text
string
predicted_class
string
confidence
float16
Clinical course and laboratory findings. The prolonged prothrombin time (PT) and partial thromboplastin time (PTT) values were consistent and factor V activity was consistently measured as low (15%); aspirin discontinuance abruptly increased the value of the area under the curve for ticagrelor from 22 to 52. AD = admission day, aPTT = activated PTT, AUC_aspirin = area under the curve for aspirin, AUC_ticagrelor = area under the curve for ticagrelor, HgB = hemoglobin, INR = international normalized ratio, IV = intravenous, PRC = packed red cell, PT = prothrombin time, PTT = partial thromboplastin time.
clinical case
99.8
The patient was readmitted due to recurrent epistaxis, hemoptysis, and cough on day 26 after coronary stenting. HgB and the AUCs of aspirin and ticagrelor were 10.7 g/dL, 19, and 22, respectively, whereas PT, PTT, and factor V activity showed no significant changes (Fig. 2). Due to the recurrent epistaxis and oozing of the nasal mucus membrane, the aspirin was stopped and only the daily maintenance dose of 90 mg of ticagrelor was maintained. HgB remained constant at 9.5 g/dL, and epistaxis and hemoptysis both stopped. The factor V activity and AUC value of ticagrelor 5 days after stopping aspirin were 12 and 52, respectively. After symptoms improved, she was discharged on day 35 with only 90 mg of ticagrelor daily.
clinical case
99.94
The HgB level was 10 g/dL and the AUC of ticagrelor was 24 at 9 weeks (Fig. 2), and the patient led healthy life for 9 months without any recurrent symptoms. At the 9-month follow-up, the AUC of ticagrelor was stabilized at 18 and the other test results were more stabilized compared with those at the start of coronary stenting (Fig. 2).
clinical case
99.94
We present the case of a patient with F5D and unstable angina who underwent coronary stenting and antithrombotic maintenance therapy with a single antiplatelet agent due to recurrent multiple mucosal bleeding shortly after coronary stenting. In particular, for the first time, we have described maintenance therapy after coronary stenting in an F5D patient. As F5D is a rare hematological disease with an incidence of approximately 1 in a million, it has no established treatment guidelines. Furthermore, there are no guidelines or studies on antithrombotic therapy after coronary stenting in F5D patients.
clinical case
99.94
Although F5D is defined as a mild and severe disease in cases with activity >5% and <1%, respectively, factor V activity does not necessarily predict bleeding severity or clinical features, and many patients may present with a mild disease even with factor V activity of <1%. However, F5D therapy depends on the factor V activity level, and is classified as bleeding control and the removal of inhibitors or antibodies of factor V. FFP or platelet transfusion is recommended to control bleeding. In particular, transfusing FFP is recommended to maintain factor V activity at >25% to 30% in cases of bleeding, invasive testing, or surgery. In this case, FFP was not transfused because coronary stenting was performed without knowledge of the degree of factor V activity or the existence of F5D. We recognized F5D due to hemoptysis, epistaxis, and hematochezia after coronary stenting. However, we first planned to reduce or modify the dosage of antithrombotic drugs due to the risk of stent thrombosis. This strategy was based on previous reports in which factor V activity was not related to bleeding severity. Moreover, successful childbirth delivery was possible without an FFP transfusion, and the success rate for bleeding control was low with FFP transfusion. Fortunately, mucosal bleeding in the present case was controlled by adjusting the antithrombotic drug dose. Treatments for the inhibitors or antibodies to factor V are another option and include plasmapheresis, immunosuppressants, steroids, intravenous globulins, and a monoclonal antibody. Such therapies were not used in our case because the inhibitors and autoantibodies for factor V were unknown and such therapies have weak evidence.
clinical case
99.75
Another notable finding is the effect of 100 mg of aspirin daily on mucosal bleeding in F5D patients. P2Y12 generally plays a key role in dual antiplatelet therapy (DAPT). In this instance, the AUC values of aspirin and ticagrelor were 8 and 10, with 100 mg of aspirin daily and 90 mg of ticagrelor twice a day, respectively. The AUC value for ticagrelor was 22 for a daily regimen of 100 mg of aspirin and 90 mg of ticagrelor, but the AUC for ticagrelor increased to 52 when only the 90 mg of ticagrelor was maintained. This observation emphasizes that ticagrelor's antithrombotic effect is affected by aspirin. In fact, the Platelet Inhibition and Patient Outcomes trial showed that daily combination of 100 mg of aspirin and ticagrelor was superior to aspirin and clopidogrel combined, but the use of 300 mg/day of aspirin wiped out the benefit. Moreover, Kirkby et al demonstrated that aspirin provides additional antiaggregatory effects when only a partial P2Y12 blockade is achieved. Daily 90 mg of ticagrelor administration means incomplete P2Y12 inhibition and 100 mg of aspirin has the enhancing effect of suppressing platelet aggregation for the daily dose of 90 mg of ticagrelor. Aspirin discontinuance with a daily dose of 90 mg of ticagrelor causes ticagrelor to have a weakened antiplatelet effect, which can be expressed as an increased AUC value for ticagrelor. Meanwhile, it is unclear whether the disappearance of mucosal bleeding after stopping aspirin can be ascribed to the antithrombotic effect of aspirin itself or its excessive dose (100 mg/day). However, as the daily dose of 100 mg of aspirin may have had a strong antithrombotic effect in a patient with F5D and mucosal bleeding; aspirin may have to be stopped to control mucosal bleeding when simultaneously taking a P2Y12 receptor blocker.
study
99.44
The initial loading dose of aspirin, a P2Y12 receptor blocker, and intravenous heparin were used to prevent thrombosis (particularly stent thrombosis) during coronary stenting and standard maintenance antithrombotic therapy after implanting a DES composed of DAPT (aspirin and a P2Y12 receptor blocker). Generally, the optimal range for the antithrombotic effect is known as mid-third platelet inhibition. F5D was detected after the coronary stenting procedure; thus, the routine DAPT protocol was applied to coronary intervention, and a decrease in HgB was observed that started on the day of coronary stenting. This finding indicates that a weaker antithrombotic effect is recommended for coronary intervention in patients with F5D. In particular, recurrent mucosal bleeding for daily doses of 90 mg of ticagrelor and 100 mg of aspirin suggests that a weakened antithrombotic effect is also required during maintenance therapy after coronary stenting. The disappearance of mucosal bleeding after stopping the 100 mg of aspirin means that ticagrelor may be more effective than aspirin for controlling mucosal bleeding and maintaining antithrombosis in F5D patients after coronary stenting.
study
99.9
Another unusual feature of this case beyond the antithrombotic therapy is ischemic colitis. Unlike mucosal bleeding, such as epistaxis or hemoptysis, ischemic colitis is ascribed to tissue hypoperfusion through the reduced blood supply to the mesenteric artery rather than bleeding tendency or a coagulation disorder. In this patient, tissue hypoperfusion may occur due to transient blood flow reduction, suggesting old age, abdominal aortic calcification, and a sudden drop in the hemoglobin level. Thus, it is unclear whether the ischemic colitis seen here was related to F5D. Furthermore, additional evidence is needed regarding whether antithrombotic therapy presents a risk factor for F5D patients developing ischemic colitis.
clinical case
99.9
Finally, the result of mixing tests in which test plasma is combined with normal plasma was normal, and the absence of inhibitors, including anticoagulants, lupus anticoagulant, or another inhibitor type, was documented. However, we neither performed genetic testing nor investigated autoantibodies for factor V due to the testing cost and the fact that it would provide no additional benefit.
clinical case
99.8
Pancreatic ductal adenocarcinoma (PDAC) is the most common and the deadliest form of pancreatic cancer, comprising 85% of all cases, with a five-year survival rate of just 6.7% [1, 2]. PDAC commonly arises from precancerous lesions called pancreatic intraepithelial neoplasias (PanINs) . These lesions are characterized by very high occurrence of mutations in the KRAS oncogene that are also maintained throughout disease progression and found in over 90% of PDAC cases . These findings indicate that KRAS could be a robust therapeutic target in PDAC. Indeed, murine pancreatic cancers with activated KRAS (e.g. KRASG12D and KRASG12V) exhibit oncogene addiction, whereby suppression of KRAS activity induces cell death in advanced tumors and regression of early PanINs [5–7]. However, efforts to directly inhibit KRAS activity in human tumors have thus far been unsuccessful . Moreover, clinical and preclinical studies have demonstrated the complexities of inhibiting the well-characterized downstream RAF/MEK/ERK and PI3K/AKT pathways [9–14]. These findings highlight the complexity of the signaling networks downstream of activated KRAS and suggest potential roles for post-transcriptional mechanisms that may buffer signaling downstream of KRAS. Therefore, a deeper understanding of the factors that influence KRAS-driven tumor initiation and progression in the pancreas is greatly needed.
review
99.8
MicroRNAs (miRNAs) are highly conserved short non-coding RNAs that influence gene expression post-transcriptionally and regulate development, normal physiology and disease . MiRNAs have been demonstrated to regulate the initiation and progression of many malignancies by controlling oncogenic and tumor suppressive pathways . Among the earliest described oncogenic miRNAs were members of the mir-17~92 cluster [17, 18]. mir-17~92 has been implicated in a variety of cancer contexts , and inhibition of members of this cluster has been shown to impair tumor growth and survival [20, 21]. Profiling of human pancreatic tumors and pancreatic cancer cell lines has shown that miRNAs encoded by the mir-17~92 cluster and its paralogs–mir-106b~25 and mir-106a~363–are upregulated in tumors compared to normal pancreatic tissue or chronic pancreatitis [22, 23]. In addition, mir-17~92-encoded miRNAs are induced in precursor PanIN lesions, implicating them in early stages of PDAC development , and miR-17 overexpression has been associated with reduced pancreatic cancer patient survival . The mir-17~92 cluster was initially identified as oncogenic over a decade ago [18, 26]. Subsequent studies have demonstrated critical roles for this microRNA cluster in several malignancies including B-cell lymphoma, retinoblastoma, medulloblastoma, hepatocellular carcinoma and neuroblastoma [20, 21, 27–34]. Individual miRNAs within the cluster have been associated with specific tumorigenic properties. Of note, the miR-19 microRNAs have been associated with tumor cell invasion and metastasis in gastric cancer , lung cancer and colon cancer .
review
99.9
Studies in pancreatic cancer cell lines additionally demonstrated roles for the mir-17~92 cluster in PDAC cell proliferation, transformation and invasion [25, 38, 39]. However, to date no studies have been performed to evaluate the role of the cluster in vivo during pancreatic tumor initiation and progression. Given the upregulation of these miRNAs in human pancreatic cancers and their validated role as oncogenes in a variety of contexts, we hypothesized that they contribute to KRAS-induced pancreatic tumorigenesis. Therefore, we experimentally tested the requirement for mir-17~92 in a mouse model of pancreatic cancer.
study
99.94
We find that deletion of mir-17~92 impairs MEK/ERK signaling in PanIN lesions and this correlates with the presence of fewer PanINs, as well as their regression over time. In addition, we find that mir-17~92 miRNAs, in particular miR-19 family miRNAs, promote PDAC cell invasion by regulating the formation of extracellular matrix-degrading invadopodia rosettes. Together, these findings illustrate important roles for mir-17~92 miRNAs during multiple phases of PDAC development and progression.
study
100.0
Prior miRNA expression profiling studies of human PDAC specimens demonstrated elevated expression of components of the mir-17~92 cluster in PDAC. However, the results from these studies were somewhat inconsistent, potentially reflecting the significant stromal and immune cell component of pancreatic tumors. To ascertain whether mir-17~92 miRNAs have elevated expression in PDAC cells, we profiled a panel of PDAC cell lines as well as the immortalized pancreatic epithelial cell line HPNE. We find that mir-17~92 miRNAs are consistently overexpressed in PDAC cell lines (Supplementary Figure 1). Thus, we set out to identify the role of this microRNA cluster in pancreatic tumorigenesis in vivo in genetically engineered mouse models.
study
100.0
To determine the effect of mir-17~92 loss on pancreatic development, we induced pancreas-specific deletion of the mir-17~92 cluster using the conditional mir-17~92flox allele and the recombination driver Ptf1a-Cre [40, 41]. qRT-PCR of RNA from whole pancreata showed a strong reduction in the levels of mir-17~92 constituent miRNAs in compound mir-17~92flox/flox, Ptf1a-Cre mice (Supplementary Figure 2A). The observed residual mir-17~92 expression is likely from endocrine cells, many of which are derived from a PTF1A-independent lineage [40, 42], as well as a small population of acinar and ductal cells that have avoided recombination due to the fact that Cre drivers are not 100% efficient . Expression from the paralogous mir-106b~25 cluster is unaffected (Supplementary Figure 2A). Despite efficient depletion of mir-17~92 miRNAs from the pancreas, organ size and exocrine and endocrine architecture and composition were unperturbed (Supplementary Figure 2B). These findings demonstrate that mir-17~92 is not required for normal pancreas development.
