signaling , and other neuronal communication related genes . The methods we develop here for quantifying CGE , and for scoring the contribution of individuals genes to overall CGE , yield biologically interpretable results from incomplete binary gene expression data . With improvements in gene annotation quality and specificity , and increases in genome coverage , similar methods could be used in future work to characterize the biological basis of a range of neuronal connectivity patterns . The availability of spatial maps of gene expression with genome-wide coverage has allowed the relationship between gene expression and connectivity to be investigated in species ranging from C . elegans through to mouse and human . For example , Krienen et al . [42] showed that the topography of transcriptional expression of a small number of human supragranular enriched genes mirrors the large-scale brain network organization of rs-fMRI in the healthy human brain , and Romme et al . [117] showed that schizophrenia-related structural disconnectivity is significantly correlated to the expression profiles of schizophrenia risk genes . Recent work has demonstrated a relationship between spatial gene expression maps and cortical hierarchy using structural MRI imaging in macaque and human [118] . Spatial maps of gene transcription will continue to play a key role in uncovering species-conserved mechanisms underlying brain connectivity . Computational methods to extract relationships between network organization and gene expression can help understand the molecular processes underlying neuronal connectivity . Previous research has related gene expression data in C . elegans to axonal connectivity patterns , focusing on pairwise relationships of genes that might underpin axonal connectivity . Both Kaufman et al . [60] and Baruch et al . [119] developed statistical models to predict the postsynaptic partners of individual neurons in C . elegans ( using k-nearest neighbors and boosted decision tree models , respectively ) . This research found that the targets of some neurons are easier to predict than others [60] , and that the prediction can be done with good accuracy using only a small subset of genes [119] . Our results demonstrate differences in CGE across different topological classes of connections , and highlight genes that make the biggest contribution to these differences . More detailed investigations of these relationships ( e . g . , across C . elegans development ) may shed light on the molecular logic underlying the establishment and maintenance of neuronal connectivity . It is reasonable to expect that the principles of neural organization may differ from the scale of individual neurons to the scale of macroscopic brain regions ( in which each brain region contains millions of neurons ) . However , many of our results in C . elegans suggest a striking conservation of many fundamental spatial trends in neural connectivity and CGE across scales and species . For example , connection probability decreases with spatial separation between brain areas in rodents and primates [79 , 80] ( including in macaque [81] , human [82] , mouse [38] , and rat [83] ) , for individual neurons in mouse primary auditory cortex [84] , and between neurons in C . elegans ( cf . Fig . S1 of [53] ) . Unlike mammalian brains , where all neurons are confined to a spatially contiguous organ , neurons are distributed across nearly the entire length of C . elegans , including a dense cluster of neurons in the head and in the tail . Despite these distinct morphologies , we report a qualitatively similar spatial dependence of connection probability with separation distance for many classes of connections in C . elegans , including those within the head , body , and tail , indicating that this distance-dependence may be a generic property of evolved neuronal systems that must balance the energetic cost of long-range connections with their functional benefit [17 , 23 , 33 , 55] . Less frequently characterized is the spatial dependence of CGE , with available evidence indicating that more proximal brain areas exhibit more similar gene expression patterns than more distant brain areas in the mouse brain [38] and human cortex [42 , 85 , 86] . Some of the spatial trends in CGE found in the 948 genes analyzed here mirror these trends of bulk regions of macroscopic mammalian brains . It is therefore possible that these spatial dependences of connectivity and CGE may not be simply due to bulk spatial trends in macroscopic brains containing millions of neurons , but may reflect conserved organizational principles that hold across species and spatial scales . Our results highlight the importance of treating nervous systems as spatially embedded objects , as many seemingly non-trivial properties of brain organization may be well approximated by simple , isotropic spatial rules [22 , 50 , 79 , 82 , 120] ( see also [17 , 23] ) . Our analysis indicates that CGE patterns in C . elegans show many surprising similarities to previous work in the mesoscale mouse connectome [38] , despite: ( i ) involving different gene expression annotation data ( comprehensive in situ hybridization expression data across ∼20000 genes in mouse versus literature-curated annotations across ∼1000 genes in C . elegans ) , ( ii ) being a different type of neural system ( from the spatially continuous macroscopic brain of mouse , to the spatially separated nervous system of C . elegans ) ; ( iii ) orders of magnitude differences in spatial scale . The findings were also robust to a range of data processing choices , including different representations of the connectome ( e . g . , directed/undirected , or excluding electrical synapses ) , and across alternative metrics for quantifying transcriptional similarity . What could drive this highly conserved association between CGE and hub connectivity ? Here , we took advantage of the rich and diverse information available for each neuron of the C . elegans connectome to begin to address this question . We show that CGE between hub neurons is not determined by their neuronal type ( i . e . , the fact that all hubs are interneurons rather than sensory or motor neurons ) , since CGE between hub neurons is higher than between other pairs of interneurons . The effect cannot be attributed to the modular organization of the network either , since CGE between hubs in the same module is higher than between other pairs of neurons in the same topological module , with a similar increase in CGE for pairs of hubs in different modules . We also show that the effect is not driven by similarities in the birth time nor lineage distance of hub neurons , which exhibit higher CGE than other early-born neurons ( prior to hatching ) and are not closer in their lineage . Moreover , the abundance of cholinergic signaling of hub neurons cannot explain the effect . Rather , the CGE between pairs of hub neurons in C . elegans may be related to the specific functional role of these cells . Namely , 60% of them are command interneurons , which play a vital role in coordinating forward and backward locomotion in C . elegans [54] . The overlap between command interneurons and hub neurons has interesting parallels with the human cortex , where polymodal association areas tend to be the most highly connected network elements [1] . Association areas sit at atop the cortical hierarchy and support complex behaviors by integrating information from diverse neural systems [121] . Locomotion is arguably one of the most complex behaviors expressed by C . elegans . Thus , the association between hub status and command interneurons may reflect the specialization of these neurons for supporting higher-order functions in the behavioral repertoire of C . elegans . It is as yet unclear whether CGE between network hubs , regardless of species and scale , is simply a byproduct of tightly coupled hub activity , or some shared morphological or development characteristic between hubs that we have not captured in the present analysis . More comprehensive transcriptomic data ( e . g . , obtained through systematic single-neuron RNA sequencing ) , measured through development and coupled with measures of neuronal activity , would allow us to address these questions . Additionally , we cannot rule out the possibility that gene annotations have been influenced by the nature of the curated data that we have used here . Given their functional similarity , command interneurons might have been tested as a group in a set of experiments for the expression of particular genes and consequently assigned similar expression signatures . More precise and systematic measurement of neuron-specific gene expression patterns would be required to address this question . Studies of gene expression often assume that expression levels correspond to protein abundance , but this assumption does not always hold [122–124] . Thus , analyses of transcriptomic data can be viewed as a relatively efficient approach for investigating potential links between molecular function and nervous system organization , that can be more strongly verified using subsequent proteomic analysis . In this work we developed methods to relate correlations in binary gene expression data to pairwise connectivity and subsequently score and evaluate the contribution of individual genes to these patterns . Compared to continuous in situ hybridization measurements of the expression of > 17000 genes in the mouse brain [125] , or microarray measurements of > 20000 genes in the human brain [126 , 127] , which permit more detailed analysis [38 , 48 , 96–98 , 128] , working with C . elegans gene expression data is challenging due to its low coverage ( < 5% coverage of the worm genome ) , binary indications of expression , and incompleteness ( an inability to distinguish missing data from lack of expression ) . Moreover , the data have different qualifiers related to the certainty of gene expression annotations ( see S1 Text ) , requiring choices to be made to appropriately balance sensitivity and specificity . Although gene enrichment analyses did not have enough power to detect significant effects here , top GO categories point us towards biologically relevant categories related to neuronal connectivity , neurotransmitters , and metabolism . We note , however , that the incomplete coverage of the genome in our annotated dataset may mask many true GO associations . Our single gene analysis identified specific genes contributing to increases in CGE for connected pairs of neurons and for connections involving hub neurons . In line with our expectations , genes regulating both chemical and electrical signaling , namely glutamate receptor and innexin genes , were implicated in general connectivity . In addition , we also find multiple cell adhesion molecule genes and transcription factors that regulate neuronal development and fate specification—both groups are important for forming neuronal connections . High overlap between genes encoding adhesion molecules and transcription factors implicated in regulating both general and hub connectivity highlights that related mechanisms might be used in both cases . While we were not able to test GO categories related to neurotransmitter signaling comprehensively , due to insufficient coverage of gene expression annotations , single gene analysis revealed the importance of acetylcholine genes , which may be related to the fact that acetylcholine is the dominant neurotransmitter in hub neurons .
|
Some elements of neural systems possess many more connections than others , marking them as network hubs . These hubs are often densely interconnected with each other , forming a so-called rich-club that is thought to support integrated function . Recent work in the mouse suggests that connected pairs of hubs show higher levels of transcriptional coupling than other pairs of brain regions . Here , we show that hub neurons of the nematode C . elegans also show tightly coupled gene expression and that this effect cannot be explained by the spatial proximity or anatomical location of hub neurons , their chemical composition , birth time , neuronal lineage or topological module affiliation . Instead , we find that elevated coexpression is driven by the identity of most hubs of the C . elegans connectome as command interneurons , a specific functional class of neurons that regulate locomotion . These findings suggest that coupled gene expression is a highly conserved genomic signature of neural hubs that may be related to the specific functional role that hubs play in broader network function .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"caenorhabditis",
"nervous",
"system",
"electrophysiology",
"neuroscience",
"animals",
"motor",
"neurons",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
"brain",
"mapping",
"experimental",
"organism",
"systems",
"interneurons",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"gene",
"expression",
"connectomics",
"cellular",
"neuroscience",
"eukaryota",
"neuroanatomy",
"cell",
"biology",
"anatomy",
"synapses",
"physiology",
"neurons",
"genetics",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neurophysiology",
"organisms"
] |
2018
|
Hub connectivity, neuronal diversity, and gene expression in the Caenorhabditis elegans connectome
|
Cyclic AMP-dependent pathways mediate the communication between external stimuli and the intracellular signaling machinery , thereby influencing important aspects of cellular growth , morphogenesis and differentiation . Crucial to proper function and robustness of these signaling cascades is the strict regulation and maintenance of intracellular levels of cAMP through a fine balance between biosynthesis ( by adenylate cyclases ) and hydrolysis ( by cAMP phosphodiesterases ) . We functionally characterized gene-deletion mutants of a high-affinity ( PdeH ) and a low-affinity ( PdeL ) cAMP phosphodiesterase in order to gain insights into the spatial and temporal regulation of cAMP signaling in the rice-blast fungus Magnaporthe oryzae . In contrast to the expendable PdeL function , the PdeH activity was found to be a key regulator of asexual and pathogenic development in M . oryzae . Loss of PdeH led to increased accumulation of intracellular cAMP during vegetative and infectious growth . Furthermore , the pdeHΔ showed enhanced conidiation ( 2–3 fold ) , precocious appressorial development , loss of surface dependency during pathogenesis , and highly reduced in planta growth and host colonization . A pdeHΔ pdeLΔ mutant showed reduced conidiation , exhibited dramatically increased ( ∼10 fold ) cAMP levels relative to the wild type , and was completely defective in virulence . Exogenous addition of 8-Br-cAMP to the wild type simulated the pdeHΔ defects in conidiation as well as in planta growth and development . While a fully functional GFP-PdeH was cytosolic but associated dynamically with the plasma membrane and vesicular compartments , the GFP-PdeL localized predominantly to the nucleus . Based on data from cAMP measurements and Real-Time RTPCR , we uncover a PdeH-dependent biphasic regulation of cAMP levels during early and late stages of appressorial development in M . oryzae . We propose that PdeH-mediated sustenance and dynamic regulation of cAMP signaling during M . oryzae development is crucial for successful establishment and spread of the blast disease in rice .
Heterotrimeric G protein signaling utilizes cyclic AMP ( cAMP ) as a second messenger , to mediate the transduction of extracellular stimuli to the intracellular downstream signaling components in several eukaryotes , including the pathogenic fungi . The cAMP pathway is a highly conserved signaling module that influences and regulates a range of fundamental cellular processes in growth , development and morphogenesis . In response to ligand-stimulated GPCRs , spatially segregated “point sources” of cAMP are generated through GαS based activation of membrane anchored adenylyl cyclases . In response to extracellular stimuli , multiple point sources of cAMP are generated throughout the cell , leading to the gradual accumulation and increase in the basal or steady-state levels of cAMP . On attaining a critical threshold concentration , cAMP can further activate several important effectors ( foremost being cPKA , a cAMP-dependent Protein Kinase A ) , which in turn mediate a wide array of downstream physiological effects [1] . The inactivation of cAMP to 5′-AMP is carried out through enzymatic hydrolysis , by phosphodiesterases ( PDEs ) . Such an inactivation of cAMP regulates the overall strength and intensity of the signaling cascade and is also necessary for efficient signal compartmentalization and termination [2] , [3] . In order to achieve this , PDEs are targeted to specific intracellular sites or signaling complexes and are known to localize not only to the cytosol , but also to a variety of membrane , nuclear and cytoskeletal locations [4] , [5] , [6] , [7] , [8] . In addition , PDEs establish and shape concentration-dependent gradients of cAMP at distinct regions within a cell [9] , [10] , [11] . Thus , PDEs play an important role in regulating the specificity , amplitude and temporal duration of cAMP signaling [2] , [12] . In fungi and yeasts , cAMP signaling cascade has been co-opted for a multitude of cellular processes and development . For example , in yeasts like S . cerevisiae , cAMP regulates nutrient sensing , pseudohyphal differentiation , cell cycle progression and stress signaling [13] , [14] , [15] , [16] , [17] , [18] . In S . pombe , mating , sporulation and gluconeogenesis are controlled through the cAMP pathway [19] , [20] , [21] , [22] . In pathogenic fungi , such as C . albicans and C . neoformans , cAMP signaling influences sexual differentiation , stress tolerance and several important aspects of virulence [23] , [24] , [25] , [26] , [27] , [28] , [29] , [30] , [31] , [32] , [33] . In the plant-pathogenic fungus U . maydis , cAMP governs dimorphic transition in addition to virulence [34] , [35] , [36] , [37] , [38] . Morphogenesis , cell polarity and asexual development are regulated through cAMP in N . crassa and A . nidulans [39] , [40] , [41] , [42] , [43] , [44] , [45] , [46] , [47] , [48] , [49] . Fluctuations in cAMP levels are modulated by cAMP phosphodiesterases in yeasts ( including the dimorphic pathogens ) but not in filamentous fungal species . S . cerevisiae contains a low-affinity phosphodiesterase ( Pde1; with a high Km towards cAMP ) and a high-affinity phosphodiesterase ( Pde2; with a low Km towards cAMP ) [50] , [51] , [52] , [53] . Pde1 regulates cAMP levels induced by glucose stimulation or intracellular acidification , and is in turn regulated through phosphorylation by cPKA [54] , [55] , [56] . Pde2 regulates basal or steady state levels of cAMP , in addition to protecting the yeast cells from extracellular cAMP [17] , [54] , [57] . Although poorly understood , Pde2 regulation is mediated through the cPKA pathway . Neither of the two PDEs is indispensable for cell growth under standard culture conditions but are primarily needed to overcome stress and nutritional starvation [17] , [58] , [59] . The fission yeast S . pombe , lacks a Pde2 ortholog , but possesses a Pde1 protein , which functions to regulate cAMP levels induced by glucose stimulation in a manner likely dependent on cPKA [21] , [60] . A definite role has not yet been assigned to the Pde1 in C . albicans development . However , the C . albicans Pde2 has been functionally well characterized [23] , [25] , [61] . Similar to the effects caused in S . cerevisiae , deletion of Pde2 leads to elevated levels of intracellular cAMP , increased responsiveness to exogenous cAMP and sensitivity to heat shock [23] , [27] . In addition , Pde2 mutant displays defects in cell wall and membrane integrity , and is more sensitive to a range of antifungal agents [24] , [25] . Furthermore , the pde2 mutant shows enhanced filamentation but is avirulent in a murine model of systemic Candidiasis [23] . The Pde1 and Pde2 have been functionally characterized in Cryptococcus neoformans , which represents another well-studied human fungal pathogen . Unlike in S . cerevisiae and C . albicans , Pde2 deletion in C . neoformans results in only subtle and mild phenotypic defects . In contrast , Pde1 regulates the basal levels of cAMP but does not respond to cAMP induced in response to glucose induction . However , deletion of PDE1 conferred only mild phenotypic defects . Furthermore , Pde1 activity is regulated through cPKA derived phosphorylation [30] . Mouse strains deficient in PDE activity are viable , but exhibit poor growth , increased prenatal mortality or female infertility [62] , [63] . Taken together , the phenotypes displayed by the PDE deletion strains in either fungi or in mammals , imply that , rather than executing a strict regulatory role , PDEs function to modulate and streamline cAMP signaling . M . oryzae is an ascomycete fungus that causes the blast disease in rice and several other monocot species [64] . It reproduces asexually when stimulated by light , typically producing three to four conidia each on stalk-like structures known as conidiophores [65] , [66] , [67] . These asexual spores or conidia aid in the spread of the blast disease . Under conditions of high humidity , the asexual spore produces a germ tube which senses and responds to host surface stimuli by forming an infection structure known as an appressorium [68] . In M . oryzae , cAMP signaling is required for conidiation and appressorium initiation [69] , [70] , [71] . Although an Adenylate cyclase ( mac1 ) function has been shown to be crucial for M . oryzae pathogenesis , key downstream modulators of cAMP signaling are yet to be identified [72] . In this study , we were interested in deciphering the roles of phosphodiesterases in modulating the cAMP levels and signaling during various stages of M . oryzae development . Towards this end , we identified a high- and low-affinity cAMP phosphodiesterase in M . oryzae and further characterized gene-deletion strains of both the cAMP phosphodiesterases . Our results show that PdeH plays an important role in regulating the steady-state levels of cAMP during asexual , pathogenic and invasive growth in M . oryzae . We analyzed the transcriptional regulation and localization of PDEs and demonstrate that cAMP signaling is compartmentalized in M . oryzae and that it responds rapidly to modulate and maintain the appropriate levels of cAMP during growth , development and infection .