study
100.0
To assess the impact of mir-17~92 deletion on the development and progression of precursor precancerous PanIN lesions, we crossed mir-17~92flox/flox, Ptf1a-Cre mice onto the LSL-KrasG12D background . Breeding pairs were designed to cross mir-17~92flox/wt, LSL-KrasG12D mice with mir-17~92flox/wt, Ptf1a-Cre mice in order to generate littermate mir-17~92wt/wt, LSL-KrasG12D, Ptf1a-Cre and mir-17~92flox/flox, LSL-KrasG12D, Ptf1a-Cre mice (hereafter ‘KC’ and ‘17KC’). Littermate KC and 17KC animals were maintained until four or nine months of age and subsequently euthanized to obtain pancreata for histological analysis. At four months of age, the tissue area occupied by PanIN lesions was not significantly different between the two groups as illustrated by hematoxylin and eosin staining and quadchrome staining (Figure 1A–1D, 1I). However, the extent of normal acinar tissue was significantly greater in the pancreata of 17KC mice (Figure 1I). Nine month-old KC mice displayed higher PanIN burdens compared to younger KC mice and nine month-old 17KC littermates (Figure 1E–1I). In contrast to the trend observed in KC mice, we observed that the pancreata of nine month-old 17KC mice exhibited less PanIN tissue and more healthy acinar tissue by area than four month-old 17KC mice (Figure 1F, 1H, 1I). Of note, PanIN lesions in KC and 17KC pancreata demonstrate no significant differences in proliferation or apoptosis markers and no differences were observed in the proliferative rate of adjacent acinar tissue (Supplementary Figure 3). Together, these data suggest that loss of mir-17~92 impairs the maintenance of PanIN lesions.
study
100.0
Histological evaluation of PanIN lesions in pancreata from 4-month old (A–D) and 9-month old (E-H) KC and 17KC mice. Representative images from hematoxylin and eosin (A, B, E, F) and quadchrome (simultaneous Alcian Blue and Sirius Red staining; C, D, G, H) stains are shown. (I) Quantification of normal and neoplastic cell types as a percentage of total tissue area. The quadchrome stain marks collagen red, mucin blue, cytoplasm yellow-brown, and nucleic acids black. Arrowheads identify examples of PanIN lesions. Ac = Acinar tissue. Number of organs (n) analyzed for each group were 4 mo KC (4), 4 mo 17KC (7), 9 mo KC (3), 9 mo 17KC (2). Scale bar = 0.25 mm. p values by Student's t test: * < 0.05, ** < 0.01. Error bars represent standard deviation from the mean.
study
100.0
Recent work suggested that inhibition of MEK/ERK signaling promotes the regression of PanIN lesions . As determined by immunohistochemical staining for phosphorylated ERK, we observed that 17KC PanINs display marked reduction in MAPK signaling (Figure 2A–2D). In contrast to the difference in phosphorylated ERK, phosphorylation of the upstream kinase MEK was not different between the two genotypes, suggesting that signaling through the MEK/ERK cascade is impacted at the level of ERK but not further upstream (Figure 2E–2H). Moreover, ectopic mir-17~92 expression in murine PanIN cell lines [46, 47] increased p-ERK levels under both serum replete and serum starved conditions (Figure 3A, 3B), confirming that mir-17~92-encoded miRNAs regulate this pathway.
study
100.0
(A) Representative immunoblot for pERK and total ERK in the PanIN cell lines RP2294 and AH2375 stably infected with PIG-mir-17~92 or empty vector. (B) Average densitometry for three experiments performed in (A). Error bars represent standard deviation from the mean.
study
99.94
Analysis of the phosphorylation status of AKT at the Thr308 and Ser473 sites that are phosphorylated by PDK1 and mTORC2 also did not show any differences between genotypes (Figure 2I–2P). Taken together with the phosphorylation of MEK, these data suggest that the alterations in ERK phosphorylation do not reflect an overall reduction in KRAS signaling, and further suggest that mir-17~92 regulates PanIN maintenance by specifically influencing ERK pathway activity downstream of KRASG12D and MEK.
study
100.0
We utilized the TargetScan database to identify known and putative mRNA targets of mir-17~92 that are implicated in the regulation of ERK phosphorylation. We identified the dual specificity phosphatases DUSP2, DUSP7 and DUSP10, which are known to regulate MAP kinase phosphorylation, as potential targets of cluster-encoded miRNAs [49, 50]. However, immunostaining of PanIN lesions with antibodies against these phosphatases did not show any differences between mir-17~92 wild type and deficient PanINs (data not shown). Moreover, the enhanced ERK phosphorylation observed in PanIN cell lines following ectopic mir-17~92 expression was not associated with changes in DUSP7 or DUSP10 protein levels (Supplementary Figure 4). DUSP2 was undetectable by immunoblotting or qRT-PCR in these cell lines (data not shown). Thus, the detailed mechanisms regulating ERK phosphorylation in PanINs downstream of mir-17~92 remain unknown.
study
100.0
To assess whether mir-17~92 deletion impairs progression to carcinoma, we accelerated the KC model by including conditional loss of one copy of Trp53 (LSL-KrasG12D, Trp53flox/wt, Ptf1a-Cre and mir-17~92flox/flox, LSL-KrasG12D, Trp53flox/wt, Ptf1a-Cre; hereafter “KPC” and “17KPC” mice). We observed that KPC and 17KPC mice display similar overall survival and tumor size (Figure 4A, 4B). Rates of liver metastasis are also equivalent between groups (Supplementary Figure 5A, 5B). The carcinomas identified in mice of both genotypes displayed a mixture of glandular and undifferentiated histology; no difference in the relative frequencies of the histologic types was identified between the two groups (Supplementary Figure 5C–5E). Histological evidence of invasion was also equally prevalent in both groups, and variously involved the stomach/duodenum, liver, colon, and spleen (Supplementary Figure 5F–5J). The histopathology findings are summarized in Supplementary Table 1. Additionally, tumors across both groups exhibited similar rates of proliferation and apoptosis as demonstrated by Ki67 and cleaved caspase 3 (CC3) staining (Supplementary Figure 6). Moreover, KPC and 17KPC tumors display equivalent MEK/ERK pathway activation as measured by immunostaining for phosphorylated ERK (Supplementary Figure 7A, 7B). Further, cell lines derived from KPC and 17KPC tumors display similar levels of ERK phosphorylation and similar levels of DUSP7 and DUSP10 expression (Supplementary Figure 7C, 7D). Together, these findings suggest that heterozygous deletion of Trp53 compensates for the loss of mir-17~92 and promotes disease progression in mir-17~92 deficient animals.
study
100.0
(A) Kaplan-Meier survival plot for KPC and 17KPC mice. (B) Primary tumor burden identified in these mice upon euthanasia. (C) Kaplan-Meier survival plot comparing metastatic and localized KPC mice. (D) Kaplan-Meier survival plot comparing metastatic and localized 17KPC mice. (E) Kaplan-Meier survival plot comparing localized KPC and localized 17KPC mice. (F) Total primary tumor burden of localized mice. Number of animals (n) for each group were as follows: KPC (18), 17KPC (18), metastatic/localized KPC (5/13), metastatic/localized 17KPC (7/11), female/male KPC (7/11), female/male 17KPC (10/8). p values: * < 0.05. Error bars represent standard deviation from the mean.
study
100.0
In the KPC mouse model, mice reliably generate aggressive primary tumors that invade adjacent tissues and sporadically metastasize: thus, morbidity and euthanasia result from the effects of either the primary tumor or its metastases. To ascertain whether mir-17~92 deletion differentially impacted primary versus metastatic disease processes, we stratified survival data based on the presence or absence of grossly visible metastases at euthanasia. The majority of KPC mice lacked metastases and were sacrificed at relatively young ages due to effects of the primary tumor (“localized” mice), whereas a minority presented later with metastatic disease (“metastatic mice”) (Figure 4C), suggesting that primary KPC tumors cause significant morbidity prior to the development of metastases. However, this was not observed in 17KPC mice, where localized and metastatic mice display similar survival curves (Figure 4D). Indeed, the survival of metastatic mice is not significantly different between KPC and 17KPC mice (data not shown), but the survival of localized KPC mice was significantly worse than that of localized 17KPC mice (Figure 4E). Moreover, localized 17KPC mice also possessed larger tumors than localized KPC mice (Figure 4F). Together, these data suggest that mir-17~92 contributes to the morbidity and mortality caused by primary KPC tumors and does not impact time to metastasis.
study
100.0
The observed differences in survival among animals with localized disease could be the result of reduced or delayed invasive potential of 17KPC primary tumors. Thorough histological step sectioning of all primary tumors and adjacent tissues obtained in the study demonstrated equivalent evidence of invasion between KPC and 17KPC animals at the time of euthanasia (data not shown). This suggests that local invasion is a common endpoint in our survival study, but requires more time to develop in 17KPC mice, resulting in the longer survival and larger tumors of localized 17KPC mice compared to localized KPC mice. Localized pancreatic cancer in mice can cause morbidity with biliary obstruction and jaundice (evident in the ears, footpads, or pancreas) or gastrointestinal (GI) obstruction, as seen by gross abdominal distension and GI lumen distension upstream of an adhesion with the absence of downstream luminal contents on necropsy. While these features were commonly observed in KPC mice, 17KPC mice never presented with any form of GI or biliary obstruction (Supplementary Figure 8). Other characteristics of tumor presentation such as the presence of adhesions to adjacent organs, intraperitoneal bleeding, jaundice or ascites demonstrated no significant differences between KPC and 17KPC mice (Supplementary Figure 8). Together, these in vivo findings suggest that loss of the mir-17~92 cluster may impact PDAC cell invasion, a feature associated with later stages of disease.
study
100.0
To better understand the biology of mir-17~92 deficient pancreatic cancer cells, we generated a collection of cell lines from KPC and 17KPC tumors. Evaluation of cell invasion in transwell assays demonstrated that 17KPC cell lines have reduced ability to invade through Matrigel relative to KPC cell lines (Figure 5A). However, no differences were observed between the two genotypes in their ability to migrate across uncoated membranes (Figure 5B). Additionally, there were no significant differences in the proliferation, anchorage independent growth, or survival phenotypes of KPC and 17KPC cell lines (Figure 5C–5F). These data agree with the suggestion that 17KPC tumors display delayed invasion in vivo, and suggest a specific defect in the ability of tumor cells to manipulate extracellular matrix.
study
100.0
Characterization of in vitro phenotypes of PDAC cell lines derived from KPC and 17KPC tumors. (A) Invasion activity in transwell assays, displayed as invasion index [(number of invading cells/number of migrating cells) × 100]. (B) Migration activity in transwell assays. (C) Proliferation rate. (D) Colony formation in a soft agar. (E) Cell survival in serum-replete medium. (F) Cell survival in serum-free medium. Cell line nomenclature is noted as cage#animal# (e.g. 9025#2 is the cell line derived from the primary tumor of mouse #2 from cage #9025). All error bars represent SD. p value: * < 0.05.
study
100.0
In cancer cells, invasion activity is associated with the presence of specific matrix-degrading, cell adhesion structures called invadopodia [51, 52]. Invadopodia can be identified by the organization of their core cytoskeletal protein components–actin and cortactin–into punctate or rosette-shaped structures that are functionally associated with localized sites of elevated metalloproteinase activity [51, 53–55]. To better understand the nature of the invasive defect that we observed in 17KPC cell lines, we analyzed invadopodia formation by immunofluorescence. We found that invadopodia in KPC and 17KPC cell lines take the form of rosettes and that KPC cell lines display significantly higher rates of invadopodia rosette formation than 17KPC lines (Figure 6A–6I). In agreement with reduced invadopodia numbers, 17KPC cell lines also degrade less FITC-labeled gelatin matrix than KPC lines (Figure 6J–6N). These data suggest that loss of mir-17~92 decreases invasion, at least in part, as a result of reduced matrix-degrading capacity.
study
100.0
Immunofluorescence staining for the invadopodia constituent proteins cortactin (A, E), actin (B, F) and paxillin (C, G) in representative KPC and 17KPC cell lines. Merged images are shown in (D) and (H). D′ and H′ are higher magnification views of panels D and H. Scale bar = 10 um. Quantification of invadopodia rosettes is shown in (I). n = 3 for each cell line. Areas of FITC-gelatin degradation, identified as dark regions, are shown for representative KPC and 17KPC cell lines (J, L). (K, M) Co-staining for actin and DNA (DAPI). (N) Quantification of gelatin degradation. n = 3 for each cell line. Error bars represent standard error of the mean. p values: * < 0.05, ** < 0.01, **** < 0.0001.
study
100.0
The mir-17~92 cluster encodes six miRNAs encompassing four miRNA seed families (Figure 7A), implicating thousands of predicted mRNA targets as downstream effectors of the cluster's invasive program. To aid in our determination of which miRNA families may be most important in the invasive phenotype, we evaluated nine KPC and nine 17KPC cell lines for their expression of miR-17, -18, -19, and -92 family miRNAs across the three cluster paralogs: mir-17~92, mir-106b~25, and mir-106a~363. Quantitative RT-PCR demonstrated that 17KPC cell lines are indeed null for miRNAs from mir-17~92, however they retain robust expression from mir-106b~25 (Figure 7B, 7C). In fact, the mir-106b~25 locus is sufficient to drive expression of miRNAs for the miR-17 and miR-92 families to levels close to those observed in KPC lines, suggesting that loss of the miR-17 and -92 families may not be primarily responsible for the invasive defect of 17KPC cell lines. In contrast, miR-19 family miRNAs can only be expressed from the mir-17~92 and mir-106a~363 clusters, and 17KPC lines were found to completely lack expression of this miRNA family (Figure 7E). Based on the partial residual expression of the miR-17 and miR-92 families (Figure 7B, 7C), and the generally very low expression of the miR-18 family (Figure 7D, note y axis units), we hypothesized that loss of the miR-19 family was responsible for the defective invasion of 17KPC cell lines.
study
100.0
(A) Schematic representation of the mir-17~92 cluster and its paralogs mir-106b~25 and mir-106a~363. Constituent miRNAs are color-coded according to their seed families. (B–E) Quantitative RT-PCR measurement of mature miRNA expression across eighteen cell lines derived from primary KPC and 17KPC tumors. Some miRNAs share sufficient sequence similarity that standard oligonucleotides amplify both species equally (e.g. miR-19a and miR-19b), and therefore not all miRNAs are individually plotted. p values: **** < 0.0001. Error bars represent standard deviation from the mean.
study
100.0
To validate that miR-19 family miRNAs play an important role in invasion, we utilized antagomirs–short oligonucleotides that bind and inactivate miRNAs–to specifically knock down miR-19 function in KPC lines with high invasive capacity and varying levels of miR-19 expression . We first confirmed miR-19 antagomir activity using a β-galactosidase (β-Gal) reporter containing multiple miR-19 binding sites within the 3′UTR that allow translational suppression in the presence of miR-19. Pooled antagomirs against miR-19a and miR-19b enhanced reporter activity in a KPC cell line, but not in a 17KPC cell line (Supplementary Figure 9).
study
100.0
MiR-19 antagomirs significantly inhibited KPC cell line invasion, but not migration, consistent with an invasion-specific effect for miR-19 family miRNAs (Figure 8A, 8B). This response inversely correlated with the level of endogenous miR-19 family expression, suggesting a dosage response (Figure 8C). Indeed, the cell line with the highest expression of miR-19, 9415#2, was resistant to antagomirs at a concentration of 50 nM, but responded when treated with antagomirs at 100 nM (Figure 8A, 8C). We next ascertained whether miR-19 regulates invasion in human PDAC cells. The human pancreatic cancer cell lines MIA Paca-2 and PANC-1 are also invasive and express relatively high levels of miR-19 (Figure 8F). Treatment of these lines with miR-19 antagomirs reduced their invasive capacity in a transwell assay without affecting their migration ability (Figure 8D, 8E). Furthermore, treatment with miR-19 antagomirs was sufficient to reduce the number of invadopodia rosettes formed in KPC cells (Figure 8G, Supplementary Figure 10) and also decreased the gelatin degradation capacity of these cell lines (Figure 8H, Supplementary Figure 11). These data demonstrate that miR-19 miRNAs regulate PDAC cell invasion.
study
100.0
(A) Transwell invasion through Matrigel of KPC cell lines treated with control or miR-19-targeting antagomirs at the indicated concentrations. (B) Transwell migration activity of the cell lines shown in (A). (C) Quantitative RT-PCR measurement of baseline mature miR-19 levels in KPC cell lines. (D) Transwell invasion through Matrigel of the human PDAC cell lines MIA Paca-2 and PANC-1 treated with control or miR-19-targeting antagomirs at the indicated concentrations. (E) Transwell migration of activity of the cell lines shown in (D). (F) Quantitative RT-PCR measurement of baseline mature miR-19 levels in MIA Paca-2 and PANC-1 cells; miR-19 levels in the KPC cell line 9415#2 are shown for comparison. Invadopodia rosette formation (G) and FITC-gelatin degradation (H) in the KPC cell lines 9910#1 and 9248#1 treated with control or miR-19-targeting antagomirs. n = 3 for each cell line. p values: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001. (A–F) Error bars represent standard deviation from the mean. (G, H) Error bars represent standard error of the mean.
study
100.0
Given its rate of mutational activation in PDAC and past observations of oncogene addiction, KRAS and components of its downstream signaling pathways represent robust therapeutic targets in this devastating malignancy. To date, efforts to directly inhibit KRAS function have been unsuccessful ; thus efforts in the field have turned to targeting key downstream pathways [9–11]. However, a deeper understanding of the mechanisms responsible for the transforming effect of KRAS could inform more effective therapeutic strategies. Several microRNAs, including those in the mir-17~92 cluster, display increased expression in PDAC as well as precursor PanIN lesions [22–24], suggesting that these miRNAs may play a role in tumorigenesis. However, to date, the functional importance of these miRNAs has not been evaluated in vivo.
review
99.9
Here we report that deletion of the mir-17~92 miRNA cluster results in the regression of KRASG12D-driven PanIN lesions and the expansion of normal acinar tissue in place of neoplastic cells over time. In addition, we observe that mir-17~92-null PanIN lesions have reduced ERK phosphorylation, and PanIN cell lines with ectopic mir-17~92 expression display elevated p-ERK levels. However, mir-17~92-null PanIN lesions display no changes in MEK phosphorylation or PI3K/AKT signaling, suggesting a specific impact of mir-17~92 on ERK activation. Importantly, apoptotic rates are not increased in mir-17~92-null PanINs, suggesting that these lesions are not lost by apoptosis. Moreover, the proliferation rate is unchanged in the adjacent acinar tissue of 17KC mice compared to that of KC mice, indicating that exocrine recovery is not due to the expansion of residual acinar cells in these animals. Together, these findings suggest that mir-17~92 loss promotes the redifferentiation of PanINs into acinar cells. While sophisticated lineage tracing experiments will be required to validate this hypothesis, our data are in agreement with recently published data demonstrating that ERK pathway activity downstream of KRASG12D is critical for PanIN maintenance, and loss of ERK signaling triggers PanIN regression into normally differentiated exocrine tissue . Importantly, the effects of miRNAs on signaling pathway output are generally smaller in magnitude than those observed with small molecule inhibitors, which potentially explains the extended timeline of PanIN regression in 17KC mice compared to the rapid regression described by Collins, et al. using MEK inhibitors .
study
100.0
The mechanisms underlying the regulation of ERK activity by mir-17~92 are unknown. We explored the possibility that mir-17~92 increases ERK activation via suppression of dual-specificity phosphatases (DUSPs), which are well-known regulators of numerous MAPK family proteins, including ERK . In particular, DUSP2, DUSP7 and DUSP10 suppress ERK activity and are demonstrated or predicted targets of mir-17~92 miRNAs [48, 58–60]. Interestingly, a recent publication linked miR-92 and DUSP10 to PDAC cell proliferation in vitro, suggesting that there may indeed be an important role for this regulatory axis in pancreatic tumorigenesis . However, immunohistochemical staining of DUSP2, DUSP7 and DUSP10 failed to demonstrate any difference between KC and 17KC PanIN lesions, and modulation of mir-17~92 levels in PanIN cell lines also failed to change the levels of these phosphatases. Thus, the precise mechanisms through which mir-17~92 regulates ERK phosphorylation remain unknown. Additional studies using inducible expression or repression of mir-17~92 miRNAs, coupled to mRNA and proteomic profiling approaches, will aid in elucidating the mechanisms by which this cluster regulates ERK phosphorylation in the early stages of pancreatic tumorigenesis.
study
100.0
We observed similar overall survival, rates of metastasis, and histological prevalence of invasion at sacrifice between KPC and 17KPC mice, suggesting that Trp53 loss can compensate for mir-17~92 deletion. In the KPC model, mice typically develop aggressive PDAC, characterized by local invasion, obstructive symptoms involving the gastrointestinal or biliary systems, and metastasis, which we observe in the KPC animals of this study. In contrast, we find that 17KPC mice exhibit less aggressive primary disease, as demonstrated by longer survival with larger primary tumors in the absence of metastases, and the absence of obstructive gastrointestinal or biliary symptoms. These data suggest that mir-17~92 plays a role in PDAC invasiveness. However, we cannot exclude the possibility that the absence of obstructive disease in the 17KPC mice reflects a difference in the initial anatomical site of these primary tumors compared to KPC tumors. Perhaps the tumors in 17KPC mice arise in the tail of the pancreas, allowing them to reach greater size before obstructing the duodenum or bile duct; whereas tumors occurring in KPC animals predominantly occur in the head of the pancreas, predisposing those animals to early obstructive phenomena. Additional studies analyzing tumors at earlier stages and smaller sizes will be needed in order to clarify whether KPC and 17KPC tumors arise in different locations within the pancreas.
study
100.0
Analysis of a panel of tumor-derived cell lines demonstrated that 17KPC cells are specifically defective in their ability to form invadopodia rosettes and invade through Matrigel in a transwell assay, providing a potential link to our in vivo observations. Using antagomirs against miR-19, we demonstrated that miR-19 family miRNAs are key drivers of invadopodia formation and the invasive capacity of human and murine pancreatic cancer cells. Thus, we have identified a novel role for miR-19 in pancreatic cancer cells. In fact, few studies exist that link miR-19 to cancer cell invasion in any tumor type [36, 37]. Future studies will be required to validate the significance of miR-19 family miRNAs in the invasive phenotype of PDAC in vivo. In addition, expression profiling and proteomic studies will be required to identify the mechanisms through which miR-19 miRNAs regulate invadopodia formation and function. TIMP2, CST3, and TGM2 are all predicted targets of miR-19 that are also suppressors of invasion [37, 39, 62, 63] and could potentially explain the reduced invasiveness of 17KPC cell lines. Indeed, TGM2 is linked to miR-19-mediated invasion in colorectal cancer cells . However, immunoblotting and qRT-PCR experiments failed to detect any difference in expression between KPC and 17KPC cell lines (data not shown). Matrix metalloproteinases are also key drivers of invasion , and a survey of MMP-2, -7, -9, and MT1-MMP by qRT-PCR demonstrated no difference between KPC and 17KPC cell lines (data not shown). Given that our data point to a major effect of miR-19 on invadopodia formation and/or stability, and since none of the above factors are known to influence invadopodia, these negative findings are perhaps not surprising. Instead, they highlight the importance of the unbiased approaches stated above for elucidating the mechanisms responsible for the observed phenotypes.
study
99.94
Importantly, our experiments do not preclude significant roles for the other miRNAs encoded within the mir-17~92 cluster in PDAC invasion. Indeed, other members of the cluster are computationally predicted to regulate genes previously implicated in cellular invasion. Thus, studies that confirm or exclude roles for these miRNAs in PDAC invasion will also be of importance.
study
100.0
Our studies reported here are potentially in conflict with the recently published work of Heeschen and colleagues [65, 66]. In their work, the authors report that ectopic mir-17~92 expression promotes the proliferation of pancreatic cancer stem cells, resulting in their premature depletion and consequent reduced cell transformation and tumorigenicity. In contrast, our studies demonstrate that ectopic mir-17~92 increases MEK/ERK signaling, a feature required for the maintenance of established PanIN lesions [6, 67], suggesting that elevated mir-17~92 levels should enhance pancreatic cancer development. However, we have not evaluated the self-renewal capacity of mir-17~92 expressing PanIN cell lines, nor have we tested their tumorigenic capacity upon implantation into recipient mice. Future experiments in genetically engineered PDAC mouse models with enhanced mir-17~92 expression may be required to resolve this discrepancy.
study
99.94
Together, our findings demonstrate important roles for mir-17~92-encoded miRNAs during early stages of pancreatic tumorigenesis, as well as tumor progression and invasion. These findings provide functional support for the observed elevated expression of these miRNAs in precursor PanIN lesions and invasive PDAC. Further dissection of the cluster to identify the roles played by individual miRNAs during PanIN maintenance and PDAC invasion will be required in order to identify the critical target genes and pathways regulated by this miRNA cluster during pancreatic tumorigenesis.
study
100.0
The Ptf1a-Cre , mir-17~92flox , LSL-KrasG12D , and Trp53flox mouse strains have been described previously. Health status of all animals was monitored at least three times per week, and animals were euthanized when they displayed signs of distress or high tumor burden. Animals were maintained in specific pathogen-free facilities with abundant food and water. Mice were randomly assigned to the studies. Group sizes were estimated based on investigators’ prior experience. Euthanized mice that did not have pancreatic tumors were censored from the analysis. Mice of both genders were used in all studies. The pathologist was blinded to mouse genotypes for quantification of histopathologic lesions. All animal experiments were reviewed and approved by the University of Massachusetts Medical School Institutional Animal Care and Use Committee.
study
99.94
Tissues were fixed in 10% neutral-buffered formalin. Five-micrometer sections on charged glass slides were cleared through Xylenes (Fisher #X3P) to 100% ethanol and rehydrated through a graded alcohol series to distilled water. Immunostaining was performed as previously described . A list of all antibodies used for immunohistochemical stains appears in Supplementary Table 2. Stains were developed using ABC (Vector Labs #PK-6101) and Nova Red (Vector Labs #SK-4800) kits according to manufacturer's instructions.
other
99.9
A licensed pathologist who was blinded to tissue genotypes performed quantification of tissue areas. Four images were quantified per section representing 1) the area of greatest neoplastic progression, 2) the area of lowest neoplastic progression, and 3,4) areas of the pancreas that were consistent with the average progression for that tissue section. Acini, ducts, PanINs, ADM lesions, and stromal tissue not including blood vessels were manually outlined using ImageJ software and are graphed as the percentage area of all areas quantified.
study
99.94
All cell lines were maintained at subconfluent densities in high glucose DMEM (Life Technologies #11965) supplemented with 10% fetal bovine serum (Atlanta Biologicals #S11150) and 100U/ml penicillin/streptomycin (Pen/Strep: Life Technologies #15140) (herein ‘complete media’). Murine PDAC cell lines were generated in the Lewis lab. Murine PanIN cell lines were obtained from Nabeel Bardeesy (Massachusetts General Hospital). Human PDAC cell lines were obtained from ATCC, except for the Pa01c through Pa18C cell lines, which were obtained from Bert Vogelstein (Johns Hopkins University) . Migration and invasion assays were performed as previously described using 8-micrometer-porous transwell inserts (Fisher #08-774-162) and Matrigel-coated inserts (Fisher #08-774-122) . The invasion index was calculated as previously described . Soft agar colony formation assays were performed as previously described . Proliferation assays of adherent, subconfluent cells were performed by direct live cell counting after trypsin-mediated resuspension using trypan blue exclusion over a period of 48 hours. Counts were plotted as the log2 of the cell number over time, and the proliferative rate of each line was calculated as the inverse slope of a linear regression to the data (i.e. time/doubling). Data shown are the average of greater than four experiments. P-values were calculated using the Student's t-test.