Using complete amino acid sequences , we identified orthologs of the yeast Pde1 and Pde2 in M . oryzae ( http://www . broadinstitute . org/annotation/genome/magnaporthe_grisea/MultiHome . html ) . MGG_05664 ORF ( GQ869475 ) was predicted to encode an 893 aa protein , showing 27% identity to Pde2; whereas MGG_07707 ( GQ869476 ) was a 585 aa polypeptide with 32% overall identity to yeast Pde1 . The complete cDNA sequence was determined in each instance ( Genbank accession GQ869476 and GQ869476 , respectively ) and the predicted amino acid sequences confirmed . MGG_05664 showed the conserved PDE class I consensus sequence [73] permitting us to designate it as a high-affinity phosphodiesterase ( Figure 1A ) . Based on the conserved PDE class II consensus [30] , MGG_07707 was likewise judged to be the low-affinity cAMP phosphodiesterase in M . oryzae ( Figure 1B ) . In order to avoid confusion with Pde1 , a P-Type ATPase [74] , we hereafter refer to the low-affinity phosphodiesterase as PdeL and the high-affinity phosphodiesterase as PdeH in M . oryzae . To understand the role of cAMP phosphodiesterases during growth and morphogenesis in M . oryzae , we generated gene-deletion mutants of PDEH and PDEL in the B157 wild type background . Agrobacterium-mediated gene targeting was used to replace the entire ORF of the PDEH or PDEL gene with the hygromycin-resistance ( Figure S1A ) or the bialaphos-resistance marker cassettes ( Figure S1B ) , respectively . Selected transformants were screened by locus-specific PCR and confirmed by Southern blotting ( Figure S1C and S1D ) . Two independent strains for pdeHΔ and pdeLΔ were used for further investigations . When grown on Prune agar ( PA ) medium for 7 days at 28°C , the pdeLΔ was similar to the wild-type strain and showed no obvious defects in aerial or radial growth and colony morphology ( Figure 1C ) . In contrast , the pdeHΔ colony was flat due to reduced aerial hyphal growth , and displayed enhanced pigmentation and marginally slower radial growth at analogous time points ( Figure 1C ) . To further understand the relationship between PDEH and PDEL , we generated a pdeHΔ pdeLΔ double mutant ( Figure S1E ) . Compared to the pdeHΔ , the double deletion mutant showed a flat colony appearance with enhanced pigmentation and severely reduced aerial hyphal growth ( Figure 1C and S3A ) , thus underscoring the importance of proper PdeH and PdeL function in M . oryzae . In addition , radial growth was compromised in the double mutant . Next , we attempted a genetic complementation of pdeHΔ by introducing a full-length PDEH fused to RFP at the N-terminus under native regulation . The complemented pdeHΔ strain showed complete suppression of the aerial and radial growth defects exhibited by the pdeHΔ strain ( Figure S2A and 2B ) , suggesting that the phenotypic defects therein were solely due to the loss of PdeH function . Taken together , these results indicate that PdeH is essential for proper aerial hyphal growth and development in M . oryzae . Furthermore , we infer that PdeL plays only a minor role in regulating vegetative and aerial growth in M . oryzae . Asexual reproduction commences in M . oryzae with the formation of specialized aerial conidiophores that go on to produce conidia in a sympodial manner [65] . We were particularly curious about the conidiation status of pdeHΔ since it showed reduced aerial growth . Rather surprisingly , conidiation was two to three-fold higher in the pdeHΔ , when compared to pdeLΔ or the wild type ( Figure 2A ) . Furthermore , the pdeHΔ formed 5–8 conidia per conidiophore unlike the wild type conidiophores , which typically produce 3 conidia each ( Figure 2C ) . The pdeLΔ pdeHΔ strain interestingly was severely blocked in conidiation ( 98±1 . 0% reduced in conidiation when compared to the wild type ) and failed to produce proper conidia and instead formed highly pigmented aberrant structures of varied morphologies ( Figure 2A , 2C and S3A ) or two celled conidia-like structures at a very low frequency ( ≤2%; Figure 2C inset and S3C; p<0 . 05 ) . Next , we quantified conidiophore formation in the above four strains and found that the pdeHΔ produced twice the number of conidiophores per microscopic field , compared to the pdeLΔ or the wild type ( Figure 2B ) . The pdeHΔ pdeLΔ was also defective in conidiophore differentiation , and resulted in highly pigmented structures that were misshapen and failed to grow further and form conidia ( Figure 2B , 2C and S3A ) . The pdeLΔ and the complemented pdeHΔ ( RFP-PdeH expressing strain ) produced nearly the same number of conidia and conidiophores as the wild type and did not show any apparent defects during any stage of asexual development ( Figure 2A and 2B ) . Taken together , our results suggest that PdeH regulates multiple aspects of asexual development , primarily influencing conidiophore differentiation and conidia formation . Furthermore , we concur that PdeH function ( and by inference cAMP levels ) likely determines the pattern and number of conidia formed on conidiophores . Furthermore , we conclude that the simultaneous loss of PDE functions leads to deleterious effects in asexual differentiation in M . oryzae . The inability of the pdeHΔ pdeLΔ to form proper conidiophores and conidia suggests a regulatory albeit ancillary function for PdeL during asexual development . In order to assess whether the phenotypic defects in pdeHΔ were due to elevated levels of cAMP , we quantified and compared the steady-state levels of cAMP during the conidiation phase in the pdeHΔ , pdeLΔ , pdeHΔ pdeLΔ and the wild type ( Figure 3A ) . First , we measured cAMP levels in cultures grown in the dark for 3 d and compared it to those in cultures exposed to constant illumination for a period 24 h ( after a 3 d incubation in the dark ) . We found that even in the absence of light stimulation ( darkness ) , the pdeHΔ strain accumulated ∼3 fold ( p<0 . 005 ) and ∼2 . 6 fold ( p<0 . 005 ) higher levels of cAMP , compared to the wild type or the pdeLΔ respectively ( Figure 3A ) . Under similar conditions , the double deletion mutant accumulated ∼3 fold ( p<0 . 001 ) higher cAMP compared to pdeHΔ and ∼10 fold higher levels compared to the wild type or pdeLΔ ( p<0 . 001 ) . Furthermore , during conidiation ( 24 h photo-induced ) the cAMP levels were significantly up regulated in all the strains except the double deletion mutant: 2 fold ( p<0 . 05 ) in the wild type and pdeLΔ , and by 1 . 4 fold ( p<0 . 01 ) in the pdeHΔ when compared to the respective dark grown cultures . However , in the pdeHΔ pdeLΔ , the cAMP levels were consistently higher , irrespective of the growth conditions ( Figure 3A ) . Thus , the pdeHΔ pdeLΔ and pdeHΔ mutants accumulate significantly higher levels of cAMP compared to the wild type or pdeLΔ . We thus conclude that PdeH is an important regulator of cAMP signaling during asexual development in the rice blast fungus . Next , we measured the steady-state levels of cAMP in the pdeHΔ , pdeLΔ and the wild type during pathogenic development ( Figure 3B and 3C ) . The intracellular levels of cAMP could not be estimated reliably during the pathogenic phase of development in the pdeHΔ pdeLΔ strain , as it failed to conidiate properly and the resultant aberrant structures were scarce . The various developmental stages and time points that we considered for the cAMP measurements were as follows: ungerminated conidia ( 0 h ) , appressorium initiation ( 6 h ) , mature appressorium ( 21 h ) , penetration stage ( 24 h ) and infection hyphae formation stage ( 28–30 h ) [75] . The levels of cAMP were ∼2 . 5 fold ( p<0 . 005 ) higher in freshly harvested pdeHΔ conidia when compared to the wild type or the pdeLΔ ( Figure 3B ) . At 6 hours post inoculation ( hpi ) , the aforementioned strains accumulated higher levels of cAMP , but the overall level was significantly higher in the pdeHΔ ( 1 . 5 fold; p<0 . 005 ) . In mature appressoria ( 21 hpi ) the cAMP levels decreased across all the strains , except the pdeHΔ , which sustained a two-fold higher level of cAMP compared to wild type or pdeLΔ ( Figure 3B ) . At 24 hpi , the cAMP levels decreased further by two-fold across all the strains with the exception of the pdeHΔ , which continued to accumulate elevated cAMP levels . At the stage of host penetration ( 24–30 hpi ) , the pdeHΔ maintained comparatively high cAMP levels ( 2 to 5 fold; Figure 3C ) unlike the wild type or pdeLΔ . We therefore infer that the overall steady-state levels of cAMP are regulated in two distinct phases in M . oryzae: upregulated during the early stages ( appressorium initiation and formation ) , while being down regulated at the late stages ( host invasion ) of pathogenic development . We thus conclude that PdeH-based regulation helps maintain functionally relevant levels of cAMP during infection-related morphogenesis . We propose a minor role for PdeL in regulating cAMP levels in M . oryzae . However , based on the cAMP levels observed in the pdeHΔ pdeLΔ and the pdeHΔ , we suggest that PdeL function likely gains more importance in the absence of PdeH . Inductive surface cues such as hardness and hydrophobicity , which naturally prevail on leaves , can be mimicked using artificial membranes ( GelBond , Lonza Walkersville Inc . , USA ) for in vitro appressorial assays . The wild type is incapable of appressoria formation on non-inductive surfaces , but can do so only in the presence of exogenous cAMP or inhibitors of cAMP phosphodiesterases , suggesting that increased levels of cAMP may play a fundamental role in initiating early signaling events in appressorium formation [69] . Therefore , we asked whether elevated cAMP could influence precocious initiation of pathogenic development in the pdeHΔ . We quantified the efficiency of appressorium formation on inductive or non-inductive surface in the pdeHΔ , pdeLΔ and the wild type . The pdeHΔ elaborated appressoria on inductive ( 90±2 . 1%; Figure 4A and 4B ) as well as on non-inductive surfaces ( 66±2 . 0%; Figure 4A and 4B ) , with a reasonably high frequency . In contrast , the wild type or the pdeLΔ conidia could form appressoria on inductive surface ( 75–80% ) , but failed to do so on non-inductive surface ( 4–7%; Figure 4A and 4B ) . Thus , the pdeLΔ behaves similar to the wild type and required proper inductive cues for pathogenic development . The two-celled conidia from the pdeHΔ pdeLΔ elaborated appressoria on inductive ( Figure S3C ) and non-inductive surfaces , however comparable quantifications were not possible due to the very low numbers of such structures formed . We thereby infer that the increased levels of intrinsic cAMP in the pdeHΔ are likely sufficient to uncouple surface dependency from pathogenic differentiation in M . oryzae . Interestingly , we observed that a significant number of pdeHΔ conidia formed germ tubes that were extremely short or barely visible , prior to appressorium initiation . We scored and systematically classified the lengths of the germ tubes into three categories: short ( short or barely visible ) , normal ( equal to the length of the conidium ) or long ( longer than the conidium ) prior to appressorium formation . We found that nearly 45±1 . 0% of the pdeHΔ conidia produced extremely short germ tubes prior to appressorium formation ( Figure 4A; arrows ) , versus 25±1 . 8% in pdeLΔ strain and 13±1 . 7% in the wild type . Nearly 49±1 . 7% of the pdeHΔ conidia formed germ tubes of normal length , versus 69±2 . 3% in pdeLΔ and 74±2 . 2% in the wild type . Long germ tubes were seen in 6±1 . 7% of pdeHΔ compared to 6±2 . 0% and 13±1 . 8% in pdeLΔ and wild type conidia respectively . Based on the above results , we construe that PdeH influences surface sensing and guides germ tube growth during the early stages of infection-related morphogenesis in M . oryzae . We observed that the loss of PdeH derails cAMP-associated surface signaling and significantly accelerated appressorium formation . Time-course analysis revealed that pdeHΔ is significantly advanced in all stages of pathogenic development . On inductive surfaces , the wild type conidia underwent germ tube hooking at 3–4 hpi ( Figure 5; white arrow ) , followed by tip swelling and growth into an immature appressoria by 5–6 hpi . By 8 hpi , the wild type formed melanized appressoria ( Figure 5; black arrow ) . Under identical conditions , the pdeHΔ conidia initiated germ tube hooking as early as 2 hpi ( Figure 5; white arrow ) , formed immature appressoria by 3–4 hours and finally melanized appressoria in about 5 hours ( Figure 5; black arrow ) . Thus , the pdeHΔ not only initiated appressoria earlier but also formed melanized appressoria more rapidly , at least 3 h in advance compared to the wild type . The pdeLΔ behaved similar to the wild type taking about 8 hours to form melanized appressoria . Our findings suggest that precocious elevated cAMP levels can disrupt the temporal regulation of the processes that lead to proper initiation , development and maturation of appressoria in M . oryzae . In M . oryzae , nuclear division or mitosis has been shown to precede appressorium formation , and arresting DNA replication or mitotic entry prevents appressorium formation [76] . Considering that a majority of pdeHΔ conidia precociously formed appressoria on inductive surfaces , we were interested in the mitotic status of the nuclei in pdeHΔ at different stages of pathogenic development . In wild type , mitosis generally occurred between 5–6 hpi , followed by migration of a daughter nucleus into the developing appressorium ( Figure S4A ) . Interestingly , appressorial morphogenesis progressed rapidly and independent of nuclear division up to 3 hpi in the pdeHΔ conidia ( Figure S4B; white arrows ) . By 4 hpi , mitosis generally occurred in the nucleus at the junction of the terminal cell of the conidium and the developing appressorium , followed by migration of a of the daughter nucleus into the maturing appressorium ( Figure S4B; asterisk ) . We thus infer that elevated cAMP levels likely result in appressorial morphogenesis prior to nuclear division in the germinating pdeHΔ conidia . In order to assess the ability to cause blast disease , we spray inoculated three-week old rice seedlings ( variety CO39 ) with conidia from pdeHΔ or pdeLΔ strain . The wild type served as a positive control , and disease symptoms were evaluated nine days post inoculation . Rice seedlings inoculated with the wild type or pdeLΔ conidia showed numerous typical spindle-like , gray centered lesions that merged into one another . On the other hand , pdeHΔ conidia failed to infect the host efficiently and to cause typical blast lesions . Instead , the pdeHΔ formed minute brown speckles ( Figure 6 ) that failed to coalesce and resembled the hypersensitive response ( HR ) seen on a moderately resistant host plant . Barley leaf explants inoculated with various dilutions of wild-type or pdeLΔ conidia developed typical blast symptoms comprising spindle-shaped lesions with gray centers ( Figure 6 ) . Under similar conditions , pdeHΔ conidia failed to cause comparable disease symptoms at all conidial dilutions tested ( Figure 6 ) . Next , we tested if the aberrant structures formed by pdeHΔ pdeLΔ during the conidiation phase could cause disease on barley explants . Equivalent numbers of conidia from the wild type were used as control in parallel . The double deletion mutant failed to cause any disease symptoms on the inoculated leaves , unlike the wild type that displayed typical disease lesions ( Figure S3B ) . The complemented pdeHΔ ( RFP-PdeH expressing strain ) was found to be as virulent as the wild type or the pdeLΔ on barley ( Figure S2C ) . Based on the disease assays on rice and barley , we conclude that PdeH ( and by inference cAMP levels ) is an important regulator of pathogenesis and disease severity during M . oryzae host interactions . Although , the pdeHΔ could efficiently form appressoria ( ∼90% ) on rice and barley leaves , it failed to establish proper blast disease . Hence , we investigated the in planta development and quantified the host penetration ability of each strain ( as judged by aniline blue staining for papillary callose deposits ) of pdeHΔ in comparison to the wild type and pdeLΔ on barley leaves . At 22 hpi , only 3±0 . 