study
100.0
For antagomir experiments, miRCURY LNA Power Inhibitors (Exiqon #4101004-100, #4103258-100, and #199006-100) were transfected into 2.5 × 105 cells per well of a 6-well plate using Superfect (Qiagen #301305) according to manufacturer's specifications at a final antagomir concentration of either 50 or 100 nM.
study
99.94
Invadopodia rosette formation analysis and FITC-gelatin matrix degradation assays were performed as previously described . Briefly, to visualize invadopodia rosette formation, cells were plated on fibronectin-coated slips (10μg/ml; Corning) for 24hrs, fixed, permeabilized and stained with anti-cortactin antibodies (1:200; Merck Millipore #05-180), anti-paxillin antibodies (1:200; Santa Cruz Biotechnology #sc-5574), TRITC-phalloidin (F-actin)(1:1000; Invitrogen #R415) and DAPI (nuclei) (1:1000; Sigma-Aldrich #D9542). A minimum of 150 cells per sample were scored for rosette formation, which were defined by cortactin and actin colocalization. n = 3.
study
99.94
For FITC-gelatin degradation analysis, coverslips were coated with 50μg/ml poly-L-lysine (Sigma-Aldrich #P8920) in PBS, and then incubated with 0.5% glutaraldehyde (Sigma-Aldrich #G6257) in PBS. The slips were coated with 1:40 fluorescent 488 gelatin (‘FITC-gelatin’) (Life Technologies #G13186) diluted with 0.2% unlabeled gelatin (Sigma-Aldrich) in PBS for 30 min. at 37°C. Cells were plated for 24 hours, fixed and stained as above. For quantitation of matrix degradation, nine random fields (20x objective) were imaged from each sample (n = 3). ImageJ software was used to threshold the cell area as well as areas of matrix degradation (black areas in FITC-gelatin) and to calculate a ratio of matrix degradation to cell area. Statistical analyses were performed using a one-way ANOVA with a Tukey correction.
study
100.0
Quantitative RT-PCR for miRNAs was performed as previously described . After RNA isolation with TRIzol reagent (Invitrogen #15596), genomic contaminants were removed using DNAse (Life Technologies #AM1907). DNA-free RNA was then polyadenylated (New England Biolabs #M0276) and subsequently reverse transcribed (Invitrogen #18080) using a pool of specially designed primers at a concentration of 50 uM to generate cDNA copies of polyadenylated miRNAs. A complete list of primer sequences appears in Supplementary Table 3. All subsequent steps of the cDNA synthesis were conducted according to kit directions.
study
99.9
PCR for miRNAs was performed as follows: 1) denaturation at 94°C for 15 seconds, 2) annealing at 55°C for 30 seconds, and 3) extension at 70°C for 34 seconds, for a total of 40 cycles. The reverse primer for all miRNA reactions was the sequence of the universal tag present in all 12 RT primers. The forward primer for each miRNA was the mature miRNA sequence given by miRBase . PCR reactions were carried out on an ABI Step One Plus machine in 10ul volumes using SYBR Green (VWR #95072).
other
99.3
CT values were calculated for all miRNA PCR reactions at a uniform threshold of absorbance across all experiments and controlled to CT values for the endogenous reference (snoRNA234 in mouse-only experiments, U6 in human and cross-species comparisons). The expression of individual miRNAs is presented as relative snoRNA234 units in order to convert the ΔCT value for each miRNA into a relative molar measure (calculated as 2ΔCT).
study
100.0
Two dimensional artificial honeycomb lattice provides a facile platform to explore many novel properties of magnetic materials in one system.1, 2 It includes the ice analog of magnetism, spin ice, spin liquid, and an unusual spin solid state, depicted by the distribution of magnetic vortex loops of opposite chiralities.3, 4 The complex variety of entropy controlled magnetic phases that are predicted to arise in an artificial honeycomb lattice as a function of reducing temperature cannot be realized in a “3D” bulk material of geometrically frustrated origin. According to recent theoretical reports, the honeycomb lattice behaves as a paramagnet at high temperature, corresponding to a gas of ±1 and ±3 magnetic charges.5, 6 As temperature is reduced, the system crosses over into a spin‐ice type state, manifested by “two‐in & one‐out” (or vice‐versa) configuration where two of the moments along the honeycomb element point to the vertex and one points away from the vertex (or vice‐versa). At further reduction in temperature, a new ordering regime, characterized by the topological “charge order” of ±1 magnetic charges, develops (depicted in Figure 3). Thermal energy is expected to be comparable to the strength of the dipolar interaction (≈D) in the charge ordered regime. The transition from a local spin ice to the charge ordered state is chiral in nature, as a number of mobile closed loops of each chirality develop.3, 5, 6 At much lower temperature, the system is predicted to evolve into a “spin‐ordered” state of chiral vortex loops with zero entropy density, also called the “spin solid” state. It represents a novel phase of magnetic matter with zero entropy and magnetization.7
study
100.0
In this letter, we report the investigation of the spin solid state in artificial honeycomb lattice of connected ultrasmall elements. The typical dimension of a connecting element is 12 nm (length) × 5 nm (width) × 10 nm (thickness). Detailed measurements of polarized neutron reflectometry and glancing incidence small angle neutron scattering have revealed the development of additional scattering of magnetic origin from in‐plane correlations when cooling down to T = 7 K. The diffuse band‐type scattering is reasonably well described by the numerical modeling of the spin solid state configuration where magnetic moments along the connecting permalloy elements of the honeycomb lattice manifest an alternating order of the vortex loops of opposite chiralities. Development of the spin solid state is independently confirmed by the temperature‐dependent micromagnetic simulations that show a temperature‐dependent evolution of the underlying spin correlation in an artificial honeycomb lattice with similar size of connecting elements. The system tends to attend this novel order as T → 0 K.
study
100.0
The current experimental efforts in accessing the spin solid state in artificial honeycomb lattice is mainly based on the employment of the electron‐beam lithography method for sample fabrication, which generally leads to small sample sizes with large elemental parameters. Such large element size honeycomb lattice typically manifests an inter‐elemental interaction energies of ≈104 K.1, 8, 9, 10 More recently, a new design of artificial honeycomb lattice, consisting of very thin (few angstroms) and well‐separated permalloy elements of large size (≈500 nm in length and 20–50 nm in width) that reduces the interelemental energy significantly, was proposed to access the spin solid state in the disconnected geometry.11, 12 In the disconnected honeycomb lattice, the underlying physics is dictated by the Block transition mechanism. On the other hand, the physical connection between the lattice elements in connected honeycomb lattice facilitates the propagation of domain wall from one vertex to another, which is necessary for inducing the spin flip and, hence, the continuous progression of a new magnetic phase as temperature is reduced.8, 13, 14 Here, we take a different approach and create large throughput new artificial honeycomb lattice of connected ultrasmall elements. The ultrasmall size of the connecting elements automatically reduces the interelemental magnetostatic energy between 12 and 15 K. Therefore, it is quite suitable for the investigation of temperature‐dependent evolution of magnetic phases using reciprocal space probes, such as neutron scattering method.
study
100.0
The fabrication of artificial honeycomb lattice samples involved the synthesis of hexagonal diblock copolymer templates and near parallel deposition of permalloy material on top of the honeycomb structured silicon substrates in an ultrahigh vacuum chamber. Similar diblock copolymer templates are extensively used to fabricate nanostructured materials.15 Under suitable physical conditions, a diblock copolymer tends to self‐assemble where one component tends to develop long‐range periodic structures.16, 17 Additionally, the flexibility in tuning the structural properties and lattice parameters, by simply varying the composition and molecular weight of the diblock copolymer, allow to create a plethora of nanomaterials.18, 19 Some of the notable examples include the fabrication of nanodot, nanoring, and nanoparticle assemblies.15, 17, 19, 20 More recently, researchers have used diblock templates and glancing angle deposition to create directional hierarchical structures of metal nanoparticles.17 An atomic force microscopy image of a typical honeycomb sample is shown in Figure 1 a (see the Experimental Section for detail). A small angle X‐ray scattering measurement in the grazing incidence angle configuration (GISAXS) confirms the high structural quality of the sample. GISAXS measurements can provide information about the structural properties of a system.21, 22, 23 The GISAXS measurements were performed using a Ga Kα source with a wavelength of 1.34 Å at an incident angle of 0.15°. A 1 mm thick stainless steel foil was used to attenuate the reflected beam. As shown in Figure 1b, GISAXS measurements show a primary spacing of 31 nm, which is consistent with the atomic force microscopy image within the calibration error. The second and third peaks visible in the scattering pattern occur at multiples of 3 and 2 of the primary peak, corresponding to a 2D hexagonal lattice. The higher order peaks seem to be overshadowed by the background in the data, which is most likely arising due to the possible inhomogeneity in the sample. The dimension of the constituting element of the honeycomb lattice is not perfect, rather varies a little bit from the average size of 12 nm in length and 5 nm in width. However, the small variation in the element size is not expected to affect the underlying physics much, as the interelemental energy will only change marginally (less than 2 K for a variation of up to 2 nm). From the modeling of the GISAXS data, the large domain size of long‐range structural order in the honeycomb lattice was confirmed (paracrystal correlation length of 250 nm).
study
99.94
Structural characterization of artificial honeycomb lattice. a) Full size atomic force microscopy image of typical artificial honeycomb lattice, derived from diblock porous template combined with reactive ion etching (see the text for detail). The bond length, width, and lattice separation are ≈12, 5, and 31 nm, respectively. b) Grazing incident X‐ray scattering recorded with an incidence angle of 0.15° using Ga Kα. 2D plots, as shown below, are horizontal and vertical integrations of the areas marked as red and green boxes in the image. Numerical simulations, using the same structural parameters as for the neutron models (discussed below), are shown in the same graph for comparison and describe the main features and their positions accurately.
study
100.0
To directly investigate the correlation between magnetic moments along the honeycomb elements, we have performed polarized neutron experiments, namely reflectometry, off‐specular and grazing incidence neutron scattering (GISANS) measurements.21, 24, 25, 26 Together, they allow us to explore the underlying magnetic correlations27 at a length scale of ≈5 nm to 10 μm in the honeycomb lattice. In Figure 2 , we plot the off‐specular intensity measured using spin‐up (+) and spin‐down (−) neutrons at T = 300 K and 7 K. Here, the y‐axis represents the out‐of‐plane scattering vector Qz = 2πλ(sinαi+ sinαf) while, the difference between the z‐components of the incident and the outgoing wave vectors pi − pf = 2πλ (sinαi − sinαf) is drawn along the x‐axis. Thus, vertical and horizontal directions correspond to the out‐of‐plane and in‐plane correlations, respectively (for detail information, see Lauter et al.28). The specular reflectivity lies along the x = 0 line. Clearly, the off‐specular scattering plots show remarkable differences between high and low temperature measurements. At T = 300 K, the specular intensity is more than two orders of magnitude stronger compared to the off‐specular data, which is the expected behavior for most systems. There is also some scattering in the off‐specular regions caused by the paramagnetic nature of the moment and the honeycomb structure itself. The difference between the spin‐up and the spin‐down components in the off‐specular reflections is similar to that observed in the specular data (see Figure S1, Supporting Information). It indicates a paramagnetic or weak ferromagnetic character (MSLD = 5.9 × 10−6 Å−2 ↔ M = 1.92 × 105 A m−1) of the honeycomb lattice at room temperature. Upon cooling the sample to T = 7 K, the off‐specular signal increases significantly (notice the logarithmic color scale). No specular beam can be distinguished from the off‐specular background, anymore. As the nuclear structure factor will not change noticeably upon cooling, this can only be explained by a significant change in the magnetic characteristics of the system.
study
100.0
Polarized neutron scattering measurement on artificial honeycomb lattice. a–d) Off‐specular neutron reflectometry simulated in DWBA for the various models shown as sum of all neutron spin states. Starting with a) a paramagnetic spin‐gas state, the correlations of spins (see Figure 3) increase up to d) the spin solid state. The corresponding polarized neutron data recorded at e) T 300 K and f) 7 K are shown as sum of spin‐up and spin‐down components. In all graphs, the x‐ and y‐axes indicate the difference (p i − p f) and sum (Q z) of the out‐of‐plane components of the incident and outgoing neutron wave vectors that correspond to lateral and depth correlations, respectively. The specular reflection at room temperature, indicating the paramagnetic state (e), is replaced by a broad diffuse scattering, indicating correlation due to ≈100 nm structure size (f). Unlike the paramagnetic case, the simulated spin solid state exhibits strong scattering along the horizontal axis of similar magnitude as the specular reflectivity. The simulated profiles well describe the experimental results. The steps in the data represent change in the incidence angle, typical of a time of flight (TOF) measurement.
study
100.0
The broad feature along the horizontal axis in the neutron reflectometry data at T = 7 K indicates the development of in‐plane magnetic correlations in the artificial honeycomb lattice (Figure 2). In order to further understand the underlying magnetic structure, experimental data is compared to the numerically simulated patterns for theoretically predicted magnetic phases, namely paramagnetic state as well as spin ice (ice‐1), charge ordered configuration (ice‐2) and spin solid states, all depicted in Figure 3 . The different magnetic states were simulated in our Distorted Wave Born Approximation (DWBA) model (see the Experimental Section for more information).29, 30 As shown in the lower panel of Figure 2, the off‐specular scattering increases with the amount of spin–spin correlation in the system. Although the difference between ice‐2 (magnetic charge ordered state) and spin solid is small, a reasonable agreement with the experimental data is obtained for the spin solid state. The spin solid state is modeled by arranging the vortex loops of opposite chiralities in an alternating order, as described above. The simulated patterns indeed exhibit broad bands of diffuse scattering along the x‐axis, as observed in the experimental data (also see the Supporting Information for detail about the modeling). We note that the interelemental energy in the newly designed artificial honeycomb lattice is ≈12 K. This is the characteristic energy necessary for a magnetic charge ordered state to develop, followed by the spin solid state as T → 0 K. Therefore, the observed increase in intensity is consistent with the expected behavior in our sample.
study
100.0
Magnetic correlation models used for the DWBA simulations. While in all models the number of spins on the lattice is equal to the number of honeycomb edges, the correlations are limited to a) single spins for the paramagnetic, b) spin triangles with 2‐in & 1‐out or 1‐in & 2‐out arrangements for ice‐1, c) spin vortices of either left or right chiralities for ice‐2, and d) long range ordered arrangement of chiral vortex loops for the spin solid state. Pink and yellow balls at the vertex represent ±1 unit of magnetic charge.
study
99.94
To further understand the magnetic behavior at low temperature, we have also performed GISANS experiments with limited intensity available, as the beamline is not optimized for this method. The signal‐to‐background ratio is not large enough to observe distinguishable features in the reciprocal space map (see Figure S2, Supporting Information). We have performed integrations along the Q y axis using a band from Q z = 0.025 Å−1−0.045 Å−1 for the models depicted in Figure 3 and the measured data, which are shown in Figure 4 . The 300 K data have a clear peak around Q y = 0.02 Å−1, which corresponds to the nuclear structure as well as the total scattering in the gas and ice‐1 states. Upon cooling down to 7 K, additional intensity develops around Q y ≈ 0.012 Å−1 that is only expected for the ice‐2 and/or the spin solid states (also see Figures S3–S5, Supporting Information for detail simulations). For the sample to manifest an Ice‐2 state, a finite intensity at Q y ≈ 0.025 Å−1 should have been observed. Rather, it is missing in the experimental data. The observed intensity profile, albeit limited, is reasonably consistent with the predicted q‐dependence of intensity in the spin solid state or, a phase mixture of spin solid state (majority phase) with ice states (minority phase).
study
100.0
Cuts through the simulated (top) and measured (bottom) GISANS maps on a range from Q z = 0.025 Å−1 to 0.045 Å−1. There is an increase in intensity around Q y ≈ 0.012 Å−1 that corresponds well to the simulated increase in the spin‐solid state. While there is an increase in the ice‐2 state for this region as well, the additional intensity at 0.03 Å−1 is not observed in the experiment. The initial 300 K state could well be a mixture of spin‐gas and ice‐1 state.
study
100.0
Next, we have performed temperature‐dependent micromagnetic simulations on an artificial honeycomb lattice of similar element size and thickness to independently verify the development of a spin solid state at low temperature. The honeycomb lattice was simulated using 0.2 nm × 0.2 nm mess size on the μMAG platform,31, 32 with magnetic field applied in‐plane to the lattice. The simulation utilizes the Landau–Lifshitz–Gilbert equation of magnetization relaxation in a damped medium. It is given by33, 34, 35 (1)dmdt = − γm × heff+ αm × dmdt,where γ is the gyromagnetic ratio and α is damping constant. In the above equation, thermal fluctuation is introduced by Langevin dynamics.36 The Langevin dynamics utilizes the concept of white noise of Gaussian form, Γ(t) with a mean at zero, arising due to the thermal fluctuation. Accordingly, each site experiences a white noise as temperature increases. At each time step, the instantaneous thermal field on each site is given by, htherm = Γ(t) 2kBTαγμsΔt, where μs is magnetic moment of simulation element i. The thermal field adds to the original effective field via h(T > 0)eff = h(T = 0)eff + h therm, where h eff(T = 0) = −δH/δm. The Hamiltonian, H, of the system consists of four terms: exchange energy, uniaxial anisotropic energy, magnetostatic energy and the Zeeman energy. For the simulation, we have used exchange stiffness A = 1.0 × 10−11 J m−1, uniaxial anisotropy strength K 1 = −1.0 × 103 J m−3, and damping constant α = 0.2. The simulated hysteresis curves at various temperatures are shown in Figure 5 . Qualitative differences between magnetic hysteresis curves at different temperatures are clearly observed in this figure. The plot of M/M s versus h, where M s is saturation magnetization, depicts a more subtle transition in applied field as temperature reduces. At low temperature (simulation at T = 0 K), magnetic hysteresis not only manifests a sharp transition near zero field but also develops a small plateau near M = 0. The simulated magnetization profile in this state is characterized by the vortex configuration, which is the key element of the spin solid state. Unlike the development of vortex state near zero field at low temperature, the magnetization profile at T = 100 K depicts short‐ranged ordered spin ice state (see the color map). The micromagnetic simulations further confirm the development of spin solid state.
study
100.0
Temperature‐dependent micromagnetic simulation. Micromagnetic simulations at T = 0 , 100, 200, and 300 K show qualitative differences in magnetic hysteresis curves. A small plateau near zero field and magnetization at T = 0 K is identified with a magnetic configuration of chiral vortex loops, tending to form a spin solid state in connected honeycomb lattice. The simulated profile near zero field at T = 100 K exhibits spin‐ice configuration. The chiral vortex loop disappears as temperature increases.
study
100.0
In summary, we have presented experimental investigation of magnetic correlations at low temperature in newly designed artificial honeycomb lattice. The experimental results were independently verified by the temperature‐dependent micromagnetic simulations. Among the various magnetic phases that are expected to arise as a function of reducing temperature in artificial honeycomb lattice, magnetic charge ordered state and spin solid state hold greater significances.37 Both states are somewhat unique to this 2D structure that involve chiral vortex loops. While the charge ordered state is expected to develop below magnetostatic dipolar interaction temperature, given by D ≈ k BT, the spin solid state arises as T → 0 K. Based on our experimental and theoretical researches, we infer that the magnetic moments along the honeycomb element in the newly fabricated honeycomb lattice tend to develop the spin solid state, compared to magnetic charge ordered state (ice‐2 phase), as temperature reduces below the interelemental energy, T ≈ 12 K. Our results also suggest the highly competing nature of novel magnetic phases in artificial honeycomb lattice of connected elements. Future experimental researches, such as the estimation of entropy per element, are desirable to further understand the development of the spin solid state6, 7 A real time imaging technique, such as Lorentz microscopy, can provide direct evidence of spin solid state in this system. Future efforts in this regard are specially desirable.
study
100.0
Sample Fabrication and Characterizations: Fabrication of artificial honeycomb lattice involves the synthesis of porous hexagonal diblock template on top of a silicon substrate, calibrated reactive ion etching using CF4 gas to transfer the hexagonal pattern to the underlying silicon substrate, and the deposition of magnetic material (permalloy) on top of the uniformly rotating substrate in near‐parallel configuration (≈1°) to achieve the 2D character of the system. The sample fabrication process utilized diblock copolymer polystyrene(PS)‐b‐poly‐4‐vinyl pyridine (P4VP) of molecular weight 23K Dalton with the volume fraction of 70% PS and 30% P4VP. The self‐assembly of diblock copolymer was driven by microphase separation arising from the immiscibility of the polymer blocks. A microphase separated diblock copolymer film can take various forms from spherical to cylindrical to lamellar, depending upon the volume fraction of each block. At this volume fraction, the diblock copolymer tends to self‐assemble, under right condition, in a hexagonal cylindrical structure of P4VP in the matrix PS.38 A 0.5% PS‐b‐P4VP copolymer solution in toluene was spin casted onto cleaned silicon wafers at 2500 rpm for 30 s and placed in vacuum for 12 h to dry. The samples were solvent annealed at 25 °C for 12 h in a mixture of THF/toluene (80:20 v/v) environment. The process results in the self‐assembly of P4VP cylinders in a hexagonal pattern within a PS matrix. The average diameter of a P4VP cylinder was ≈ 12 nm and the center‐to‐center distance between two cylinders was ≈ 30 nm, also consistent with that reported by Park et al.38 Submerging the samples in ethanol for 20 min releases the P4VP cylinders yielding a porous hexagonal template. The diblock template was used as a mask to transfer the topographical pattern to the underlying silicon substrate. The top surface of the reactively etched silicon substrate resembles a honeycomb lattice pattern. This property is exploited to create metallic honeycomb lattice by depositing permalloy, Ni0.81Fe0.19, in near parallel configuration in an electron‐beam evaporation. For this purpose, a new sample holder was designed and setup inside the e‐beam chamber. The substrate was rotated uniformly about its axis during the deposition to create uniformity. This allowed evaporated permalloy to coat the top surface of the honeycomb only, producing the desired magnetic honeycomb lattice with a typical element size of 12 nm (length) × 5 nm (width) × 8 nm (thickness). Atomic force microscopy image of a typical honeycomb lattice is shown in Figure 1a (also see Figure S6 in the Supporting Information where it is shown that the average roughness in the thickness of a honeycomb element is less than 0.5 nm). The center‐to‐center spacing between neighboring honeycombs is ≈ 30 nm. Thus, each honeycomb is about 30 nm wide. Further details about the fabrication procedure can be found somewhere else. GISAXS, at an incident angle of 0.15o, confirmed the good structural quality of the sample.
study
100.0
Neutron Scattering Measurements: Neutron scattering measurements were performed on a 25 mm × 25 mm surface area sample at the magnetism reflectometer, beam line 4A of the Spallation Neutron Source, at Oak Ridge National Laboratory. The instrument utilizes the time of flight technique in a horizontal scattering geometry with a bandwidth of ≈ 2.8 Å (wavelength varying between 2.2 and 5.0 Å). The beam was collimated using a set of slits before the sample and measured with a 2D position sensitive 3He detector with 1.5 mm resolution at 2.5 m from the sample. The sample was mounted on the copper cold finger of a close cycle refrigerator with a base temperature of T = 7 K. Beam polarization and polarization analysis were performed using reflective supermirror devices, achieving better than 98% polarization efficiency over the full wavelength band. For reflectivity and off‐specular scattering, the full vertical divergence was used for maximum intensity and a 1% Δθ/θ ≈ ΔQ z/Q z relative resolution in horizontal direction. These measurements were carried out with five q‐ranges in a total of 380 min. For GISANS experiments, both directions were collimated to reach a symmetric resolution of 0.13° in Q y and Q z directions. Counting times for these measurements were 8 h per instrument setting (temperature, incident angle).
study
100.0
Scattering Modeling: The DWBA was used as implemented in BornAgain29 to model the GISAXS, GISANS, and off‐specular scattering intensity using the same model with different instrument and material parameters. The specular neutron reflectivity data at T = 300 K (see Figure S1, Supporting Information), fitted using the GenX software,30 were used as a basis for the model generation. For more details about the used model see the Supporting Information and corresponding theoretical background in refs. 23, 30, 39, 40, 41, 42.
study
99.94
Conditional gene knockout (cKO) is a genetic technique that enables the investigation of gene function in a temporally regulated manner in specific tissues or cell types (Lobe and Nagy, 1998; Rajewsky et al., 1996; Rossant and McMahon, 1999). Although the global KO approach is useful for investigating the first developmental function of a gene of interest, a conditional approach is often essential for deciphering the tissue-specific function of a gene in later biological events, avoiding the confounding effects of embryonic lethality or other morphogenetic defects that can result from global gene deletion. The cKO approach has been used most extensively in mice, owing to the availability of embryonic stem (ES) cells. ES cells possess the unique capacity of efficient homologous recombination (HR) (Koller and Smithies, 1992; Thomas and Capecchi, 1987), thereby providing a platform for engineering loxP-flanking (floxed) genomic segments that can be excised by Cre recombinase. However, recent advances in genome editing technologies have provided new possibilities for efficiently engineering conditional alleles in any model organism without the need for ES cells (Bedell et al., 2012; Brown et al., 2013; Dickinson et al., 2013; Wang et al., 2013).
review
99.9
The zebrafish has been widely used as a model organism in developmental biology and regeneration research. However, the first floxed allele of an endogenous zebrafish gene generated using transcription activator-like effector nuclease (TALEN)-mediated HR was only recently reported (Hoshijima et al., 2016). To expand the range of genetic tools that can facilitate cKO analysis in zebrafish, we explored an alternative approach using a Cre-dependent genetic switch referred to as FLEx (Schnütgen et al., 2003). FLEx is characterized by strategically arranged wild-type (WT) and mutant loxP sites that facilitate the induction of Cre-dependent stable inversion of a gene trap cassette (Schnütgen et al., 2003). Although similar approaches have been used for conditional gene trap mutagenesis in zebrafish (Clark et al., 2011; Ni et al., 2012; Jungke et al., 2016; Trinh et al., 2011), no reports have described the precise targeting of a FLEx cassette at defined loci. In the present report, we describe a streamlined method to generate a conditional allele using a FLEx-based genetic switch via in vivo HR. Using this approach, we successfully generated a targeted sonic hedgehog a (shha) cKO allele to analyze shha function during heart development and regeneration. Our findings demonstrate the utility of the FLEx-based method for cKO analysis of gene function in zebrafish.