8% of the wild type appressoria formed penetration pegs compared to 10±1 . 2% in case of the pdeHΔ strain , indicating a significant advancement ( p<0 . 005 ) in the ability of the pdeHΔ appressoria to generate penetration pegs ( Figure 7A ) . At 24 hpi , 75±2 . 8% of the wild type and 73±3 . 8% of pdeHΔ appressoria formed penetration pegs . As is evident from the graphs in Figure 7A , 90±5 . 0% of the wild type penetration pegs proceeded to form infection hyphae at 36 hpi . In contrast , only 14±1 . 7% of the penetration pegs developed into infection hyphae in pdeHΔ . These observations suggest that the pdeHΔ is not defective in host penetration , but shows severe reduction in differentiating infection hyphae . Further observations ( 48 hpi ) revealed that 35±3 . 6% of pdeHΔ and 85±3 . 2% of the wild type penetration pegs advanced to form infection hyphae ( Figure 7A ) . Furthermore , the pdeHΔ infection hyphae were significantly reduced in their in planta growth and colonization , with a majority being restricted to the first invaded cell ( Figure 7B ) . In contrast , infection hyphae elaborated by the wild type achieved cross-wall penetration and spread . The pdeLΔ strain behaved in a manner similar to the wild type at all the time points tested ( Figure 7B ) . The aberrant structures formed by pdeHΔ pdeLΔ during the conidiation phase failed to germinate and form appressoria even after 72 hpi , and as a consequence did not elicit any response from the host ( Figure S3D ) . Interestingly , the two-celled conidia formed by the double deletion mutant successfully penetrated the host tissue as early as 22 hpi ( Figure S3E ) . Further observations at 36 , 48 and 72 hpi revealed that the double deletion mutant was blocked at the penetration stage , and failed completely to elaborate infection hyphae ( Figure S3D and 3E ) . These results indicate that the primary reason for the failure of the pdeHΔ to cause typical blast lesions is due to its compromised ability to form proper infection hyphae and to further advance its growth and spread in the host tissue . We thus hypothesize that PdeH-dependent regulation of cAMP signaling is key to successful host colonization by M . oryzae . In M . oryzae , cAMP signaling is known to regulate appressorium initiation and development [69] , [71] . Extensive work carried out on PDEs in mammalian cells and in Dictyostelium , has described the existence of feedback loops ( positive and negative ) between varying cAMP levels and PDE gene expression [77] , [78] . We were interested in elucidating if PDE transcripts were differentially regulated in response to fluctuating cAMP levels during pathogenic development in M . oryzae . Quantative Real-Time RTPCR analysis was used to measure the PDE transcript levels during different developmental stages during infection . The time points and tissues used were 0 h ( freshly harvested conidia ) , 3 h ( conidial germination and growth ) and 6 h ( appressorium initials ) . In addition , we also performed Real-Time RTPCR analysis on inoculated barley leaves at different stages of in planta growth . We included MGG_04795 ( BAS1 ) as a positive control , since it has been shown to be highly upregulated during infection [79] . Comparative analysis of the Real-Time RTPCR data showed that PDEH transcript levels were significantly down regulated during the early stages of pathogenic development at 3 h ( 2 . 5 fold ) and at 6 h ( 2 fold ) compared to freshly harvested wild-type conidia ( Figure 8A ) . A similar trend of reduction , 1 . 4 fold at 3 h and 2 . 5 fold at 6 h , was evident for the PDEL transcript . Compared to the 21 h time point , the PDEH transcript levels were consistently higher across all other time points tested: 1 . 5 fold at 24 h , 3 fold at 29 h and 5 fold at 48 h ( Figure 8B ) . However , the levels of PDEL transcript did not show significant changes at similar time points . This suggests that the PDEH transcript levels are differentially regulated during early as well as the late stages of pathogenic differentiation in M . oryzae , likely in response to the varying cAMP levels at these stages of development . We measured the levels of the PDEL transcript in pdeHΔ during pathogenic development . Conversely , the PDEH transcript levels were assessed in the pdeLΔ background . We used similar time points and fungal cell types as in the previous RTPCR experiment . PDEH transcript levels in the pdeLΔ were comparable to those in the wild type ( Figure 8C and previous experiment ) . At the early stage , the PDEH transcript was down regulated in both the wild type and pdeLΔ strains ( 2 fold at 3 h and 2 . 5 fold at 6 h ) and at the late stage , the PDEH transcript was upregulated: 2 fold at 24 h and 5 fold at 48 h ( Figure 8C ) . However , compared to its levels in the wild type , the PDEL transcript were notably abundant in the pdeHΔ , during the early as well as the late stages of pathogenic development , across all time points and conditions tested ( Figure 8D ) . We postulate that PDEH ( and by inference PdeHp function ) may be more responsive to even small changes in the levels of cAMP , whereas PDEL ( and likely PdeL activity ) likely responds to substantially elevated levels of cAMP , such as those prevalent in the pdeHΔ strain . Our results showed that the loss of PdeH function leads to higher basal levels of intracellular cAMP , thus affecting various aspects of asexual and pathogenic development . Therefore , we addressed whether exogenous addition of cAMP ( 8-Br-cAMP ) or IBMX to the wild type would mimic the pdeHΔ defects . 8-Br-cAMP ( a membrane permeable variant of cAMP ) or IBMX ( a phosphodiesterase inhibitor ) have been extensively used in various studies to artificially cause the enhancement of endogenous cAMP levels [69] , [71] , [80] , [81] . As shown in Figure 9A , wild type cultures grown in the presence of 8-Br-cAMP or IBMX showed a visible reduction in aerial hyphal growth , a defect reminiscent of the pdeHΔ strain . We further explored if addition of 8-Br-cAMP to the wild type would enhance conidiation therein . Wild type cultures were initially grown in the dark for 2 d in the presence of 8-Br-cAMP and later exposed to constant illumination for a period of 7 d prior to quantification of conidia . An untreated wild type and pdeHΔ served as controls . Compared to the untreated control , wild type treated with 10 mM 8-Br-cAMP showed a marginal increase in conidial numbers; however 50 mM 8-Br-cAMP treatment evoked a 1 . 6 fold ( p≤0 . 01 ) increase in conidia production ( Figure 9B ) . This increase in conidiation although considerable , was not as high ( 2 . 3 fold; p<0 . 005 ) as exhibited by the pdeHΔ . Furthermore , barley explants inoculated with wild-type conidia in the presence of 8-Br-cAMP showed a dose-dependent reduction in lesion size and lacked disease severity when compared to the control inoculations ( Figure 9C ) . In contrast to the reduced lesion size and disease progression caused by 10 mM 8-Br-cAMP , the wild type conidia treated with 2 . 5 mM IBMX showed absolutely no disease symptoms on barley explants ( Figure 9C ) . About 75±4 . 0% of 8-Br-cAMP-treated wild type appressoria successfully formed penetration pegs at 24 hpi; however by 36 hpi only 40±3 . 4% of the penetration pegs further developed into infection hyphae . By 48 hpi the number of infection hyphae formed increased to 47±3 . 0% ( Figure 9D and 9F ) . At the corresponding time point , 90±3 . 6% penetration pegs formed by the control untreated sample had developed into infection hyphae ( Figure 9E and 9F ) . IBMX-treated wild-type conidia formed appressoria efficiently on leaf tissues , but failed to elaborate penetration pegs even after 36 hpi . At 72 hpi , unlike the 8-Br-cAMP treated wild type , which elaborated infection hyphae , the IBMX-treated wild type failed to develop infection hyphae ( Figure 9F ) . Thus , 8-Br-cAMP- or IBMX-treated wild type significantly mimicked the pdeHΔ defects in aerial growth , conidiation , disease development and severity . These results support our previous experimental observations and our hypothesis that the defects exhibited by the pdeHΔ strain are indeed due to enhanced basal levels of cAMP , caused due to the loss of PdeH function in M . oryzae . We expressed an RFP-PDEH translational fusion construct in the pdeHΔ strain to track the subcellular localization and understand the spatial and temporal regulation of PdeH . This strategy was aimed at fluorescently tagging PdeH and to serve as a complementation tool that could potentially suppress the defects exhibited by pdeHΔ strain ( Figure S2A and S2B ) . Quantifications revealed that the number of conidiophores and conidia produced by the RFP-PDEH expressing pdeHΔ strain was comparable to the wild type ( Comp pdeHΔ , Figure 2B ) . The RFP-PDEH strain lost the ability to elaborate appressoria on non-inductive surfaces . Infection assays showed that , unlike the pdeHΔ , the RFP-PDEH expressing strain had regained the ability to efficiently infect barley leaf explants at various spore concentrations tested ( Figure S2C ) . These results indicate that the RFP-PdeH fusion protein was indeed functional and able to significantly suppress the pdeHΔ defects in conidiation and pathogenesis . The expression and localization of RFP-PdeH was then examined at different stages of asexual and pathogenic development . Vegetative hyphae and developing conidiophores showed a predominantly weak cytosolic distribution of RFP-PdeH ( Figure 10A and 10B ) . Next , we looked at the distribution of RFP-PdeH at different stages of pathogenic development ( Figure 10C ) . Conidia were harvested from the RFP-PDEH strain , inoculated on plastic cover slips and observed using epifluorescence microscopy . In freshly harvested conidia ( 0 h ) , RFP-PdeH localized as cytosolic punctae mostly in the terminal cell of the conidium . At 2 hpi , RFP-PdeH foci were predominant in the terminal cell as well as in the developing germ tube ( Figure 10C; arrow ) . After 4 hpi , RFP-PdeH sustained its distinct punctate localization throughout the terminal cell of the conidium and the germ tube . Furthermore , there was a notable signal although weak , from the rim of the hooking germ tube , indicating a probable association with the plasma membrane ( Figure 10C; arrow ) . The RFP-PdeH signal was weak and indiscernible in the conidia at 6 hpi and 8 hpi , but localized as randomly distributed punctae throughout the developing appressorium , and also showed a possibly weak association with the appressorial membrane . We excluded the possibility that the membrane localization was an artifact of melanization , since tricyclazole treated RFP-PdeH appressoria retained the weak association with the plasma membrane in addition to the distinct cytosolic punctae . ( Figure S5A; arrows ) . In mature melanized appressoria ( 21 hpi; Figure 10C ) , the RFP-PdeH displayed a predominantly vacuolar localization . The RFP-PdeH was uniformly distributed through out the cytosol in the infection hyphae within the rice leaf sheath at 36 hpi ( Figure 10D ) . Thus , RFP-PdeH displays a predominantly cytosolic distribution during asexual differentiation , whereas it localizes as distinct cytosolic foci , ( and weakly to the plasma membrane of the germ tubes ) during pathogenic development . RFP-PdeH was uniformly distributed throughout the cytosol in the infection hyphae during the biotrophic phase . In order to better visualize the intracellular distribution and dynamics of PdeH , we expressed a GFP-PDEH translational fusion construct driven by the MPG1 promoter [82] in the pdeHΔ strain . Similarly , we generated a strain expressing GFP-PdeL fusion protein under the MPG1 promoter . We examined the expression and localization patterns of the PROMpg1-GFP-PdeH fusion protein during different stages of asexual and pathogenic development . Compared to the weak RFP-PdeH signal , the PROMpg1-GFP-PdeH showed a relatively strong cytoplasmic signal in the vegetative mycelia ( Figure 11A ) . During asexual development , the GFP-PdeH fusion protein was predominantly cytosolic ( Figure 11B ) , a pattern comparable to that displayed by the RFP-PdeH strain at a similar stage of development . To gain further insight into the dynamics , we made time-lapse observations of the PROMpg1-GFP-PdeH at different stages of pathogenic development encompassing the following time points: 2–3 hpi ( conidial germination; Video S1 and Figure 11C ) , 4–5 hpi ( hooking stage; Video S2 and Figure 11D ) , and 6–7 hpi ( appressorium development; Video S3 and Figure 11E ) . Time-lapse analysis revealed that the GFP-PdeH fusion protein was associated with vesicular structures , which were highly dynamic and mobile ( Video S1 and Figure 11C ) . Furthermore , co-staining with a nuclear dye ( Hoechst 33342 ) confirmed a peri- and extra- nuclear localization of the GFP-PdeH foci ( Figure S5B ) . At 4–5 hpi , GFP-PdeH localized to regions of the plasma membrane of the hooking germ tube in addition to being associated with highly mobile vesicles shuttling between the conidium and the germ tube ( Video S2 and Figure 11D ) . In addition , GFP-PdeH was vesicular and enriched at the plasma membrane of the appressorium at 6 hpi ( Video S3 and Figure 11E ) . Next , we examined the subcellular distribution of the PROMpg1-GFP-PdeL fusion protein in M . oryzae . Rather surprisingly , at all the stages of asexual and pathogenic development the GFP-PdeL localized predominantly to the nucleus ( Figure 12A to 12F ) . The nuclear destination of GFP-PdeL was confirmed by co-staining with DAPI ( Figure 12D and 12E ) . Thus , the predominant compartmentalization of PdeH within the cytosol ( including the vesicular and plasma membrane localization ) and of PdeL in the nucleus suggests that the high- and low-affinity PDEs differentially regulate and likely modulate distinct intracellular pools of cAMP . In summary , our results strongly suggest that appropriate and timely modulation of cAMP signaling , mainly by the PdeH , is critical for regulating the disease causing ability , as well as several important aspects of asexual development in M . oryzae .