study
99.94
We modified FT1 (Ni et al., 2012), a FLEx construct used for random gene trap mutagenesis in zebrafish, and generated a donor vector for the generation of zebrafish with an invertible gene trap cassette via HR (Zwitch) (Figure 1A and Figure 1—figure supplement 1A). Zwitch consists of a removable, lens-specific enhanced green fluorescence protein (EGFP) tag (LG) (Lee et al., 2005) and an invertible splice acceptor site conjugated to a gene expressing a red fluorescent protein (TagRFP) via a 2A self-cleaving peptide sequence (Figure 1A and Figure 1—figure supplement 1B). The region encompassing LG and the gene trap cassette is flanked by unique restriction enzyme sites in which the right arm (RA) and left arm (LA) of homologous sequences are inserted (Figure 1A and Figure 1—figure supplement 1A).10.7554/eLife.24635.002Figure 1.Generation of the shha conditional allele using Zwitch.(A) Schematic of Zwitch. (B) Schematic of the zebrafish shha locus and TALEN used to induce DNA DSBs in intron 1. Exons are indicated by filled boxes with numbers. The binding sites for the TALEN pair are highlighted in blue, and the XbaI site in the spacer region is highlighted in green. (C) The efficiency of the TALENs in introducing DSBs. XbaI digestion of PCR products amplified from the genomic DNA of embryos injected with TALEN mRNAs. The efficiency of the TALEN pair in inducing DSBs (67%) was quantified from the gel image using ImageJ software. (D) Schematic of the strategy used to target shha via TALEN-mediated homologous recombination with pZwitch-shha-int1. (E) The screening process for founders. (F) Genomic PCR analysis of the Zwitch insertion with the correct orientation. (G) Southern blot analysis of the Zwitch-modified shha allele. BGHpA, bovine growth hormone polyadenylation signal; cryaa, α A-crystallin; LA, left arm; RA, right arm.DOI: http://dx.doi.org/10.7554/eLife.24635.00210.7554/eLife.24635.003Figure 1—figure supplement 1.pZwitch vector.(A) Schematic of pZwitch. (B) Sequence of the splice acceptor and P2A components of pZwitch (filled bar in A). The splice acceptor sequence was derived from pFT1 (Ni et al., 2012). pZwitch for the +2 and+3 reading frames was generated by inserting T and TCGAT, respectively, at the indicated site. BP, branch point; ESE, exonic splice enhancer; ISE, intronic splice enhancer; SAC, splice acceptor consensus.DOI: http://dx.doi.org/10.7554/eLife.24635.00310.7554/eLife.24635.004Figure 1—figure supplement 2.DNA sequence of the shhact allele.(A) DNA sequence at the 5′ end of the LA homology sequence of pZwitch-shha-int1. (B) DNA sequence at the 3′ end of the RA homology sequence of pZwitch-shha-int1.DOI: http://dx.doi.org/10.7554/eLife.24635.004
study
100.0
(A) Schematic of Zwitch. (B) Schematic of the zebrafish shha locus and TALEN used to induce DNA DSBs in intron 1. Exons are indicated by filled boxes with numbers. The binding sites for the TALEN pair are highlighted in blue, and the XbaI site in the spacer region is highlighted in green. (C) The efficiency of the TALENs in introducing DSBs. XbaI digestion of PCR products amplified from the genomic DNA of embryos injected with TALEN mRNAs. The efficiency of the TALEN pair in inducing DSBs (67%) was quantified from the gel image using ImageJ software. (D) Schematic of the strategy used to target shha via TALEN-mediated homologous recombination with pZwitch-shha-int1. (E) The screening process for founders. (F) Genomic PCR analysis of the Zwitch insertion with the correct orientation. (G) Southern blot analysis of the Zwitch-modified shha allele. BGHpA, bovine growth hormone polyadenylation signal; cryaa, α A-crystallin; LA, left arm; RA, right arm.
study
100.0
10.7554/eLife.24635.003Figure 1—figure supplement 1.pZwitch vector.(A) Schematic of pZwitch. (B) Sequence of the splice acceptor and P2A components of pZwitch (filled bar in A). The splice acceptor sequence was derived from pFT1 (Ni et al., 2012). pZwitch for the +2 and+3 reading frames was generated by inserting T and TCGAT, respectively, at the indicated site. BP, branch point; ESE, exonic splice enhancer; ISE, intronic splice enhancer; SAC, splice acceptor consensus.DOI: http://dx.doi.org/10.7554/eLife.24635.003
study
82.25
(A) Schematic of pZwitch. (B) Sequence of the splice acceptor and P2A components of pZwitch (filled bar in A). The splice acceptor sequence was derived from pFT1 (Ni et al., 2012). pZwitch for the +2 and+3 reading frames was generated by inserting T and TCGAT, respectively, at the indicated site. BP, branch point; ESE, exonic splice enhancer; ISE, intronic splice enhancer; SAC, splice acceptor consensus.
other
98.06
10.7554/eLife.24635.004Figure 1—figure supplement 2.DNA sequence of the shhact allele.(A) DNA sequence at the 5′ end of the LA homology sequence of pZwitch-shha-int1. (B) DNA sequence at the 3′ end of the RA homology sequence of pZwitch-shha-int1.DOI: http://dx.doi.org/10.7554/eLife.24635.004
other
99.4
We selected shha (GenBank ID: 30269), the gene encoding Sonic Hedgehog a, to test the utility of Zwitch. We tested several pairs of TALENs and selected the one that most efficiently induced DNA double-strand breaks (DSBs) in intron 1 of shha (Figure 1B and C). mRNA encoding the TALENs was co-injected with the donor vector (pZwitch-shha-int1) into one-cell-stage embryos. The donor vector includes the RA and LA sequences in the orientation that facilitated the insertion of Zwitch in the non-mutagenic orientation (Figure 1D). The injected fish were examined for LG expression at 7 and 45 days post-fertilization (dpf), and fish that maintained LG expression at 45 dpf were raised to adulthood and outcrossed with WT fish (Figure 1E). We screened 32 LG+ potential founders, finding that 19 of them (59%) transmitted LG expression (Figure 1E). PCR demonstrated that Zwitch was inserted in the correct non-mutagenic orientation in 17 (89%) of the 19 LG+ founders (Figure 1E and F). The precise location of the insertion was further verified using DNA sequencing (Figure 1—figure supplement 2A and B) and Southern blot analysis (Figure 1G). We subsequently established a conditional gene trap line, Tg(shha:Zwitch)vcc8Gt, which is referred to as shhact hereafter. We explained key steps for using Zwitch for other genes in the subsection titled ‘Targeting Zwitch into other genes’ in the Materials and methods.
study
100.0
To characterize the shhact allele (Figure 2A), we injected one-cell-stage shhact/+ embryos with a Cre expression vector (pUbb-iCRE-GFP; Figure 2—figure supplement 1A and B). Genomic PCR and reverse transcription (RT)-PCR detected the gene trap cassette inversion (Figure 2B) and alternative shha splicing in embryos injected with Cre DNA (Cre+) but not in uninjected embryos (Cre−) (Figure 2C). DNA sequencing verified the precise inversion (Figure 2—figure supplement 2A) and in-frame transcription of shha-P2A-TagRFP mRNA (Figure 2—figure supplement 2B). We injected Flp mRNA into one-cell-stage embryos of the offspring of an incross of shhact/+ fish and detected the excision of the FRT-flanked LG cassette using genomic PCR (Figure 2D and E). EGFP fluorescence was lost or reduced at approximately the expected frequency of the homozygous or heterozygous shhact genotype (Figure 2F and G).10.7554/eLife.24635.005Figure 2.Characterization of the shhact allele.(A) Schematic of Cre-dependent conversion of Zwitch from the non-mutagenic orientation to the mutagenic orientation. Cre activation induces an inversion between loxP or lox5171 sites and the subsequent excision of loxP or lox5171-flanking DNA sequences (Schnütgen and Ghyselinck, 2007), thereby permanently converting Zwitch into the mutagenic form and inducing aberrant shha splicing. (B) PCR analysis of the Zwitch inversion. Genomic DNA from 72 hpf Cre+ and Cre− shhact/+ embryos was analyzed using PCR. (C) RT-PCR analysis of shha expression in 72 hpf Cre+ and Cre− shhact/+ embryos. (D) Schematic of Flp-mediated excision of the FRT-flanked LG tag in the shhact allele. (E) Genomic PCR analysis of Flp-injected (+) or uninjected (−) embryos from a cross of shhact/+ adults. PCR using F3 and R1 primers detected shhact alleles in both samples. Flp mRNA was synthesized from linearized pCS2-FLPo (Materials and methods). (F) Representative image of embryos injected with Flp mRNA. Arrows indicate the lens. (G) Quantification of phenotypes of the embryos analyzed in F. A total of 87 Flp-injected (+) and 99 uninjected embryos (−) were analyzed (****p<1.0 × 10−8 Fisher’s exact test). dpf, days post-fertilization.DOI: http://dx.doi.org/10.7554/eLife.24635.00510.7554/eLife.24635.006Figure 2—figure supplement 1.Cre expression vector.(A) Schematic of pUbb-iCRE-GFP. (B) A representative image of embryos injected with pUbb-iCRE-GFP expressing EGFP (right) on approximately 80% of the body surface (left). Embryos expressing EGFP at similar or greater levels were selected for the analysis presented in Figure 3A–C.DOI: http://dx.doi.org/10.7554/eLife.24635.00610.7554/eLife.24635.007Figure 2—figure supplement 2.DNA sequence of the inverted shhact allele and its transcript.(A) DNA sequence of the inverted shhact allele. (B) DNA sequence of shha-P2A-TagRFP mRNA.DOI: http://dx.doi.org/10.7554/eLife.24635.00710.7554/eLife.24635.008Figure 2—figure supplement 3.TagRFP expression from the inverted shhact allele.Weak expression of TagRFP was detected in cells in the ganglion cell layer (GCL) in the retina of ubb:Cre-GFP; shhact/+ embryos but not in shhact/+ and ubb:Cre-GFP embryos. Confocal projections of z-stacks are shown in three panels from the right, and the bright field image of in situ hybridization of shha mRNA is shown in the far left panel. Dotted lines, approximate border of the GCL. Scale bar, 50 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.008
study
100.0
(A) Schematic of Cre-dependent conversion of Zwitch from the non-mutagenic orientation to the mutagenic orientation. Cre activation induces an inversion between loxP or lox5171 sites and the subsequent excision of loxP or lox5171-flanking DNA sequences (Schnütgen and Ghyselinck, 2007), thereby permanently converting Zwitch into the mutagenic form and inducing aberrant shha splicing. (B) PCR analysis of the Zwitch inversion. Genomic DNA from 72 hpf Cre+ and Cre− shhact/+ embryos was analyzed using PCR. (C) RT-PCR analysis of shha expression in 72 hpf Cre+ and Cre− shhact/+ embryos. (D) Schematic of Flp-mediated excision of the FRT-flanked LG tag in the shhact allele. (E) Genomic PCR analysis of Flp-injected (+) or uninjected (−) embryos from a cross of shhact/+ adults. PCR using F3 and R1 primers detected shhact alleles in both samples. Flp mRNA was synthesized from linearized pCS2-FLPo (Materials and methods). (F) Representative image of embryos injected with Flp mRNA. Arrows indicate the lens. (G) Quantification of phenotypes of the embryos analyzed in F. A total of 87 Flp-injected (+) and 99 uninjected embryos (−) were analyzed (****p<1.0 × 10−8 Fisher’s exact test). dpf, days post-fertilization.
study
100.0
10.7554/eLife.24635.006Figure 2—figure supplement 1.Cre expression vector.(A) Schematic of pUbb-iCRE-GFP. (B) A representative image of embryos injected with pUbb-iCRE-GFP expressing EGFP (right) on approximately 80% of the body surface (left). Embryos expressing EGFP at similar or greater levels were selected for the analysis presented in Figure 3A–C.DOI: http://dx.doi.org/10.7554/eLife.24635.006
study
99.94
(A) Schematic of pUbb-iCRE-GFP. (B) A representative image of embryos injected with pUbb-iCRE-GFP expressing EGFP (right) on approximately 80% of the body surface (left). Embryos expressing EGFP at similar or greater levels were selected for the analysis presented in Figure 3A–C.
study
99.94
10.7554/eLife.24635.008Figure 2—figure supplement 3.TagRFP expression from the inverted shhact allele.Weak expression of TagRFP was detected in cells in the ganglion cell layer (GCL) in the retina of ubb:Cre-GFP; shhact/+ embryos but not in shhact/+ and ubb:Cre-GFP embryos. Confocal projections of z-stacks are shown in three panels from the right, and the bright field image of in situ hybridization of shha mRNA is shown in the far left panel. Dotted lines, approximate border of the GCL. Scale bar, 50 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.008
study
99.94
Weak expression of TagRFP was detected in cells in the ganglion cell layer (GCL) in the retina of ubb:Cre-GFP; shhact/+ embryos but not in shhact/+ and ubb:Cre-GFP embryos. Confocal projections of z-stacks are shown in three panels from the right, and the bright field image of in situ hybridization of shha mRNA is shown in the far left panel. Dotted lines, approximate border of the GCL. Scale bar, 50 μm.