Upon perception of light , the aerial hyphae of the rice-blast fungus M . oryzae , enter the asexual differentiation pathway to form conidiophores , which ultimately produce 3–4 conidia , each , in a sympodial manner [65] , [67] , [83] , [84] . The pattern and number of conidia produced per conidiophore is highly regulated and important for proper conidial function [67] , [71] , [84] . Heterotrimeric G proteins and cAMP signaling have been shown to be important regulators of asexual development in M . oryzae [71] , [85] , [86] and several other filamentous fungi such as Aspergillus , Cryphonectria , Neurospora [44] , [87] , [88] , [89] , [90] , [91] . The cAMP signaling is also a key regulator of the pathogenic differentiation in M . oryzae [71] , [81] , [85] , [92] , [93] . The cAMP PDEs constitute a large family of enzymes that hydrolyse cAMP and provide the sole means of inactivating this important second messenger in eukaryotic cells . In addition , PDEs function to regulate the specificity , intensity and temporal duration of cAMP signaling [2] , [4] , [94] , [95] , [96] , [97] , [98] . While a functional role for the high- and low-affinity PDEs in filamentous fungi such as Neurospora or Aspergillus has not yet been defined , our preliminary analysis does indicate that these fungi do possess PDE orthologs ( Figure 1A and 1B ) with the conserved consensus sequences typical of both the variants . However , specific roles have been defined for cAMP signaling in C . albicans and C . neoformans , representing the dimorphic or pseudo-filamentous fungi . In addition to regulating the development of hyphal cell wall and membrane structures , the high-affinity PDE in C . albicans has been shown to regulate intracellular levels of cAMP during stress conditions and virulence . However , a role for the low-affinity PDE has not yet been established [23] , [24] , [25] , [27] , [61] , [99] . Rather surprisingly , a reversal in function has been suggested for the PDEs in C . neoformans , wherein the low-affinity PDE modulates the cAMP signaling [30] . In the present study , we investigated the implications of PDE-based regulation of the intracellular cAMP during asexual and pathogenic development in M . oryzae . We analyzed in detail the effects of gene-deletion of the two cAMP phosphodiesterase-encoding genes PDEH and PDEL ( either independently or in combination ) . Our data suggests unique and non-redundant roles/functions for PdeH and PdeL; but underscores the overall importance of PdeH as a critical regulator of cAMP signaling in M . oryzae . Furthermore , PdeH and PdeL appear to be differentially compartmentalized: PdeL being predominantly nuclear while PdeH is largely cytosolic but confined to perinuclear regions as vesicular foci during conidial germination and to the developing appressorial membrane during the later stages of infection-related morphogenesis . Such an association of a PDE to membranous regions around the nucleus , has been demonstrated through fractionation studies in yeast [100] . Unlike PdeL , which we found to be dispensable for asexual development and virulence in M . oryzae , PdeH was a key regulator of cAMP levels during these two important phases of growth . This suggests that PdeH is able to compensate for the loss of PdeL . However , simultaneous loss of both PDEs was deleterious to proper growth and asexual development , suggesting that above a particular threshold , excess cAMP is detrimental . Our results further suggest that PdeH regulates cAMP signaling at two critical steps during M . oryzae pathogenesis namely: infection-structure formation and host invasion . This hypothesis is based on the direct assessment of cAMP levels , at the initial ( increased ) and final ( decreased ) stages of infection in the wild type and the PDE mutant strains of M . oryzae . An initial increase in cAMP [69] , [71] is a likely consequence of a concomitant decrease in PDEH transcript levels during the early phase of appressorium formation . Such a decrease in PDE function could also be a consequence of post-translational modifications ( as demonstrated in yeast and various mammalian isoforms [2] , [12] , [30] , [54] , [94] ) and/or the dynamic recycling of the PdeH-containing multivesicular foci away from the active growth or signaling zones in the germ tube tips . Nonetheless , the down-regulation of cAMP at appressorium maturity is extremely important since it has a strong bearing on successful colonization of host tissue by M . oryzae . We demonstrate that higher levels of exogenous cAMP retard in planta development in the wild type at the late stages of infection , a phenotype reminiscent of the loss of PdeH function during rice-blast disease . On the contrary , increased intracellular cAMP levels ( pdeHΔ or exogenous addition ) enhance asexual development in M . oryzae . However , there appears to be a threshold and dose-dependent response to high cAMP , since loss of both PDEs ( simultaneous ) led to aberrant fungal structures and a complete cessation of proper conidiation . The genetic data and biochemical analysis presented here support the model that PdeH ( like Rgs1 , Regulator of G protein Signaling [71] ) , negatively affects the cAMP signaling in M . oryzae . We show that overall cAMP levels are not significantly affected in the pdeLΔ either during conidia or appressoria formation , and that PdeH is the major hydrolyzing enzyme to control the baseline levels of cAMP in these important cell types . However , taking into consideration the phenotype and the elevated cAMP levels in the double deletion mutant , it is possible that PdeL gains importance under conditions when the intracellular levels of cAMP are very high , like in the pdeHΔ strain . We propose that PdeH may not be directly involved in cell shape and morphogenesis , but that the pdeHΔ phenotype is brought about by the inappropriate or untimely activation of signaling through cAMP accumulation . For instance , majority of the pdeHΔ appressoria were initiated without any discernable germ tube emergence or extension from the terminal cells of the conidia . It is tempting to speculate that cAMP signaling is involved in cessation of germ tube growth prior to initiating the hooking stage important for appressorium initiation . Our preliminary data ( Figure S4 ) also suggests that increased cAMP levels likely override the requirement for nuclear division prior to initiation of appressoria . However , additional experiments are required to further investigate the relevance of the pdeHΔ defects in understanding infection-related morphogenesis in conjunction with cell cycle progression in the rice-blast fungus . It is plausible that PdeL activity likely responds to enriched gradients or pools of cAMP present in the nucleus , as has been suggested by Huston . E et al [101] . Furthermore , the principles of compartmentalized cAMP signaling may well apply to highly polarized cell types in M . oryzae , namely conidia , appressoria and infection hyphae , wherein relevant changes in cAMP levels may be controlled in space and time and could be anchored within subcellular compartments such as the nucleus or the outer membranes . It is also possible that the M . oryzae adenylate cyclase ( mac1 ) [72] , does not produce cAMP constitutively but only in small bursts confined to the active growth and signaling zones , in response to inductive stimuli . Our data indicates that cell signaling in M . oryzae responds to rapid albeit small changes in cAMP levels , since loss of PdeH leads to increased accumulation of intracellular cAMP . Furthermore , our findings clearly establish that the overall modulation of the non-nuclear cytosolic pools of cAMP by the high-affinity phosphodiesterase at two distinct phases ( up-regulation during early stages and down-regulation during the late stages ) is vital for proper pathogenic and infectious development in M . oryzae . We do not rule out dynamic fluctuations ( albeit minor ) of cAMP within these two major stages of pathogenic differentiation . Furthermore , it is possible that the reduced in planta growth of the pdeHΔ in M . oryzae is a likely consequence of the increased sensitivity towards stress conditions encountered within the host plant . For instance , the cAMP pathway is a major regulator of stress signaling in Candida and the pde2Δ mutant is more sensitive to stress , particularly peroxide and cadmium [99] . Future experiments would address the issues related to associations ( if any ) between stress signaling and cAMP levels in M . oryzae .
The wild type M . oryzae strain , B157 , was obtained from the Directorate of Rice Research ( Hyderabad , India ) . M . oryzae strains were cultured either on Prune agar medium at 28°C ( PA; per liter: 40 mL prune juice , 2 . 5 g lactose , 2 . 5 g sucrose , 1 g yeast extract , and 20 g agar ) or Complete medium ( CM; per liter: 0 . 6% yeast extract , 0 . 6% casein hydrolysate , and 1% sucrose ) . Either CM agar ( for hygromycin selection ) or Basal Medium ( BM; per liter: Yeast nitrogen base without amino acids 1 . 6 g , asparagine 2 g , glucose 10 g , NH4 NO3 1 g , and 20 g of agar , for ammonium glufosinate selection ) was used for the selection of fungal transformants . To assess the growth and colony characteristics , wild type as well as the deletion strains were cultivated on PA medium at 28°C for one week . For quantitative analysis of conidiation , fungal strains were cultivated on PA medium in the dark for a day , followed by incubation under constant illumination for 7 d at room temperature . Mycelia used for genomic DNA or total RNA extraction were harvested from cultures grown in liquid for 2–3 days at 28°C as described [102] . Conidia were harvested by scraping the surface of the colonies with inoculation loops in the presence of sterile water , and the fungal biomass collected in Falcon tubes ( BD Biosciences , USA ) . The suspension was vortexed for a minute to ensure complete detachment of conidia from the mycelia , and then filtered through two layers of Miracloth ( Calbiochem , San Diego , USA ) . The conidia were pelleted by centrifugation at 3000 rpm for 15 minutes at room temperature . The conidial pellet thus obtained was washed twice and re-suspended in sterile water . The radius of the colony was initially measured to calculate the surface area of the colony . Conidia produced by a given colony were quantified using a hemocytometer and reported as the total number of conidia present per unit area of the colony . For appressorial assays , the harvested conidia were re-suspended at 105 conidia per mL in sterile water . Droplets ( 20 µl ) of conidial suspension were placed on plastic cover slips or hydrophilic side of GelBond membrane ( Lonza Walkersville Inc . , USA ) and incubated under humid conditions at room temperature . The total number of appressoria was quantified after 16 h . Microscopic observations were made using an Olympus BX51 epifluorescence compound microscope with bright field optics . For pathogenicity assays and assessment of blast lesions , a dilution series of the conidial suspension was inoculated on detached barley leaves , and incubated for 7–9 days in a growth chamber ( 22°C , 16 h light/8 h dark ) . Spray inoculations on rice cultivars were conducted as previously described [103] . For host penetration assays , conidial suspensions in sterile water were inoculated on barley leaf explants and assessed after 22 h , 24 h , 36 h and 48 h . Penetration pegs and infection hyphae were detected by staining for papillary callose deposits using Aniline blue[104] . Fungal structures were stained with acid fuchsin as described [74] . Given their lack of sufficient numbers , we could not carry out spray inoculations of the rice seedlings with the aberrant structures formed by the double deletion mutant ( during the conidiation phase ) . However barley infection assays were carried out with pdeHΔ pdeLΔ with a conidial load normalized to at least 50 two-celled conidia-like structures per droplet . A parallel wild type control with equivalent conidial load was included . The cyclic AMP analog , 8-Br-cAMP ( BioLog , Germany ) was first added at 0 h and again supplemented after 20 hpi ( a final concentration of 10 mM or 50 mM ) . Stock solutions of 8-Br-cAMP ( 100 mM ) were made in water , while IBMX ( 25 mM ) ( Sigma Aldrich , USA ) was dissolved in 99% ethanol [71] , [80] . For tricyclazole ( Cluzeau Info Labo , France ) treatment , conidia from the requisite strain were inoculated on cover slips in the presence of 8 µg/mL tricyclazole [105] . Nuclear staining was carried out using DAPI ( diamidino-2-phenylindole; Sigma Aldrich , USA ) essentially as described [106] , although with minor modifications to suit M . oryzae conidia or germlings . Freshly harvested conidia were appropriately diluted and inoculated on hydrophobic plastic cover slips in a moist chamber at room temperature ( RT ) for 1–2 hours . The cover slips with the adherent germlings were then fixed with formaldehyde ( 3 . 7% final concentration ) , for 5–10 minutes at RT . The fixed samples were then washed gently with distilled water , prior to treatment with Triton X-100 ( 0 . 1% final concentration ) for 1 minute at RT . The samples were washed thrice with distilled water , and finally incubated in the dark for 10–15 minutes , with a solution of DAPI dissolved in water at a final concentration of 1 µg/ml , and visualized using the Olympus BX51 epifluorescence microscope . Nuclear staining during live imaging of GFP-PdeH was achieved using Hoechst 33342 ( Sigma Aldrich , USA ) at a concentration of 1 µg/ml ( in water ) for 20 minutes . Cell wall and septa were stained using calcofluor white in water at a final concentration of 10 µg/ml . Samples were washed twice with sterile distilled water prior to visualization by epifluorescence imaging using the Olympus IX71 microscope . Standard molecular manipulations were performed as described [107] . Fungal genomic DNA was extracted using Master Pure Yeast DNA purification kit ( Epicenter Biotechnologies ) . Plasmid DNA was isolated with Geneaid High Speed Plasmid Mini kits . Homology searches of DNA/protein sequences were performed using the BLAST program [108] and multiple sequence alignments carried out with ClustalW [109] and Boxshade ( http://bioweb . pasteur . fr/seqanal/interfaces/boxshade . html ) . Deletion mutants of PDEH ( NCBI accession XP_360290 ) or PDEL ( NCBI accession XP_367803 ) were generated using the standard one-step gene replacement strategy . Genomic DNA fragments ( about 1 kb each ) representing the 5′ and 3′ UTR of PDEH ( MGG_05665 ) gene were amplified by PCR , ligated sequentially to flank the hygromycin phosphotransferase gene ( HPH1 ) cassette , in pFGL44 , to obtain pFGLpdeHKO . The following primers were used to amplify the 5′and 3′ UTR of the PDEH gene: PDEH-5F ( 5′- CAGAGAGAATTCAGCACCAGCATGGCACCACTATC ) , PDEH-5R ( 5′- CAGAGATCTAGACAAAGAGCGTCCAGTCATAAGACT ) , PDEH-3F ( 5′- CAGAGACTGCAGGTTCAGTACTACTGTTCACTCAGAT ) , PDEH-3R ( 5′- CAGAGAAAGCTTACGCATTACCCAATGTTGGCATC ) . pFGLpdeLKO was obtained by amplifying approximately one kb fragments of genomic DNA corresponding to the 5′ and 3′ UTR of PDEL ( MGG_07707 ) . The PCR fragments obtained were cloned in pFGL97 to flank the bialaphos resistance gene cassette ( BAR ) . The 5′ UTR was amplified using the following primer pairs: PDEL-5F ( 5′CAGAGAGGTACCCCGTTTGCTACCTGTGGCCAACG ) , PDEL-5R ( 5′- CAGAGAGGATCCCCGCCCGTCCCGTCTAGCCCAGCTGGGCT ) ; The 3′ UTR was amplified using the following primer pairs PDEL-3F ( 5′- CAGAGACTGCAGTGCGACCCTATGACAGTCCCCT ) , PDEL-3R ( 5′- CAGAGAAAGCTTGAGGCCGCCAATGCCACGAGCGC ) . Underlined text in the primer sequences represents the restriction enzyme sites used for cloning purposes . The sequences of pFGLpdeHKO as well as the pFGLpdeLKO gene replacement constructs was confirmed and the plasmids were introduced into wild type B157 strain via Agrobacterium T-DNA-mediated transformation to specifically replace the PDEH or PDEL genes with HPH1 or BAR respectively . Resistance to hygromycin ( CM containing 250 µg/ml hygromycin , A . G . Scientific Inc , USA ) or ammonium glufosinate ( BM containing 40 µg/ml ammonium glufosinate , Cluzeau Info Labo , France ) was used to select the fungal transformants . Southern blot analysis and PCR were performed to identify the correct gene-replacement events ( pdeH::HPH1 or pdeL::BAR ) . The double mutant pdeHΔ pdeLΔ was like wise generated by introducing pFGLpdeHKO into the pdeLΔ strain . Southern blot analysis was used to confirm the gene replacement ( pdeH::HPH1 ) and copy number of the integron . To generate an in-frame translational fusion of RFP-PdeH , the promoter fragment ( about 1 kb ) of the PDEH gene was PCR amplified from genomic DNA of the wild type , using the primers ( 5′- CAGAGAGAATTCGTTCGGCTCAATTCAATTCGA ) and ( 5′- CAGAGAGAGCTCCGTGGGCCCAAAGAGCGTCCA ) . The RFP coding sequence was amplified from pDsRED-Monomeric-N1 ( Clontech , CA , USA ) using the primers ( 5′- CAGAGAGCGCTCATGGACAACACCGAGGAC ) and ( 5′- CAGAGACCATGGCCTGGGAGCCGGAGTG ) . A 3 . 9 kb fragment comprising the entire PDEH coding sequence as well as the downstream 1 kb region was amplified using the primers ( 5′- CAGAGACCATGGAGAATGCTGCCTGCAAT ) and ( 5′-AGAGAGGATCCTGCCAAGTTGTCACTTTCCAAGT ) . Underlined text in the primers corresponds to the restriction enzyme sites used for cloning . All the three PCR products obtained were cloned into pFGL97 to get pFGL-NT-Comp construct with resistance to ammonium-gluphosinate ( Cluzeau Info Labo , France ) as a fungal selectable marker . This construct was introduced as a single-copy insertion into the pdeHΔ strain . eGFP was amplified using the following primers ( 5′-CAGACCATGGTGAGCAAGGGCGAGGA ) and ( 5′- CAGACATATGCTTGTACAGCTCGTCCAT ) from pEGFP-N1 ( Clontech , CA , USA ) . A ∼3 . 0 kb fragment encoding the complete PDEH coding sequence was amplified using the primers ( 5′- CAGACATATGGAGAATGCTGCCTGCAAT ) and ( 5′- CAGATCTAGATCAACCAGCAGTGTC ) . Similarly , a ∼2 . 6 kb fragment coding for the entire PDEL sequence was amplified using the following primer pairs ( 5′- CAGACATATGGGCGAGGGCAGCGCCGAA ) and ( 5′- CAGATCTAGATCACAAGTACAATGCCTCGCCA ) . Underlined text denotes the restriction enzyme sites used for cloning . In both the cases the amplified PCR products ( eGFP and PDEH ) or ( eGFP and PDEL ) were cloned in-frame , under the constitutive MPG1 promoter and TrpC terminator in pFGL275 to obtain pFGL-PROMpg1-GFP-PDEH and pFGL- PROMpg1-GFP-PDEL constructs respectively . pFGL-PROMpg1-GFP-PDEH was introduced into the pdeHΔ strain , while the pFGL-PROMpg1-GFP-PDEL was introduced into the wild type strain via Agrobacterium T-DNA-mediated transformation . Fungal transformants was selected based on resistance towards ammonium glufosinate ( BM containing 40 µg/ml ammonium glufosinate , Cluzeau Info Labo , France ) . Transformants were screened for GFP expression and confirmed by sequencing genomic DNA and southern blot analysis . Total RNA was extracted using RNeasy Plant Mini Kit ( Qiagen , USA ) from ungerminated wild type conidia ( 0 h ) , or WT conidia germinated for 3 h and 6 h on hydrophobic surface of GelBond membranes ( Lonza Walkersville Inc . , USA ) . Total RNA was also extracted from wild type conidia inoculated on detached barley leaves at 21 h , 24 h , 29 h and 48 h . Purified RNA was treated with DNase ( Roche Diagnostics , Germany ) and were verified as DNA free by using them directly as template in a PCR assay . First-strand cDNA was then synthesized using 2 µg total RNA and AMV Reverse Transcriptase ( Roche Diagnostics , Germany ) in the presence of oligo dT18 . Reactions were performed in 10 µl volume containing 25 ng of cDNA , 0 . 5 µM of each primer and 5 µl of Power SYBR Green PCR Master Mix ( Applied Biosystems ) using 7900HT Fast Real-Time PCR system . The following cycling parameters were used , 50°C 2 minute , 1 cycle; 95°C 10 minute , 1 cycle; 95°C 15 second , 60°C 1 minute , 40 cycles . Relative abundance of transcripts was analyzed by the 2−ΔΔCt method [110] and average threshold cycle ( Ct ) normalized to β-tubulin transcript for each condition as 2−ΔCt . Fold changes were calculated as 2−ΔΔCt . The primer pairs used for quantitative RTPCR were: For MGG_04795 ( 5′- TTTGATCAGCGTTACCAAGG ) and ( 5′- CGGTGACCAACATTCTCTTG ) ; MGG_00604 ( 5′- CATGATGGCTGCTTCTGACT ) and ( 5′-CGACGAGTTCTTGTTCTGGA ) MGG_ 05664 ( 5′- GCTTGAGCGCTGGAGAATGT ) and ( 5′- TAACGAGCCGATCTGTACCA ) ; MGG_07707 ( 5′- CTTCTACAGCAGAGACACC ) and ( 5′- GCTCCTGCATAATAATGTCC ) . Each Real-Time RTPCR reaction was repeated three times independently with three biological replicates per sample . The melting curve analysis was used to determine the specificity of the amplifications . Reactions with no cDNA added ( no template controls ) were performed in parallel and monitored for primer dimers . Real-Time RTPCR primers were designed to span introns whenever possible . Quantification of intracellular levels of cAMP was essentially carried out as described [71] . Freshly harvested conidia ( 0 h ) , conidia germinated for 6 h , as well as conidia that have formed mature appressoria following 21 h , 24 h , 28 h or 30 h incubation on the inductive ( hydrophobic ) surface of GelBond membranes ( Lonza Walkersville Inc . , USA ) were harvested , frozen and lyophilized for 16 h . Lyophilized fungal biomass was individually ground into a fine powder in liquid nitrogen and equal weights of the fine powder was re-suspended in 200 µl of chilled 6% Trichloro acetic acid ( TCA ) and incubated on ice for 10 min . After centrifugation at 4000 rpm for 15 min at 4°C , the supernatant was collected and extracted four times with five volumes of water-saturated diethyl ether to remove the TCA . The remaining aqueous extract was lyophilized and dissolved in the assay buffer . In total , each assay was repeated three times independently with two biological replicates for each strain . The cAMP levels were quantified using the cAMP Biotrak Immuno-assay System ( Amersham Biosciences , USA ) . For estimation of cAMP levels , the pdeHΔ pdeLΔ mutant was grown on nitrocellulose membranes ( Millipore , USA ) placed on prune agar ( PA ) medium . These cultures were incubated at 28°C for 3 d , followed by exposure to light for 24 h . The resultant colonies were scraped gently in the presence of diluted assay buffer to harvest the mycelial and aerial growth . Care was taken not to damage or include the nitrocellulose membrane in the fungal biomass . Further processing of the sample for cAMP measurements , was essentially carried out as described above . The double deletion mutant did not display any difference in growth pattern or characteristics on the nitrocellulose membrane , compared to growth without such membranes under the same conditions . Bright field and epifluorescence microscopy utilized the Olympus IX71 or BX51 ( Olympus , Japan ) equipped with a Plan APO 100X/1 . 45 or UPlan FLN 60X/1 . 25 objective with appropriate filter sets . Images were captured using a Cool SNAP HQ camera ( Photometrics , USA ) and processed using Image J ( National Institutes of Health , USA ) , MetaVue ( Universal Imaging , USA ) and Adobe Photoshop 7 . 0 ( Adobe Inc , USA ) .
|
Magnaporthe oryzae , an economically important fungal pathogen that causes the blast disease in rice and several cereal crops , is a model organism for studying fungus-host interactions . M . oryzae reproduces asexually by producing spores , which can switch to an infectious disease-causing mode of development in response to host and environmental cues . Subsequent activation of conserved signaling modules involving the second messenger cyclic AMP has been shown to be important for pathogenic development . In this study , we used a variety of genetic and biochemical approaches to functionally characterize two enzymes ( encoded by PDEH and PDEL genes in M . oryzae ) which regulate the intracellular levels of cAMP signaling through hydrolysis . Furthermore , we show that the timely modulation and maintenance of appropriate levels of cAMP is essential for proper asexual development , and more importantly for the ability of M . oryzae to successfully infect and cause disease . Our results suggest that cyclic AMP signaling is broadly compartmentalized ( nuclear and cytosolic ) and that M . oryzae is fully capable of inducing and responding rapidly to short pulses of cAMP in the various cell types related to the above developmental events .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/microbial",
"growth",
"and",
"development",
"genetics",
"and",
"genomics/gene",
"function",
"microbiology/plant-biotic",
"interactions",
"cell",
"biology/cell",
"signaling"
] |
2010
|
PdeH, a High-Affinity cAMP Phosphodiesterase, Is a Key Regulator of Asexual and Pathogenic Differentiation in Magnaporthe oryzae
|
Mechanisms generating diverse cell types from multipotent progenitors are fundamental for normal development . Pigment cells are derived from multipotent neural crest cells and their diversity in teleosts provides an excellent model for studying mechanisms controlling fate specification of distinct cell types . Zebrafish have three types of pigment cells ( melanocytes , iridophores and xanthophores ) while medaka have four ( three shared with zebrafish , plus leucophores ) , raising questions about how conserved mechanisms of fate specification of each pigment cell type are in these fish . We have previously shown that the Sry-related transcription factor Sox10 is crucial for fate specification of pigment cells in zebrafish , and that Sox5 promotes xanthophores and represses leucophores in a shared xanthophore/leucophore progenitor in medaka . Employing TILLING , TALEN and CRISPR/Cas9 technologies , we generated medaka and zebrafish sox5 and sox10 mutants and conducted comparative analyses of their compound mutant phenotypes . We show that specification of all pigment cells , except leucophores , is dependent on Sox10 . Loss of Sox5 in Sox10-defective fish partially rescued the formation of all pigment cells in zebrafish , and melanocytes and iridophores in medaka , suggesting that Sox5 represses Sox10-dependent formation of these pigment cells , similar to their interaction in mammalian melanocyte specification . In contrast , in medaka , loss of Sox10 acts cooperatively with Sox5 , enhancing both xanthophore reduction and leucophore increase in sox5 mutants . Misexpression of Sox5 in the xanthophore/leucophore progenitors increased xanthophores and reduced leucophores in medaka . Thus , the mode of Sox5 function in xanthophore specification differs between medaka ( promoting ) and zebrafish ( repressing ) , which is also the case in adult fish . Our findings reveal surprising diversity in even the mode of the interactions between Sox5 and Sox10 governing specification of pigment cell types in medaka and zebrafish , and suggest that this is related to the evolution of a fourth pigment cell type .
Cell fate specification from multipotent stem cells is a fundamental process during embryonic development . The mechanisms determining how multipotent progenitors choose a diversity of different cell fates is key to understanding the generation of multiple cell types in the animal body . The neural crest is a population of migratory multipotent cells that form at the boundary of neural plate and non-neural ectoderm , and gives rise to a variety of cell types including pigment cells [1 , 2] . Whereas mammals and birds have a single type of pigment cells , the black or brown melanocyte , fish have multiple types of pigment cells , known as chromatophores [3 , 4] . In zebrafish there are three types–melanophores ( often referred to as melanocytes , the term we shall use throughout this manuscript hereafter ) , iridescent iridophores , and yellow/orange xanthophores [5] . Medaka has four chromatophore types , with all those from zebrafish plus reflective white ( or orange during embryonic/larval stages ) leucophores [6–8] , raising the question of whether fate specification mechanisms remain conserved through evolution . Thus , the multitude of pigment cell types within fish makes pigment cells an excellent ‘model within a model’ system for studying fate specification and its evolution , especially since they are both easily identified by their natural colors and other markers , and not essential for survival [4] . The similarity of embryonic/larval pigment pattern in medaka and zebrafish , species that last shared a common ancestor some 324 million years ago [9] , makes these two fish models highly complementary and valuable for comparative studies of neural crest . Melanocytes form three longitudinal stripes , known as the dorsal stripe ( DS ) , lateral stripe ( LS ) and ventral stripe ( VS ) in both zebrafish and medaka , and a fourth stripe on the ventral yolk sac , called the yolk sac stripe in zebrafish or yolk sac cluster ( YSC ) in medaka; xanthophores lie under the epidermis between the stripes in both species; iridophores locate on the dorsal yolk sac in clusters , called lateral patches ( LP ) , as well as on the eyes in both species . Iridophores are distributed along the dorsal midline in zebrafish , being surrounded by melanocytes , whereas leucophores occupy this position in medaka [5 , 8] . Whereas melanocytes , iridophores and xanthophores are widely distributed in teleosts [4] , leucophores have a much more restricted distribution and hence are likely an evolutionary innovation in the medaka lineage [10–13] . It has been proposed that all types of pigment cells are generated from common pigment cell progenitors ( chromatoblasts ) [14] via progressive fate-restriction processes [15 , 16] . Leucophores are present in medaka but not in zebrafish , implying an additional choice between alternative fates in medaka . Our previous studies showed a close developmental and genetic relationship between leucophores and xanthophores [17] . This led us to propose that both leucophore and xanthophore specification from the neural crest is Pax7a-dependent and that they may develop from a shared bipotent progenitor , perhaps distinguishable by its expression of pax7a . The Sry-related high-mobility group ( HMG ) transcription factor Sox5 is widely expressed in migrating neural crest cells , but functions cell-autonomously in the pax7a-expressing shared progenitors to promote a xanthophore fate choice over leucophore [18] . Genetic analyses have identified other key transcription factors involved in pigment cell specification . The best studied of these is Mitf ( microphthalmia-associated transcription factor; Mitfa in zebrafish ) , which is considered the master regulator of melanocyte development in zebrafish and mammals [19–21] . Mitf controls expression of genes required for melanocyte development , including dopachrome tautomerase ( dct ) , tyrosinase and the receptor tyrosine kinase c-kit [22] . Zebrafish mitfa mutants show reduced or no melanocytes , but an increased number of iridophores [19] . Expression of Mitf/mitfa is regulated by Sox10 , which is widely expressed in neural crest cells prior to their differentiation [23–28] . As loss of Sox10 function results in complete or partial absence of all pigment cells and most other neural-crest derivative cells in zebrafish and mouse , Sox10 is necessary for specification of not only the melanocyte lineage , but also for all non-ectomesenchymal lineages , including xanthophores and iridophores ( e . g . [23 , 29–35] and reviewed in [23] ) . Although the role of sox10 has never been explored in medaka , we predict a key role in specification of the three pigment cell types shared with zebrafish; whether that role extends to leucophores remains to be tested . Among the Sox-family transcription factors [36] , the SoxD subgroup ( Sox5 , Sox6 and Sox13 ) , is unique in lacking any known trans-activation or -repression domains; thus , they exert their functions via obligatory interactions with other transcription factors [37] . The SoxE subgroup ( Sox8 , Sox9 and Sox10 ) are key partners for these SoxD factors ( reviewed in [37] ) . In mammalian chondrocyte formation , neither Sox5/6 nor Sox9 alone efficiently elicits expression of chondrocyte-specific genes , but together they function synergistically in their transcription [38 , 39] . In contrast , in mammalian oligodendrocytes and melanoblasts , Sox5 and/or Sox6 counteract transcriptional activation by Sox9 or Sox10 [40 , 41] . In mouse , Sox5 is dispensable for melanocyte development , but its loss in Sox10 heterozygotes partially rescues a severe reduction of melanoblasts and expression of Mitf and Dct [41] . Although Sox5 and Sox10 are involved in pigment cell development in fish , it remains unclear whether Sox5 and Sox10 interact in the specification of multiple types of pigment cells in fish . Furthermore , the extent of species-specific variation in these interactions and roles within the highly diverse fish has never been explored . In this study we have compared pigment cell phenotypes of single and compound medaka and zebrafish sox5 and sox10 mutants , and sox5-misexpressing medaka . We demonstrate that Sox5 acts antagonistically against Sox10 in the specification of all pigment cell lineages in zebrafish and in melanocyte and iridophore lineages in medaka , whereas synergistic interactions between Sox5 and Sox10 play important roles in specification of xanthophore and leucophore lineages in medaka . Our findings indicate that there are distinct interactions between Sox5 and Sox10 that contribute to generation of different types of pigment cells , and that many , but surprisingly not all , of these are conserved between zebrafish and medaka . Intriguingly , the changed mode of interactions between Sox5 and Sox10 affect the xanthophore , and correlate with the evolution of a novel pigment cell type ( leucophore ) ; given the shared developmental origin of leucophores and xanthophores in medaka , we hypothesize that the switch in Sox5-Sox10 interactions reflects the evolution of this novelty .