study
99.94
To validate that shhact was a loss-of-function allele, we obtained shhact/ct embryos from incrosses of shhact/+ fish and individually genotyped the embryos using PCR (Figure 3—figure supplement 1A and B). Quantitative RT-PCR (qRT-CPR) revealed that the expression of shha and Hedgehog (Hh) target genes was strongly decreased in Cre+shhact/ct embryos but not in Cre− shhact/ct embryos, indicating that the shhact allele is functional in the absence of Cre activation (Figure 3A). Furthermore, most Cre+ shhact/ct embryos exhibited truncated pectoral fins and U-shaped somites that lacked clear horizontal myosepta (Figure 3B and C), defects reminiscent of homozygous shha mutant embryos (Schauerte et al., 1998). To confirm this result, we established shhact/+ fish carrying the transgene Tg(ubb:iCRE-GFP) (ubb:Cre-GFP), in which codon-improved Cre (iCRE) DNA was expressed by a strong, ubiquitously expressed ubiquitin B (ubb) promoter (Mosimann et al., 2011) (Figure 2—figure supplement 1A). From incrosses of this line, we obtained embryos carrying the WT and/or mutagenic shha allele and analyzed the phenotype of these embryos (Figure 3—figure supplement 2A and B). We found that pectoral fin development was largely normal in WT and heterozygous mutant embryos but severely hampered in all homozygous mutants examined (Figure 3—figure supplement 2C). We also performed semi-qRT-PCR analysis of shha expression and confirmed that its expression was reduced to nearly 50% of WT levels in the heterozygous mutants and to an undetectable level in the homozygous mutants (Figure 3—figure supplement 2D and E). Together, these results provide clear evidence that Zwitch can inducibly disrupt gene function in zebrafish.10.7554/eLife.24635.009Figure 3.Phenotype of shhact/ct embryos globally expressing Cre.(A) qRT-PCR analysis of 72 hpf Cre+ and Cre− shhact/ct embryos (n = 10 and 9). WT embryos injected with pUbb-iCRE-GFP DNA were used as a control (n = 9). Ten pooled embryos per sample were used for qRT-PCR analysis. The data are presented as the mean ± SEM (***p<0.001, Mann–Whitney U test). (B) Phenotypes of 72 hpf Cre+ and Cre- shhact/ct embryos. Arrows, pectoral fins; arrowheads, somite boundaries; brackets, horizontal myoseptum. Bright field images were captured using an MVX10 microscope. The composite images shown were generated using ImageJ software. Somite defects were observed in all embryos with severe pectoral fin defects. (C) Quantification of pectoral fin phenotypes from the embryos in B (n = 7 [WT, Cre+], n = 12 [shhact/ct, Cre−], and n = 28 [shhact/ct, Cre+]; **p<0.01, Fisher’s exact test). N.S., not significant (p=0.5392). The embryos used in A–C were selected on the basis of their high-level expression of Cre as described in Figure 2—figure supplement 1B (see also Cre DNA and mRNA injection, Materials and methods). A moderate pectoral fin defect was observed in control samples, likely due to injection artifacts.DOI: http://dx.doi.org/10.7554/eLife.24635.00910.7554/eLife.24635.010Figure 3—figure supplement 1.Genotyping PCR of shhact alleles.(A) Schematic of the shha WT allele (+) and conditional trap allele (ct) and primers sites. (B) Genotyping PCR result of single embryos from an incross of shhact/+ fish injected with Cre DNA at single-cell stage. Note that ct/ct genotype was predictable due to the developmental defects observed in shhact/ct embryos after Cre DNA injection. (C) Genotyping PCR result of single embryos from a cross of tcf21:CreER; shhact/+ fish with shhact/+ fish. We did not use embryos exhibiting a faint band for experiments (asterisk). PCR with Cre-Scr-F/R primers occasionally amplified a faint non-specific band at a lower size. ND, not determined. Please see the Genotyping and Sample preparation for embryo genotyping and analysis section in the Materials and methods for details of the screening procedures.DOI: http://dx.doi.org/10.7554/eLife.24635.01010.7554/eLife.24635.011Figure 3—figure supplement 2.Analysis of embryos carrying mutagenic shha alleles.(A) Schematic of the shha WT allele (+), non-mutagenic, conditional-trap allele (ct), and inverted mutagenic allele (m), and primers sites. (B) Genotyping PCR result of single embryos from incrosses of ubb:Cre-GFP; shhact/+ fish. (C) Pectoral fin phenotype of WT and heterozygous and homozygous mutagenic embryos in B. Embryos that did not carry ubb:Cre-GFP transgene were analyzed (Cre-GFP−). Pectoral fin development was normal in all WT embryos (0 abnormal in 5 analyzed) and most heterozygous mutants (1 moderately abnormal in 7 analyzed; p=0.3774, Fisher's exact test), but severely hampered in all homozygous mutants (3 abnormal in 3 analyzed; p<0.01, Fisher’s exact test). (D) Semi-qRT-PCR analysis of shha expression in WT and heterozygous and homozygous mutagenic embryos. The number above each lane in the gel picture indicates the Embryo ID in B. (E) Densitometric quantification of the PCR result in D. The data represent the mean ± SD (**p<0.01; Mann–Whitney U test).DOI: http://dx.doi.org/10.7554/eLife.24635.011
study
100.0
(A) qRT-PCR analysis of 72 hpf Cre+ and Cre− shhact/ct embryos (n = 10 and 9). WT embryos injected with pUbb-iCRE-GFP DNA were used as a control (n = 9). Ten pooled embryos per sample were used for qRT-PCR analysis. The data are presented as the mean ± SEM (***p<0.001, Mann–Whitney U test). (B) Phenotypes of 72 hpf Cre+ and Cre- shhact/ct embryos. Arrows, pectoral fins; arrowheads, somite boundaries; brackets, horizontal myoseptum. Bright field images were captured using an MVX10 microscope. The composite images shown were generated using ImageJ software. Somite defects were observed in all embryos with severe pectoral fin defects. (C) Quantification of pectoral fin phenotypes from the embryos in B (n = 7 [WT, Cre+], n = 12 [shhact/ct, Cre−], and n = 28 [shhact/ct, Cre+]; **p<0.01, Fisher’s exact test). N.S., not significant (p=0.5392). The embryos used in A–C were selected on the basis of their high-level expression of Cre as described in Figure 2—figure supplement 1B (see also Cre DNA and mRNA injection, Materials and methods). A moderate pectoral fin defect was observed in control samples, likely due to injection artifacts.
study
100.0
10.7554/eLife.24635.010Figure 3—figure supplement 1.Genotyping PCR of shhact alleles.(A) Schematic of the shha WT allele (+) and conditional trap allele (ct) and primers sites. (B) Genotyping PCR result of single embryos from an incross of shhact/+ fish injected with Cre DNA at single-cell stage. Note that ct/ct genotype was predictable due to the developmental defects observed in shhact/ct embryos after Cre DNA injection. (C) Genotyping PCR result of single embryos from a cross of tcf21:CreER; shhact/+ fish with shhact/+ fish. We did not use embryos exhibiting a faint band for experiments (asterisk). PCR with Cre-Scr-F/R primers occasionally amplified a faint non-specific band at a lower size. ND, not determined. Please see the Genotyping and Sample preparation for embryo genotyping and analysis section in the Materials and methods for details of the screening procedures.DOI: http://dx.doi.org/10.7554/eLife.24635.010
study
99.94
(A) Schematic of the shha WT allele (+) and conditional trap allele (ct) and primers sites. (B) Genotyping PCR result of single embryos from an incross of shhact/+ fish injected with Cre DNA at single-cell stage. Note that ct/ct genotype was predictable due to the developmental defects observed in shhact/ct embryos after Cre DNA injection. (C) Genotyping PCR result of single embryos from a cross of tcf21:CreER; shhact/+ fish with shhact/+ fish. We did not use embryos exhibiting a faint band for experiments (asterisk). PCR with Cre-Scr-F/R primers occasionally amplified a faint non-specific band at a lower size. ND, not determined. Please see the Genotyping and Sample preparation for embryo genotyping and analysis section in the Materials and methods for details of the screening procedures.
study
99.94
10.7554/eLife.24635.011Figure 3—figure supplement 2.Analysis of embryos carrying mutagenic shha alleles.(A) Schematic of the shha WT allele (+), non-mutagenic, conditional-trap allele (ct), and inverted mutagenic allele (m), and primers sites. (B) Genotyping PCR result of single embryos from incrosses of ubb:Cre-GFP; shhact/+ fish. (C) Pectoral fin phenotype of WT and heterozygous and homozygous mutagenic embryos in B. Embryos that did not carry ubb:Cre-GFP transgene were analyzed (Cre-GFP−). Pectoral fin development was normal in all WT embryos (0 abnormal in 5 analyzed) and most heterozygous mutants (1 moderately abnormal in 7 analyzed; p=0.3774, Fisher's exact test), but severely hampered in all homozygous mutants (3 abnormal in 3 analyzed; p<0.01, Fisher’s exact test). (D) Semi-qRT-PCR analysis of shha expression in WT and heterozygous and homozygous mutagenic embryos. The number above each lane in the gel picture indicates the Embryo ID in B. (E) Densitometric quantification of the PCR result in D. The data represent the mean ± SD (**p<0.01; Mann–Whitney U test).DOI: http://dx.doi.org/10.7554/eLife.24635.011
study
100.0
(A) Schematic of the shha WT allele (+), non-mutagenic, conditional-trap allele (ct), and inverted mutagenic allele (m), and primers sites. (B) Genotyping PCR result of single embryos from incrosses of ubb:Cre-GFP; shhact/+ fish. (C) Pectoral fin phenotype of WT and heterozygous and homozygous mutagenic embryos in B. Embryos that did not carry ubb:Cre-GFP transgene were analyzed (Cre-GFP−). Pectoral fin development was normal in all WT embryos (0 abnormal in 5 analyzed) and most heterozygous mutants (1 moderately abnormal in 7 analyzed; p=0.3774, Fisher's exact test), but severely hampered in all homozygous mutants (3 abnormal in 3 analyzed; p<0.01, Fisher’s exact test). (D) Semi-qRT-PCR analysis of shha expression in WT and heterozygous and homozygous mutagenic embryos. The number above each lane in the gel picture indicates the Embryo ID in B. (E) Densitometric quantification of the PCR result in D. The data represent the mean ± SD (**p<0.01; Mann–Whitney U test).