Although major vertebrate species , such as mouse and zebrafish , have only one sox10 gene , some fish species including medaka have two sox10 genes , sox10a and sox10b [42] . In medaka , like in other vertebrates , [30 , 32 , 33 , 35 , 43–45] , expression of both sox10a and sox10b was detected in neural crest cells , both prior to and during migration , and in otic vesicle ( Fig 1A–1J ) . At 6-somite stage ( 34 hpf ) , expression was detected in premigratory neural crest cells in the head and anterior trunk ( Fig 1A , 1A’ , 1F and 1F’ ) , and then gradually extended posteriorly and ventrally through the trunk ( S1A and S1B Fig ) . At 24-somite stage ( 58 hpf ) , sox10a- or sox10b-expressing neural crest cells were observed on both the medial and lateral pathway and in lines of cells that from their position in the horizontal myoseptum and their organization we identify as Schwann cell precursors associated with the posterior lateral line nerve ( Fig 1B–1E and 1G–1J ) . This in situ analysis suggested that in medaka sox10a and sox10b were expressed in neural crest cells in a pattern similar to sox5 [18] . To examine the degree of overlap of expression of medaka sox10 genes with sox5 , we used double whole mount in situ hybridization using Tyramide signal amplification ( TSA ) ( S1C–S1E Fig ) . We note the extensive overlap between sox10a and sox5 in numerous migrating neural crest cells ( S1C and S1D Fig ) and also in premigratory neural crest , which possess high multipotency [46] ( S1C and S1E Fig ) . These observations of overlapping sox10 and sox5 expression lead us to examine if Sox5 and Sox10 function together in pigment cell development . We engineered medaka mutants for sox10a and sox10b using TALEN [47] and TILLING methods [48 , 49] . By a TALEN , we established two sox10a mutant lines ( sox10aE1ins10 and sox10aE2del16 ) . The sox10aE2del16 allele has a 16-base nucleotide deletion in the second exon and so is predicted to generate a truncated Sox10a protein lacking the C-terminal of HMG DNA binding domain and the transactivation domain; the sox10aE1ins10 allele is predicted to give a similar , but even more severely truncated protein ( S2 Fig , see the legend of S2 Fig for information of sox10aE1ins10 allele ) . Similarly , we obtained the sox10bE1del7 mutant allele , which has a 7-base nucleotide deletion in the first exon resulting in lack of most functional domains . From a medaka TILLING library , we isolated two sox10b mutant alleles , sox10bN108S and sox10bF110L . They each have a distinct single nucleotide mutation which results in an amino acid substitution of a highly conserved amino acid in the HMG domain of Sox10b ( sox10bN108S: N108S and sox10bF110L: F110L , S2 Fig ) . In the following study , we used the sox10aE2del16 and sox10bN108S alleles unless specified , since we found no difference in phenotypes between E1ins10 and E2del16 alleles , nor among the N108S , F110L and E1del7 alleles . We began by gross characterization of the pigment cell phenotypes of these sox10 mutants ( Fig 1K–1O ) . The sox10a homozygotes showed reduction of the number of pigment cells , but remaining cells were of normal morphology and pigmentation . They had severe defects in melanocyte numbers in the posterior and ventral body , reduced leucophores in the posterior body , and reduction of xanthophores ( Fig 1M ) . The sox10b mutation ( s ) did not show any overt pigment cell phenotype ( Fig 1L ) . In the compound sox10a homozygote and sox10b heterozygote mutants , the pigment cell phenotypes were more severe than those of sox10a homozygotes: all pigment cell types were strongly reduced in numbers except melanocytes and leucophores in the anterior body ( Fig 1N ) . The phenotypes of the double homozygous sox10a;sox10b mutants were strikingly reminiscent of that of zebrafish colourless/sox10 mutants [29 , 30] , lacking all pigment cells , although notably leucophores remained present on the dorsal head ( Fig 1O ) . We note that in every case where a pigment cell phenotype was seen , it was restricted to the number of each pigment cell type–remaining cells looked normal in morphology and pigmentation . Thus , both Sox10a and Sox10b are required for development of pigment cells , except for leucophores on the head , and they have partially redundant functions in medaka . To ask if Sox10a and Sox10b function with Sox5 in the pigment cell development in medaka , we generated compound mutants of sox5ml-3 and sox10aE2del16 and/or sox10bN108S . We refer to sox5ml-3 homozygotes , sox10aE2del16 homozygotes and sox10bN108S homozygotes as sox5-/- , sox10a-/- and sox10b-/- , respectively . We first examined the expression of the melanoblast markers mitfa and dct to elucidate the roles of sox5 , sox10a and sox10b in melanocyte lineage specification ( Fig 2A–2H , 2M and 2N ) . As described previously [18] , numbers of mitfa- or dct-expressing cells were comparable between sox5-/- and wild-type ( WT ) embryos , suggesting that sox5 mutation has little effect on melanoblast formation ( Fig 2B , 2F , 2M and 2N ) . In sox10a-/- embryos , significant decrease of mitfa-expressing and dct-expressing cells was observed ( Fig 2C , 2G , 2M and 2N ) , indicating a key role for Sox10a in medaka melanocyte fate specification . sox10a-/-;sox5-/- double mutant embryos had fewer mitfa-expressing and dct-expressing cells than WT . However the numbers of those cells were larger in sox10a-/-;sox5-/- double mutant embryos than in sox10a-/- mutant ( Fig 2D , 2H , 2M and 2N ) , indicating that loss of Sox5 led to partial rescue of the melanoblast specification defect in the sox10a-/- embryos . We next examined iridophore specification in the mutants using leukocyte tyrosine kinase ( ltk ) as an early iridophore marker , which is known to be expressed in multipotent neural crest cells at early stages and later specifically in the iridophore lineage in zebrafish [46] . In medaka , ltk expression was detected in premigratory neural crest cells in the cranial and anterior trunk neural crest at 12-somite stage ( 41 hpf , S3A Fig ) . At later stages , ltk-expressing cells appeared on the eyes ( S3B Fig ) and in the LP on the dorsal yolk sac [8] ( S3C and S3C’ Fig ) . The position and density of these cells are consistent with those of differentiating iridoblasts . Numbers of the ltk-expressing cells on the yolk sac at 34-somite stage ( 74 hpf ) in sox5-/- embryos were indistinguishable from that in WT ( Fig 2I , 2J and 2O ) . In sox10a-/- embryos , significant reduction of the ltk-expressing cells was observed ( Fig 2K and 2O ) , but it was significantly rescued in sox10a-/-;sox5-/- double mutants ( Fig 2L and 2O ) . Our results indicate that Sox5 functions antagonistically against Sox10 in specification of both melanocyte and iridophore lineages in medaka . We then examined the long-term consequences of loss-of-function of Sox5 and Sox10 on development of pigment cells in hatchlings ( 9 dpf , S4 Fig ) . Both sox10a-/-;sox10b-/- and sox10a-/-;sox10b-/-;sox5-/- mutants displayed complete loss of melanocytes ( S4D and S4H Fig ) , indicating that Sox10 is absolutely required for melanocyte specification , even in the absence of Sox5 . Conversely , sox10b-/-;sox5-/- double mutants were not significantly different to sox10b-/- mutants , being indistinguishable from WT siblings ( S4X Fig ) , and showing that sox10b is insufficient to numerically affect melanocyte numbers , even in the absence of sox5 . Despite the severe decrease in the cells expressing the melanoblast markers at early stages ( Fig 2M and 2N ) , we found a slight increase in the total number of melanocytes in sox10a-/- mutant larvae compared to WT siblings ( S4X Fig ) . In contrast , loss of sox5 activity in the context of either sox10a-/- or sox10a-/-;sox10b+/- mutants resulted in partial rescue of the melanocyte phenotype ( S4F , S4G and S4X Fig ) , suggesting that the increased dct-expressing melanoblasts gave rise to increased melanocytes later in the double mutant , and that Sox5 acts to repress Sox10-dependent melanocyte formation . In the course of these studies we also noted interesting , but complex , region-specific differences in the effects of these mutations . In medaka sox10a-/- mutants , melanocytes were preferentially decreased in VS posterior to the caudal end of yolk sac ( posterior VS ) compared to DS , and in the LS ( S4B , S4U and S4V Fig ) , as is reported for zebrafish hypomorphic sox10 alleles [5 , 50] . Unexpectedly , quantitative assessment revealed that sox10a-/- mutants showed an increase in head melanocytes ( S4S Fig ) , and in the DS and anterior VS ( S4T and S4W Fig ) . In sox10a-/-;sox10b+/- mutants , melanocytes were almost completely absent from the LS and the posterior VS ( S4C , S4U and S4V Fig ) , but with variable , but often rather few residual melanocytes located in the DS , on the head and in the anterior VS ( S4S , S4T , S4W and S4X Fig ) . Both sox10a-/-;sox10b-/- and sox10a-/-;sox10b-/-;sox5-/- mutant larvae had no iridophores on LP ( S4M and S4R Fig ) , indicating that Sox10 is absolutely required for specification of iridophores , even in the absence of Sox5 . Although sox10a-/-;sox10b+/- mutant showed strong reduction of iridophores , loss of sox5 rescued iridophores in hatchlings ( S4L and S4Q Fig ) , suggesting Sox5 also represses Sox10-dependent specification of iridophores . In summary , Sox10 function is absolutely required for specification of melanocyte and iridophore lineages in medaka , and Sox5 antagonizes the action of Sox10 in specification of both these cell types . We then asked how xanthophore specification was affected in sox10 mutants and how this was affected by loss of Sox5 activity . Xanthophore phenotypes were observed in hatchlings , using a combination of the distinctive auto-fluorescence , morphology and location of these cells ( Fig 3A–3H ) . Compared with WT ( Fig 3A ) , sox10a-/- mutants had reduced numbers of xanthophores in the body ( Fig 3B ) . In sox10a-/-;sox10b+/- mutants , reduction was more severe ( Fig 3C ) , and sox10a-/-;sox10b-/- double mutants completely lacked xanthophores ( Fig 3D ) , suggesting that xanthophore formation depends quantitatively on the number of functional sox10 alleles . Loss of sox5 caused complete absence of xanthophores regardless of the mutation status of sox10a and sox10b ( Fig 3E–3H ) . Thus , both Sox5 and Sox10 are essential for xanthophore development . In order to assess genetic interaction between Sox5 and Sox10 in the xanthophore lineage , we combined the sox10 mutant alleles with the heterozygous sox5 mutation . Xanthophore number was examined in 5 dpf embryos using immunohistochemistry with anti-Sox5 antibody , which marks the xanthophore lineage at this stage . Validation of the Sox5-specific nature of this antibody was shown by immunohistochemistry of sox5-/- embryos , which showed the expected absence of specific signal ( Fig 3M ) . We used this anti-Sox5 antibody in double immunofluorescent studies to assess the overlap between Sox5 and Pax7 expression ( S5 Fig ) . We find that at 90 hpf , Pax7-expressing cells are of two classes , those showing Sox5 expression and those without . The former are mainly on the lateral migration pathway and are interpreted as xanthoblasts , whereas the latter can be interpreted as leucoblasts , based on our previous observations that pax7a mRNA is expressed in both the xanthophore and leucophore lineage , but that sox5-/- mutants lack laterally migrating pax7a-expressing cells [18] . We then used the Sox5 antibody to assess the phenotypes of the sox10 and sox5 mutants . Consistent with our previous result that fewer fluorescing xanthophores could be seen in 5 dpf sox5+/- mutants than in WT [18] , Sox5-positive cells were decreased in the 5 dpf mutants compared to WT embryos ( Fig 3I , 3J and 3N ) . In sox10a-/- mutants as well as in sox10a-/-;sox5+/- mutants , Sox5-positive xanthoblasts were severely reduced ( Fig 3K and 3L ) , and there was no significant difference between these mutants ( Fig 3N ) . We then assessed xanthoblasts in a weaker background; sox10a+/-;sox10b-/- mutants showed significant reduction in xanthoblasts compared to WT siblings ( Fig 3O ) . Reduction of a sox5 WT allele led to further reduction of xanthophores in the sox10a+/-;sox10b-/- mutants ( Fig 3O ) . These data indicate that sox5 and sox10 additively promote xanthophore formation in medaka . We assessed leucophore development in sox10 mutant hatchlings ( Fig 4 ) . Loss of Sox10 function had distinct effects in different locations . Although there was no significant difference in total leucophore numbers between WT and sox10a-/- mutants ( Fig 4K ) , leucophores decreased in the body and increased in the head upon loss of Sox10a ( Fig 4A , 4B , 4I and 4J ) . Leucophores were significantly further reduced in sox10a-/-;sox10b+/- mutants ( Fig 4C and 4J ) and more severely reduced in sox10a-/-;sox10b-/- mutants in the body ( Fig 4D and 4J ) . In contrast to the body , reduction of functional sox10 alleles resulted in increasing numbers of leucophores on the dorsal surface of the head ( Fig 4I ) . These surprising findings indicate that loss of Sox10 led to progressive decrease in leucophore formation in the body , but to progressive increase in the head . However , there was no difference in the total leucophore numbers between WT and sox10 compound mutants ( Fig 4K ) . We then examined the effect of the loss of Sox10 on the sox5-mutant leucophore phenotype . As described previously [18] , loss of Sox5 function alone caused significant increase in leucophores compared to WT ( Fig 4E and 4L ) . Additional loss of sox10a and sox10b in the sox5 mutants enhanced the leucophore phenotype: sox10a-/-;sox5-/- and sox10a-/-;sox10b+/-;sox5-/- mutants had more leucophores than sox5-/- mutants ( Fig 4F , 4G and 4L ) . The data suggest that Sox10 functions in parallel to Sox5 to repress leucophore formation . As sox10a-/-;sox10b-/-;sox5-/- triple mutants had comparable numbers of leucophores to WT ( Fig 4H and 4L ) , Sox10 is required for the Sox5-mediated suppression of leucophore development . The differential effect of sox10 on leucophores in the head and trunk is striking and paradoxical . In the trunk and tail , leucophores seem to behave like all other pigment cell-types , with numbers proportional to the number of functional sox10 alleles , and there being none except the odd ‘escaper’ in the absence of any functional sox10 alleles . In contrast , in the head , the pattern is the reciprocal , with leucophore number increasing progressively with loss of functional sox10 alleles . One plausible explanation of this is that any maternal contribution of sox10 might be most active in the head region , simply because the neural crest develops earlier in this region . Thus , in the head , but not in the trunk , we might expect rescue of the zygotic phenotype by maternally-supplied Sox10 activity . To test this idea , we used RT-PCR to examine medaka embryos at 2–4 cell stage ( well-before mid-blastula transition ) , as well as at stage 25 ( 18–19 somite stage ) when zygotic transcription is known to be active ( S6 Fig ) . Consistent with our hypothesis , maternal contribution of sox10a and sox10b was readily detected , as indeed was sox9a and sox9b . To understand the general role of Sox5 and Sox10 in pigment cell development in teleosts , we conducted similar genetic analyses with zebrafish mutants and compared their phenotypes with those of medaka . In zebrafish , sox10 expression is prominent throughout the early neural crest from premigratory stages , and persists in cells on the medial and lateral migration pathways , before levels become gradually decreased in differentiating cells , with the exception of cartilage and peripheral glia which maintain strong expression [23 , 30 , 51 , 52] . In contrast , sox5 expression was detected in trunk neural crest at 18 hpf ( S7A , S7C and S7E Fig ) , overlapping with sox10 expression ( S7B , S7D and S7F Fig ) , but seemed much weaker than in medaka . In an independent study using single cell transcriptional profiling , we have confirmed the expression of sox5 in flow cytometry ( FACS ) -purified neural crest cells , with highest expression corresponding to likely premigratory neural crest cells ( Subkhankulova et al . , in preparation ) . It is noteworthy that sox5 expression was transient and faint in zebrafish whereas in medaka it is retained in migrating xanthophore precursors [18] . We established a novel sox5 mutant ( sox5E4del7 ) in zebrafish by using the CRISPR/Cas9 system [53] ( S2 Fig ) . Homozygous sox5E4del7 mutants ( referred to as sox5-/- below ) showed normal pigment cell pattern during larval stages ( Fig 5A and 5B , shown for swim bladder inflation stage [54] ) . Unlike medaka sox5 homozygotes , zebrafish sox5-/- mutants did not show apparent changes in the formation of xanthophores ( Fig 5C and 5E ) nor of xanthophore precursors expressing GTP cyclohydrolase 1 ( gch ) [55] ( Fig 5D and 5F ) . In order to test genetic interaction between sox5 and sox10 in zebrafish , we generated sox5;sox10 double mutants . We employed two zebrafish mutant alleles: colourless/sox10t3 as a likely null allele exhibiting an absence of normal melanocytes with severe defects in xanthophores and iridophores [29 , 30] and sox10baz1 allele , which has a single nucleotide change causing an amino acid substitution ( V117M ) in the HMG box domain ( S2 Fig ) and is a hypomorphic mutant allele [23 , 51] ( S8 Fig ) . We hypothesized that given that the Sox10baz1 mutant protein retains weak Sox10 activity , sox10baz1 mutants give a condition more similar to medaka sox10a-/- or sox10a-/-;sox10b+/- mutants . Most sox10baz1/baz1 mutants had no melanocytes except for some visible simply as tiny residual specks of melanin , while a small minority had a few partially-differentiated melanocytes on the head ( Fig 5G ) . Loss of sox5 in sox10baz1/baz1 mutants resulted in an increased number of partially-differentiated melanocytes ( Fig 5G and 5H ) . sox10baz1/baz1 mutants have some visible xanthophores on the head and very few xanthophores in the trunk ( Figs 5I , 5J , S8E and S8E’ ) . Reduction of sox5 WT alleles in sox10baz1/baz1 mutants appears to result in increasing numbers of xanthophores , but we were not able to quantify xanthophore formation in sox10baz1/baz1 and sox5-/-;sox10baz1/baz1 mutants due to their dense localization on the dorsal head ( Fig 5J–5L ) . In sox10t3/t3 mutants , which have no or only a few fluorescing xanthophores , reduction of sox5 WT alleles significantly increased xanthophore numbers ( Fig 5M–5P ) . These results indicate that Sox5 acts to repress Sox10-dependent melanocyte and xanthophore formation . Whereas we were unable to demonstrate a significant alteration in residual iridophores in sox10baz1/baz1 mutants that had also lost one or both sox5 alleles , sox5 allele loss in sox10t3/t3 mutants rescued iridophores to a small but significant extent ( Fig 5Q and 5R ) . Taken together , these data show that Sox5 antagonizes Sox10 function in development of all three pigment cell types in zebrafish . Medaka sox5 homozygotes survive to adulthood and look normal except that they show an increase of leucophore number and decrease of xanthophore number ( Fig 6A–6C ) . Thus , the function of Sox5 in adult xanthophore and leucophore development is reminiscent of that at embryonic stages . Zebrafish sox5-/- mutants looked normal at larval stages , but had fewer horizontal melanocyte stripes at 60 dpf . Whereas WT zebrafish had five stripes , sox5-/- mutant had three to four stripes ( Fig 6D–6G ) . We examined post-metamorphic phenotypes of xanthophores and melanophores in sox5-/- mutant at juvenile to adult stages [54] . The width of melanocyte stripe ( 1D ) is marginally , but significantly larger in sox5-/- mutants compared to WT siblings , and the xanthophore interstripes ( X0 and X1D ) were significantly wider in sox5-/- mutants than in WT ( Fig 6D” , 6E” and 6H ) . We found that each interstripe in sox5-/- mutants contains an increased number of xanthophores but the melanocytes number in melanocyte stripe was not altered ( Fig 6H ) . These results suggest that adult zebrafish xanthophore formation is upregulated by loss of Sox5 activity . Thus , Sox5 acts to repress adult xanthophore formation , similar to its role in xanthophore development at embryonic/larval stages , as observed in a sox10-defective sensitive background . Our data also indicate the contrasting role of Sox5 in xanthophore development between medaka ( promoting ) and zebrafish ( repressing ) . The difference in the genetic interactions between sox10 and sox5 in xanthophore development medaka and zebrafish were surprising , but they correlate with another important difference in the pigment cells of these fish species , specifically the presence of leucophores in medaka , but not in zebrafish . In previous work we and others have shown that leucophores are best considered as a modified form of xanthophore , and that the key distinction correlates with maintenance of sox5 expression in cells that will become xanthophores [18] . Importantly , our previous data suggested that maintenance of sox5 function in shared xanthophore/leucophore progenitors was incompatible with leucophore development , although we had not directly tested this hypothesis . In the context of the data presented here showing that xanthophore formation in zebrafish does not require Sox5 activity , we suggest that medaka Sox5 may have acquired an evolutionarily novel function in driving xanthophore development , associated with the evolution of the related leucophore cell-type . To test the role of Sox5 function in promoting the xanthophore and repressing the leucophore lineages , we generated medaka transgenic lines TgBAC ( pax7a:loxp-dsred-loxp-sox5WT or ml-3 ) that misexpress either WT or sox5ml-3 mutant form of Sox5 protein in the shared xanthophore/leucophore progenitors ( controlled by use of pax7a promoter ) when in the presence of Cre recombinase ( excises dsRed cDNA at the loxp sites upstream of sox5 cDNA; Fig 7A–7C ) . When Cre mRNA was injected into sox5-/-;TgBAC ( pax7a:loxp-dsred-loxp-sox5WT ) embryos , xanthophore formation was restored in the hatchlings ( Fig 7D’ and 7E’ ) while excessive formation of leucophores was suppressed ( Fig 7D and 7E ) . We also expressed Cre mRNA using an hsp70 promoter in Tg ( hsp70:Cre ) line . When TgBAC ( pax7a:loxp-dsred-loxp-sox5WT ) ;Tg ( hsp70:cre ) embryos were treated with heat shock at the end of gastrulation , the hatchlings exhibited decreased leucophore numbers compared to those without heat shock ( Fig 7F–7H ) . The misexpression of Sox5ml-3 mutant protein failed to alter the formation of leucophores at the hatching stage ( S9 Fig ) . These results support the hypothesis that Sox5 in medaka has evolved a novel , major role in driving xanthophore fate and repressing leucophore fate within the pax7a-expressing shared progenitors in medaka that is coupled to the evolution of the novel leucophore cell-type .
Fate specification of neural crest cells is a key aspect of their development , and a model for stem cell fate choice . Fate specification of pigment cells forms a nice ‘model within a model’ , but has an interesting added complexity in fish since the neural crest cells give rise to a variety of pigment cell types , and these vary between species . Our present study sought to dissect the mechanisms of pigment cell fate specification and their evolution , focusing on the genetic interactions of sox5 and sox10 in medaka and zebrafish . Taking advantage of duplicated sox10 genes in medaka , we were able to produce a variety of sox10 dosages in medaka and thereby a wide range of phenotypes in pigment cells . Although the role is shared by the two sox10 genes , Sox10 function is required for specification of iridophores , xanthophores and melanocytes in medaka , just as in zebrafish ( and in mammals ) [25 , 30 , 32 , 33 , 56] . Thus , the mutant phenotypes are enhanced by simultaneous loss of sox10a and sox10b , and given their overlapping expression this clearly suggests functional redundancy of sox10a and sox10b in pigment cell specification . However , since medaka sox10a single mutants exhibited significant decrease in numbers of melanoblasts , iridoblasts and xanthophores , whereas sox10b single mutants appeared normal , medaka Sox10a seems to play a dominant role for specification of these pigment cells whereas Sox10b seems dispensable . In contrast to the other three pigment cell types , medaka sox10a-/-;sox10b-/- double mutants retained a considerable number of leucophores , especially on the head . Given that the mean total number of leucophores is apparently unaffected in sox10 mutant medaka , it is conceivable that leucophore development from the neural crest is completely independent of SoxE function , although to explain the shift in the distribution towards the head as sox10 allele number decreases , this hypothesis would also necessitate that endogenous leucophores in WT embryos derive from the head NC and then migrate posteriorly to populate the trunk and tail . However , given the essential role for Sox10 in the specification of all non-ectodermal fates including pigment cells in zebrafish ( reviewed in [23] ) , we prefer the more conservative hypothesis that the development of leucophores may simply be achieved using lower levels of SoxE activity , perhaps especially Sox9b , similar to that shown in zebrafish where additional loss of sox9b results in loss of the residual melanocytes and sensory neurons of the DRGs in colourless/sox10 mutants [23 , 52] . To explain the shift in leucophore distribution towards the head as the numbers of functional sox10 alleles are decreased , we propose that either a maternal contribution of SoxE proteins has a preferential effect on the earlier differentiating head NC or that Sox9b expression in the early NC is sufficient for specification of leucophores , but much less so for other pigment cell types; as the numbers of mutated sox10 alleles increases , so more and more of the head NC remain unspecified and so remain vulnerable to the residual Sox9b activity . Consistent with both of these latter hypotheses , we were able to show the presence of detectable levels of all four sox10 and sox9 gene transcripts in the early cleavage stage zygote , demonstrating maternally supplied soxE activity . We note the highly variable nature of leucophore numbers in the sox10 mutants , with some fish having essentially no leucophores , which suggests that leucophore specification is likely to be Sox10-dependent , but can be rescued in an unreliable manner in the head NC . Future study of medaka sox10a;sox10b;sox9b triple homozygous mutants would allow further testing of these ideas . Our data suggests that , in medaka and zebrafish , Sox5 function has similar effects in development of both embryonic and adult metamorphic pigment cells , but that the relative importance of Sox5 in each stage varies between species . In medaka sox5 mutant adult fish have reduced xanthophores and increased leucophores , which reflects the embryonic/larval phenotypes . However , in zebrafish , loss of Sox5 did not apparently affect the formation of any chromatophore types at embryonic/larval stages , indicating that Sox5 is dispensable for the development of embryonic pigment cells in zebrafish , similar to what has been shown in mice [41] . On the other hand , sox5-deficient adult zebrafish exhibited a clear phenotype , with increased numbers of xanthophores , but little effect on melanocytes , resulting in expanded interstripes with increased xanthophores and normal melanophore stripes . While Sox5 function is maintained after metamorphosis , our data clearly demonstrate contrasting roles for Sox5 activity on xanthophore fate specification in medaka and zebrafish , being absolutely required for xanthophore specification in medaka , but weakly antagonizing xanthophore formation in zebrafish . Although Sox5 has no known transcriptional regulatory domains , it has been shown to collaborate with other transcription factors to regulate expression of downstream target genes [38–41] . We speculate that in medaka and zebrafish , Sox5 is likely to interact with different partner transcription factors controlling xanthophore specification , or to have differential affinity with similar partners in each species , which subsequently results in constructing different gene regulatory networks and thus establishing species-specific transcriptional regulation of target genes . Sox5 counteracts Sox10 activity in the specification of melanocytes , iridophores and xanthophores in zebrafish , and of melanocytes and iridophores in medaka . Whereas loss of Sox5 alone had little impact on the formation of these chromatophore types , it partially rescued the severe reduction of these pigment cells due to Sox10 deficiency , revealing a subtle antagonistic function of Sox5 against Sox10 . This mode of Sox5 and Sox10 interaction has been reported in mouse melanocytes and in human melanoma cells [41 , 57] , suggesting that an antagonistic interaction between Sox5 and Sox10 is widely conserved among vertebrates . In melanocytes , Sox5 antagonism likely functions through direct competition with Sox10 binding at a key regulatory binding site [41] , but mechanisms seem to be varied and case-specific [37] , so detailed investigation will be required to ascertain those functioning in each pigment cell case identified here . In contrast , Sox5 and Sox10 function collaboratively in specification of xanthophores in medaka . Since sox5 is a member of a group of related ( SoxD ) genes , it is conceivable that a SoxD function in zebrafish xanthophores is performed by another member of this group; however current expression data provide no evidence for this , so that this question is ripe for future investigation ( ZFIN , sox6; https://zfin . org/ZDB-GENE-081120-6 , sox13; https://zfin . org/ZDB-GENE-100519-1 ) . Furthermore , we suggest that there may be no reason to expect conservation of this function in xanthophores . This contrast in the role of Sox5 in xanthophore specification between medaka and zebrafish is striking in that it correlates with the fact that in medaka , but not in zebrafish , xanthophores appear to derive from a shared ( bipotential ) progenitor and Sox5 is expressed in that progenitor , but also maintained at high levels in the xanthophore lineage ( Fig 3 ) [17 , 18] . To explain the difference in the interactions between Sox5 and Sox10 , we hypothesize that Sox5 functions independently in two aspects of xanthophore development ( Fig 8 ) . In the first of these xanthophore progenitors are formed from multipotent chromatoblasts under the control of antagonistic interactions between Sox5 and Sox10 . In a second step , unique to medaka , the xanthophore progenitor has acquired xanthophore/leucophore bipotency , and Sox5 plays a dominant role in xanthophore fate choice , being absolutely required for xanthophore specification . As attested by the details of the sox5 single mutant phenotypes , the role for Sox5 in the first process is relatively weak , whereas in the second step Sox5 is the dominant player . Consistent with this idea , we show here that forced sox5 expression in the shared xanthophore/leucophore progenitor is sufficient to drive xanthophore fate and to repress leucophore fate . Although this effect is substantial , in our hands we did not achieve a complete switch of leucophores to xanthophores; we attribute this to the experimental conditions used , which mean that the duration of expression of Cre is limited , and the timing of heat shock-mediated Cre induction , when some neural crest cells may already have become specified to a leucophore fate and so might be expected to be in a state that was more resistant to the effects of sox5 re-expression . Melanocytes , xanthophores and iridophores are widely-distributed amongst teleost lineages [4] . In contrast , to date leucophores have been identified in just a few species , including guppy ( Lebistes reticulatus ) , Japanese flounder ( Paralichthys olivaceus ) , darkbanded rockfish ( Sebastes inermis ) , and killifish ( Fundulus heteroclitus ) [10–13 , 23 , 58 , 59] , and hence those in medaka are likely an evolutionary novelty compared to the shared common ancestor with zebrafish . Although leucophores have been considered to be closely related to iridophores since the appearance of both cell types is dominated by reflection of light due to intracellular purine-crystal accumulation , our previous work showed that leucophores share an ontogenetic origin with xanthophores i . e . they share a ( bipotent ) progenitor [17 , 18] . Thus , we speculate that medaka evolved the leucophore by expanding the potency of a xanthophore progenitor and using loss of Sox5 expression to drive accumulation of intracellular purine crystals in a xanthophore-like cell . Future work will explore the changes in transcriptional targets and partner proteins of Sox5 , and especially the development of cooperative partnership between Sox5 and Sox10 , in order to provide insight into the acquisition of new cell types during neural crest evolution .
Medaka and zebrafish care was approved by Nagoya University ethics committee and in accordance with local and Japanese ethical guidelines or by AWERB committee of University of Bath and in accordance with the Animals Scientific Procedures Act 1986 of the UK . Medaka wild type ( WT ) strain Nagoya and the mutant strain sox5ml-3 [18] were used . Medaka sox9bK136X TILLING mutant was provided by NBRP medaka [60] . Zebrafish AB and the mutant strains sox10baz1 and sox10t3 [23 , 30] were used . In some experiments , embryos were treated with 0 . 003% phenylthiourea to inhibit melanin synthesis . Putative medaka sox10b and ltk genes were identified in Ensemble ( http://asia . ensembl . org/Oryzias_latipes/Info/Index ) : sox10b; ENSORLG00000014587 , ltk; ENSORLG00000015434 . An entire open reading frame ( ORF ) of sox10b and a partial fragment of ltk were amplified from cDNA of medaka Orange red strain using following primers: 5’-ATGTCCAGGGAGGAGCAGAGCCT-3’ ( sox10b-forward ) , 5’-TCATGGCCGCGACAGAGTTGTGTA-3’ ( sox10b-reverse ) , 5’-CTGCTCTACACTTCCTGTGTGCCTG-3’ ( ltk-forward ) and 5’-GGCAGATCTTCTGGGTGAGGAGAT-3’ ( ltk-reverse ) . The amplified fragments were subsequently cloned into pTAC-2 vector ( BioDynamics Laboratory Inc ) . A putative sox10a gene was identified by BLASTN against zebrafish Sox10 protein ( AF402677 ) in Ensembl ( http://asia . ensembl . org/Oryzias_latipes/Info/Index ) . First , the identified region was cloned and sequenced . To determine full cDNA transcript of sox10a , 3’- and 5’- RACE were performed using gene specific primers . The primer sequences are available on request . The entire ORF of sox10a was cloned using the primers; 5’-ATGTCCGGTGAGGAGCACAGCCT-3’ ( forward ) and 5’-TCAAGGTCTTGTGAGGGTGGTGTAAA-3’ ( reverse ) . The full-length sequence of medaka sox10a cDNA has been deposited in GenBank ( accession number MF276972 ) . TILLING screening was performed as described [49] . The exon 1 of medaka sox10b gene , which includes a part of HMG box domain , was amplified for the screening using primers: 5’-GCAGAGCTGAGAATTCACAAA-3’ ( forward ) and 5’-ACCTGTCGGGTCTGTCA-3’ ( reverse ) . The designing and assembling of TALENs were performed as described [47] [61] . The exon 1 and exon 2 sequences of medaka sox10a and exon 1 of medaka sox10b was submitted to the TALE-NT 2 . 0 ( https://tale-nt . cac . cornell . edu/ ) . We selected the following sequences as candidates for sox10a exon 1-specific target: 5’-TTCTGAGTCGGAGCTGAGC-3’ ( left , sense ) and 5’-TGAGAGTGAGTGACCGTCC-3’ ( right , antisense ) , sox10a exon 2-specific target: 5’-TGCTGAACGAGAACGACAA-3’ ( left , sense ) and 5’-TCCTCAGCCTCTCCGCCTC-3’ ( right , antisense ) and sox10b exon 1-specific target: 5’-TCTCAGAGGTCGAGCTC-3’ ( left , sense ) and 5’-TGGGAGCAGCTGTCATCC-3’ ( right , antisense ) . The TALEN expression vectors were linearized with NotI ( NEB ) and then used as templates for capped mRNA synthesis using the mMessage mMachine SP6 kit ( Life Technologies ) . Equal amounts of mRNAs were mixed and microinjected into a blastomere of 1-cell stage embryos . To generate small guide RNA ( sgRNA ) targeting the exon 4 of zebrafish sox5 gene , we constructed a pT7-gRNA vector ( Addgene plasmid # 46759 ) which harbors annealed oligonucleotides with the following sequences; 5’-TAGGCATGGGTTCTGGCAACTT-3’ ( sense ) , 5’-AAACAAGTTGCCAGAACCCATG-3’ ( antisense ) [62] . The vector was linearized with BamHI ( NEB ) and used as a template for sgRNA synthesis using T7 RNA polymerase ( Promega ) . Cas9 mRNA was synthesized with mMessage mMachine SP6 kit ( Life Technologies ) on NotI-digested pCS2-hSpCas9 ( Addgene plasmid #51815 ) [63] . One nl of the RNAs solution ( 300 pg/nl Cas9 mRNA and 200 pg/nl sgRNA ) was injected in the blastomere of 1-cell stage embryos . We constructed transgenic vectors using a BAC clone , which contains the entire exons of pax7a gene ( ola-008A15 , [17] ) . To obtain sox5wt or ml-3-2A-venus , sox5 sequences were amplified from cDNA of WT or ml-3 mutant by following primers and subsequently cloned into pGEM T-easy vector; WT form of sox5; 5’-GAATTCATGCTCACTGAGCCTGAGCTACCT-3’ ( forward ) and 5’-GTCGACTTGGGTGATGTGGTTCTCCTTGTC-3’ ( reverse ) , ml-3 form of sox5; 5’-GAATTCATGCTCACTGAGCCTGAGCTACCT-3’ ( forward ) and 5’-GTCGACTGGAGCTGATGGGGCCCAGCTTGG-3’ ( reverse ) . Then , sox5 fragment digested by EcoRI and SalI , and 2A-Venus fragment digested by SalI and XbaI from pBLSK-2A-venus that contains the 2A peptide sequence from Thosea asigna virus ( TaV ) . To obtain loxp-dsred-loxp-sox5wt or ml-3-2A-venus cassette , β-actin pro:loxp-dsred-loxp-egfp ( kindly provided by Dr . Kinoshita ) was first digested by SalI and SmaI and the loxp-dsred-loxp fragment inserted into PL451 digested by SalI and EcoRV . Next , to insert sox5wt or ml-3-2A-venus into the second loxp site of PL451-loxp-dsred-loxp , pCSII-sox5wt or ml-3-2A-venus was digested by EcoRI and NsiI and PL451-loxp-dsred-loxp was digested by EcoRI . The fragments produced were treated with KOD polymerase to form their end into blunt end , and then they were ligated . To make BAC construct , the sequence of pax7a exon 1 downstream of the start codon was replaced with the loxp-dsred-loxp-sox5wt or ml-3-2A-venus cassette to obtain pax7a:loxp-dsred-loxp-sox5wt or ml-3-2A-venus vectors . The BAC was modified as described previously [64] . The primers used in the construction were followings; 5’-TGAAGGCTCGATCAGTGTCCAGTTGGGTGTTTTTGCCTGGGTGGCTAGATCCTCGAGGTCGACATAACTT-3’ ( left ) , 5’-TCGAACACAAGTCCAAAAAAGAAGCACTCCTTGTCCCCTTCCTTACCTTCGTACCTGACTGATGAAGTTC-3’ ( right ) . Transgenic strains used are: Nagoya;TgBAC ( pax7a:loxp-dsred-loxp-sox5wt-2A-venus ) and Nagoya; or sox5ml-3/ml-3;TgBAC ( pax7a:loxp-dsred-loxp-sox5wt-2A-venus;pax7a:loxp-dsred-loxp-sox5ml-3-2A-venus ) . Tg ( hsp70:cre ) fish was a gift from Dr . J . Wittbrodt ( EMBL ) . Cre recombinase was induced with heat shock at 40°C for 40 min . To genotype the medaka sox10aE2del16 allele , sox10bN108S , sox10bE1del7 alleles , and zebrafish sox5E4del7 allele , we detected increased mobility of PCR amplicon in 12% polyacrylamide gel electrophoresis due to the deletion [65 , 66] . The PCR primers used are: medaka sox10aE1ins10; 5’-GCATTAATCCCTGGTGGATCC-3’ ( forward ) and 5’-GGCTGAGGCGGTGAGAGTGAGT-3’ ( reverse ) and sox10aE2del16; 5’-CTCCCTCTAGGCTGCTGAACGAGA-3’ ( forward ) and 5’-GAGACCCTGCGCCCACATTGTGAT-3’ ( reverse ) , medaka sox10bN108S; 5’-CTGTTCTTCCGCCAAATCCGACGA-3’ ( forward ) , 5’-CCACACCTGTCGGGTCTGTCA-3’ ( reverse ) , medaka sox10bE1del7; 5’-AATGTCCAGGGAGGAGCAGAGCCT-3’ ( forward ) and 5’-GTCGTCGGATTTGGCGGAAGAACA-3’ ( reverse ) , zebrafish sox5; 5’-TGAGAGGCTGTTGTCTAAGGA-3’ ( forward ) and 5’-TAAGCTCAGAGGTCACATGAA-3’ ( reverse ) . As for sox10bN108S allele , the PCR products were treated with XmnI ( NEB ) : Only WT allele is digested , which can be detected as increased mobility in 1% agarose gel electrophoresis . Whole mount in situ hybridization and plastic sectioning were performed as previously described [18] . The antisense riboprobes were as follows: dct and mitfa of medaka [18] and sox10 of zebrafish [30]; and sox10a , sox10b and ltk were synthesized from the plasmid described above using SP6 polymerase ( Promega ) after restriction enzyme digestion: sox10a and sox10b; EcoRI , ltk; XhoI ( NEB ) . For double fluorescent in situ hybridization , the probes were labeled with digoxigenin or fluorescein , and signals were detected with anti-DIG or anti-Fluor POD-conjugated antibodies with using Tyramide Signal Amplification system ( Invitrogen ) . For sectioning , the stained samples were embedded in Technovit 8100 ( Heraeus Kulzer ) and sectioned at 10μm thickness . Images of stained embryos were taken using Leica MZ APO with attached AxioCam camera ( Zeiss ) or Zeiss AXIOImager . M2 with attached Orca-frash 4 . 0 camera . Images of sections were taken using Zeiss AxioPlan2 with an AxioCam camera . Confocal images were obtained with a Zeiss LSM880 laser-scanning confocal microscope . The anti-medaka Sox5 rabbit antiserum was prepared by using DYASDNENHITQ synthetic peptide as an epitope of antigen ( corresponding to the residues 674–685 of 685 amino acid protein , accession number EF577484 ) . Embryos were fixed in 4% PFA/PBS at 4°C overnight . If necessary , samples were treated with 3% H2O2 and 0 . 5% KOH in PBS to remove residual melanin before blocking . A primary antibody was diluted 1:300 in 5% goat serum/PBS solution . Biotin-conjugated anti-rabbit IgG ( VECTOR ) was used as secondary antibody at a 1:500 dilution . Signals were detected by using VECTASTAIN ABC Elite kit ( VECTOR ) with diaminobenzidine ( DAB , MUTO PURE CHEMICALS ) . For double immunohistochemistry , in combination with anti-Sox5 rabbit antiserum , anti-Pax3/7 mouse monoclonal antibody DP311 ( [67] , kindly provided by Dr . Nipam H Patel ) was used as a primary antibody to detect Pax7 protein ( 1:200 ) . Secondary antibodies used were Alexa Fluor 488 goat anti-rabbit and Alexa Fluor 568 goat anti-mouse ( Molecular Probes , 1:500 ) . In order to count melanocytes , embryos or larvae were treated with 2mg/ml epinephrine to aggregate melanin for 30 minutes . Auto-fluorescence of xanthophores was detected by UV light exposure through DAPI filter . In order to count adult pigment cells , fish were treated with 1μM melatonin for 30 min . The images were taken as described above . Ten medaka or zebrafish embryos for each stage were used to collect total RNA with Sepazol ( SIGMA ) . Total RNAs were treated with RQ1 DNase to prevent genomic DNA contamination . One μg of total RNA for each stage and species was used for synthesis of first-strand cDNA with the M-MLV Reverse Transcriptase ( Promega ) using Random hexamer primer ( Promega ) . The resultant cDNA samples were subjected to semi-quantitative PCR amplification ( 30 cycles of 94°C for 30 sec , 60°C for 30 sec and 72°C for 1 min ) and run in an agarose gel . The band of PCR products was evaluated in comparison with a housekeeping gene , ef1α . Primer information is listed in Supplementary S1 Table .
|
How individual cell fates become specified from multipotent progenitors is a fundamental question in developmental and stem cell biology . Body pigment cells derive from a multipotent progenitor , but while in zebrafish there are three types of pigment cells ( melanocytes , iridophores and xanthophores ) , in medaka these progenitors form four ( as zebrafish , plus leucophores ) . Here , we address whether mechanisms generating each cell-type are conserved between the two species . We focus on two key regulatory proteins , Sox5 and Sox10 , which we previously showed were involved in pigment cell development in medaka and zebrafish , respectively . We compare experimentally how the two proteins interact in regulating development of each of the pigment cell lineages in these fish . We show that development of all pigment cells , except leucophores , is dependent on Sox10 , and that Sox5 modulates Sox10 activity antagonistically in all pigment cells in zebrafish , and melanocytes and iridophores in medaka . Surprisingly , in medaka , Sox5 acts co-operatively with Sox10 to promote xanthophore fate and to repress leucophore fate . Our findings reveal surprising diversity how Sox5 and Sox10 interact to govern pigment cell development in medaka and zebrafish , and suggest that this likely relates to the evolution of the novel leucophore pigment cell type in medaka .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"fish",
"medicine",
"and",
"health",
"sciences",
"vertebrates",
"pigments",
"animals",
"neuroscience",
"alleles",
"epithelial",
"cells",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"model",
"organisms",
"stem",
"cells",
"materials",
"science",
"experimental",
"organism",
"systems",
"chromatophores",
"embryos",
"research",
"and",
"analysis",
"methods",
"embryology",
"developmental",
"neuroscience",
"cell",
"potency",
"animal",
"cells",
"melanocytes",
"materials",
"by",
"attribute",
"biological",
"tissue",
"neural",
"crest",
"genetic",
"loci",
"zebrafish",
"cellular",
"neuroscience",
"eukaryota",
"neural",
"stem",
"cells",
"cell",
"biology",
"anatomy",
"phenotypes",
"genetics",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"multipotency",
"organisms"
] |
2018
|
Distinct interactions of Sox5 and Sox10 in fate specification of pigment cells in medaka and zebrafish
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.