study
100.0
To demonstrate the utility of Zwitch, we investigated the functional consequences of inactivation of shha expression in the epicardium, the mesothelial layer covering the heart. Although the epicardium was reported to produce Shh during heart development in mice (Lavine et al., 2006), the precise function of Shh has not been addressed using an epicardium-specific KO approach. Using fluorescence-activated cell sorting (FACS), we purified epicardial cells and cardiomyocytes from the dissected hearts of transgenic reporter fish (see Flow cytometry, Materials and methods) and confirmed shha expression in epicardial cells and ptch1 expression in both epicardial cells and cardiomyocytes (Figure 4A). In situ hybridization analysis also detected shha expression in the epicardium of the ventricle (Figure 4—figure supplement 1A–C). Next, we crossed shhact/+ fish with shhact/+ fish carrying the epicardium-specific inducible Cre transgene Tg(tcf21:CreER) (tcf21:CreER) (Kikuchi et al., 2011) and obtained tcf21:CreER; shhact/ct embryos after PCR genotyping of individual embryos (Figure 3—figure supplement 1C). We inactivated shha expression in the epicardium by treating 24–72 hr post-fertilization (hpf) embryos with 4-hydroxytamoxifen (4-HT; Figure 4B), which induced a nearly complete inversion of the gene trap cassette (Figure 4—figure supplement 2D). Strikingly, most embryos lacking shha expression in the epicardium (epi-KO) developed severe cardiac edema by 96 hpf (Figure 4C). Cardiac edema was unclear or extremely weak in epi-KO hearts at 72 hpf, suggesting that epicardial inactivation of shha expression leads to heart defects at later developmental stages. We could not determine whether a similar cardiac phenotype was also observed in the global shha mutant embryos, as proper characterization was not possible due to the pleiotropic effects of global inactivation of shha expression at later time points. The epi-KO embryos had a thinner myocardial wall (Figure 4D) and significantly fewer cardiomyocytes (Figure 4E). These findings indicate that epicardium-derived Shha plays a crucial role in zebrafish heart development.10.7554/eLife.24635.012Figure 4.Epicardium-specific inactivation of shha expression during heart development.(A) Semi-qRT-PCR analysis of purified cardiomyocytes (CM) and epicardial cells (Epi) from 96 hpf hearts (see also Flow Cytometry, Materials and methods). Cardiomyocyte (vmhc) and epicardial (tcf21) markers were used to confirm the specificity of cell sorting. (B) Semi-qRT-PCR analysis of shha and ptch1 expression in hearts dissected from tcf21:CreER; shhact/ct embryos treated with the vehicle (−) or 4-HT (+). (C) Phenotype of tcf21:CreER; shhact/ct embryos treated with the vehicle or 4-HT. Severe cardiac edema was observed in 4-HT–treated embryos at 96 hpf (six abnormal in eight analyzed; right, arrowheads) but not in vehicle-treated embryos (zero abnormal in eight analyzed; left, arrowheads; n = 8 each; p<0.01, Fisher’s exact test). (D) Immunofluorescence of heart sections obtained from vehicle- or 4-HT–treated tcf21:CreER; shhact/ct embryos. Insets, single-channel images of Mef2 immunofluorescence. Dotted yellow lines in insets depict the outline of the ventricle. (E) Quantification of Mef2+ nuclei from the sections obtained from the vehicle- or 4-HT–treated tcf21:CreER; shhact/ct embryos in D (n = 13 and 12). The data are presented as the mean ± SEM (***p<0.001, Mann–Whitney U test). (F) Semi-qRT-PCR analysis of epicardial marker gene expression in hearts dissected from tcf21:CreER; shhact/ct embryos treated with the vehicle (−) or 4-HT (+). (G) Immunofluorescence staining of heart sections obtained from vehicle- or 4-HT–treated tcf21:CreER; shhact/ct embryos. Raldh2 immunofluorescence was detected in tcf21:DsRed2+ epicardial cells in vehicle-treated embryos (left, arrowheads) but not in 4-HT-treated embryos (right, arrowheads). Bottom panels, single-channel images of Raldh2 immunofluorescence. (H) Semi-qRT-PCR analysis of the expression of myocardial growth factor genes in hearts dissected from tcf21:CreER; shhact/ct embryos treated with the vehicle (−) or 4-HT (+). (I) Semi-qRT-PCR analysis of shha-dependent myocardial growth factor genes in purified cardiomyocytes (CM) and epicardial cells (Epi) obtained from 96 hpf hearts. Single confocal sections are shown in D and G. ve, ventricle. Scale bar, 10 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.01210.7554/eLife.24635.013Figure 4—figure supplement 1.shha expression during heart development and regeneration.(A–C) shha expression during heart development in zebrafish. Single confocal slice images of the rectangles in A are shown in B and C. (D–F) shha expression during heart regeneration in zebrafish. Single confocal slice images of the rectangles in the right panel of D are shown in E and F. at, atrium; ba, bulbus arteriosus; es, esophagus; ve, ventricle. Scale bar, 50 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.01310.7554/eLife.24635.014Figure 4—figure supplement 2.Inversion rate measurement.(A) Schematic of the non-mutagenic and mutagenic allele, primer sites, and BglI recognition site. (B) A gel image showing the PCR products of the non-mutagenic alleles without (left) and with BglI digestion (middle) and the PCR products of the mutagenic alleles with BglI digestion (right). Genomic DNA was prepared from 96 hpf shhact/+ embryos injected with (+) or without (−) Cre mRNA at single-cell stage. Cre mRNA was prepared from linearized pCS2-iCRE-BFP, in which iCRE cDNA is linked to TagBFP via P2A peptide sequence. (C) Analysis of inversion rate and shha expression levels in single embryos. Embryos were prepared from an incross of shhact/+ fish and injected with Cre mRNA at single-cell stage. Embryos without Cre mRNA injection were used as control (Sample 1). The injected embryos were separated based on the expression of TagBFP at moderate (Sample 2) and high levels (Sample 3 and 4). Individual embryos were genotyped at 96 hpf and genomic DNA and total RNA of each embryo using TRIzol following the manufactured protocol. The upper panels are the gel images used for inversion rate measurement. The numbers on the bottom are the obtained inversion rate in each sample. The lower panels are the results of semi-qRT-PCR analysis of shha expression levels of each sample. (D) Analysis of inversion rate in epicardial cells in zebrafish embryos. Embryos were prepared from crosses of tcf21:DsRed2; shhact/+ fish with shhact/+ fish (Sample 1) and tcf21:CreER; shhact/+ fish with tcf21:DsRed2; shhact/+ fish (Sample 2). Embryos were treated with 4-HT and examined for DsRed2 expression to screen tcf21:DsRed2 transgene and lens EGFP expression to screen tcf21:CreER transgene (Kikuchi et al., 2011). Lens EGFP from tcf21:CreER transgene is considerably stronger than that from shhact alleles in embryos, which enables us to sort tcf21:CreER+ embryos with shhact background. PCR genotyping on shhact alleles were not performed; this does not affect the measurement as the inversion rate is quantified based on the PCR products of the shhact alleles. tcf21:DsRed2+ epicardial cells were isolated by FACS and 1,000 DsRed2+ cells were used for analysis. The estimated inversion rate is 100%; note that the bands from the non-mutagenic allele (N1 and N2) are undetectable in Sample 2 after BglI digestion. (E) Analysis of inversion rate in epicardial cells in the adult zebrafish ventricle. Ventricles were prepared from 4-HT–treated adult tcf21:DsRed2; shhact/+ fish (Sample 1) and tcf21:CreER; tcf21:DsRed2; shhact/+ fish (Sample 2). tcf21:DsRed2+ epicardial cells were isolated by FACS and 5000 DsRed2+ cells were used for analysis. The estimated inversion rate is 100%; note that the bands from the non-mutagenic allele (N1 and N2) are undetectable in Sample 2 after BglI digestion.DOI: http://dx.doi.org/10.7554/eLife.24635.014
study
100.0
(A) Semi-qRT-PCR analysis of purified cardiomyocytes (CM) and epicardial cells (Epi) from 96 hpf hearts (see also Flow Cytometry, Materials and methods). Cardiomyocyte (vmhc) and epicardial (tcf21) markers were used to confirm the specificity of cell sorting. (B) Semi-qRT-PCR analysis of shha and ptch1 expression in hearts dissected from tcf21:CreER; shhact/ct embryos treated with the vehicle (−) or 4-HT (+). (C) Phenotype of tcf21:CreER; shhact/ct embryos treated with the vehicle or 4-HT. Severe cardiac edema was observed in 4-HT–treated embryos at 96 hpf (six abnormal in eight analyzed; right, arrowheads) but not in vehicle-treated embryos (zero abnormal in eight analyzed; left, arrowheads; n = 8 each; p<0.01, Fisher’s exact test). (D) Immunofluorescence of heart sections obtained from vehicle- or 4-HT–treated tcf21:CreER; shhact/ct embryos. Insets, single-channel images of Mef2 immunofluorescence. Dotted yellow lines in insets depict the outline of the ventricle. (E) Quantification of Mef2+ nuclei from the sections obtained from the vehicle- or 4-HT–treated tcf21:CreER; shhact/ct embryos in D (n = 13 and 12). The data are presented as the mean ± SEM (***p<0.001, Mann–Whitney U test). (F) Semi-qRT-PCR analysis of epicardial marker gene expression in hearts dissected from tcf21:CreER; shhact/ct embryos treated with the vehicle (−) or 4-HT (+). (G) Immunofluorescence staining of heart sections obtained from vehicle- or 4-HT–treated tcf21:CreER; shhact/ct embryos. Raldh2 immunofluorescence was detected in tcf21:DsRed2+ epicardial cells in vehicle-treated embryos (left, arrowheads) but not in 4-HT-treated embryos (right, arrowheads). Bottom panels, single-channel images of Raldh2 immunofluorescence. (H) Semi-qRT-PCR analysis of the expression of myocardial growth factor genes in hearts dissected from tcf21:CreER; shhact/ct embryos treated with the vehicle (−) or 4-HT (+). (I) Semi-qRT-PCR analysis of shha-dependent myocardial growth factor genes in purified cardiomyocytes (CM) and epicardial cells (Epi) obtained from 96 hpf hearts. Single confocal sections are shown in D and G. ve, ventricle. Scale bar, 10 μm.
study
100.0
10.7554/eLife.24635.013Figure 4—figure supplement 1.shha expression during heart development and regeneration.(A–C) shha expression during heart development in zebrafish. Single confocal slice images of the rectangles in A are shown in B and C. (D–F) shha expression during heart regeneration in zebrafish. Single confocal slice images of the rectangles in the right panel of D are shown in E and F. at, atrium; ba, bulbus arteriosus; es, esophagus; ve, ventricle. Scale bar, 50 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.013
study
99.94
(A–C) shha expression during heart development in zebrafish. Single confocal slice images of the rectangles in A are shown in B and C. (D–F) shha expression during heart regeneration in zebrafish. Single confocal slice images of the rectangles in the right panel of D are shown in E and F. at, atrium; ba, bulbus arteriosus; es, esophagus; ve, ventricle. Scale bar, 50 μm.
study
99.94
10.7554/eLife.24635.014Figure 4—figure supplement 2.Inversion rate measurement.(A) Schematic of the non-mutagenic and mutagenic allele, primer sites, and BglI recognition site. (B) A gel image showing the PCR products of the non-mutagenic alleles without (left) and with BglI digestion (middle) and the PCR products of the mutagenic alleles with BglI digestion (right). Genomic DNA was prepared from 96 hpf shhact/+ embryos injected with (+) or without (−) Cre mRNA at single-cell stage. Cre mRNA was prepared from linearized pCS2-iCRE-BFP, in which iCRE cDNA is linked to TagBFP via P2A peptide sequence. (C) Analysis of inversion rate and shha expression levels in single embryos. Embryos were prepared from an incross of shhact/+ fish and injected with Cre mRNA at single-cell stage. Embryos without Cre mRNA injection were used as control (Sample 1). The injected embryos were separated based on the expression of TagBFP at moderate (Sample 2) and high levels (Sample 3 and 4). Individual embryos were genotyped at 96 hpf and genomic DNA and total RNA of each embryo using TRIzol following the manufactured protocol. The upper panels are the gel images used for inversion rate measurement. The numbers on the bottom are the obtained inversion rate in each sample. The lower panels are the results of semi-qRT-PCR analysis of shha expression levels of each sample. (D) Analysis of inversion rate in epicardial cells in zebrafish embryos. Embryos were prepared from crosses of tcf21:DsRed2; shhact/+ fish with shhact/+ fish (Sample 1) and tcf21:CreER; shhact/+ fish with tcf21:DsRed2; shhact/+ fish (Sample 2). Embryos were treated with 4-HT and examined for DsRed2 expression to screen tcf21:DsRed2 transgene and lens EGFP expression to screen tcf21:CreER transgene (Kikuchi et al., 2011). Lens EGFP from tcf21:CreER transgene is considerably stronger than that from shhact alleles in embryos, which enables us to sort tcf21:CreER+ embryos with shhact background. PCR genotyping on shhact alleles were not performed; this does not affect the measurement as the inversion rate is quantified based on the PCR products of the shhact alleles. tcf21:DsRed2+ epicardial cells were isolated by FACS and 1,000 DsRed2+ cells were used for analysis. The estimated inversion rate is 100%; note that the bands from the non-mutagenic allele (N1 and N2) are undetectable in Sample 2 after BglI digestion. (E) Analysis of inversion rate in epicardial cells in the adult zebrafish ventricle. Ventricles were prepared from 4-HT–treated adult tcf21:DsRed2; shhact/+ fish (Sample 1) and tcf21:CreER; tcf21:DsRed2; shhact/+ fish (Sample 2). tcf21:DsRed2+ epicardial cells were isolated by FACS and 5000 DsRed2+ cells were used for analysis. The estimated inversion rate is 100%; note that the bands from the non-mutagenic allele (N1 and N2) are undetectable in Sample 2 after BglI digestion.DOI: http://dx.doi.org/10.7554/eLife.24635.014
study
100.0
(A) Schematic of the non-mutagenic and mutagenic allele, primer sites, and BglI recognition site. (B) A gel image showing the PCR products of the non-mutagenic alleles without (left) and with BglI digestion (middle) and the PCR products of the mutagenic alleles with BglI digestion (right). Genomic DNA was prepared from 96 hpf shhact/+ embryos injected with (+) or without (−) Cre mRNA at single-cell stage. Cre mRNA was prepared from linearized pCS2-iCRE-BFP, in which iCRE cDNA is linked to TagBFP via P2A peptide sequence. (C) Analysis of inversion rate and shha expression levels in single embryos. Embryos were prepared from an incross of shhact/+ fish and injected with Cre mRNA at single-cell stage. Embryos without Cre mRNA injection were used as control (Sample 1). The injected embryos were separated based on the expression of TagBFP at moderate (Sample 2) and high levels (Sample 3 and 4). Individual embryos were genotyped at 96 hpf and genomic DNA and total RNA of each embryo using TRIzol following the manufactured protocol. The upper panels are the gel images used for inversion rate measurement. The numbers on the bottom are the obtained inversion rate in each sample. The lower panels are the results of semi-qRT-PCR analysis of shha expression levels of each sample. (D) Analysis of inversion rate in epicardial cells in zebrafish embryos. Embryos were prepared from crosses of tcf21:DsRed2; shhact/+ fish with shhact/+ fish (Sample 1) and tcf21:CreER; shhact/+ fish with tcf21:DsRed2; shhact/+ fish (Sample 2). Embryos were treated with 4-HT and examined for DsRed2 expression to screen tcf21:DsRed2 transgene and lens EGFP expression to screen tcf21:CreER transgene (Kikuchi et al., 2011). Lens EGFP from tcf21:CreER transgene is considerably stronger than that from shhact alleles in embryos, which enables us to sort tcf21:CreER+ embryos with shhact background. PCR genotyping on shhact alleles were not performed; this does not affect the measurement as the inversion rate is quantified based on the PCR products of the shhact alleles. tcf21:DsRed2+ epicardial cells were isolated by FACS and 1,000 DsRed2+ cells were used for analysis. The estimated inversion rate is 100%; note that the bands from the non-mutagenic allele (N1 and N2) are undetectable in Sample 2 after BglI digestion. (E) Analysis of inversion rate in epicardial cells in the adult zebrafish ventricle. Ventricles were prepared from 4-HT–treated adult tcf21:DsRed2; shhact/+ fish (Sample 1) and tcf21:CreER; tcf21:DsRed2; shhact/+ fish (Sample 2). tcf21:DsRed2+ epicardial cells were isolated by FACS and 5000 DsRed2+ cells were used for analysis. The estimated inversion rate is 100%; note that the bands from the non-mutagenic allele (N1 and N2) are undetectable in Sample 2 after BglI digestion.
study
100.0
To investigate the mechanism underlying the developmental defects in epi-KO hearts, we analyzed the expression of the pan-epicardium markers tcf21 and pard3 using semi-qRT-PCR (Plavicki et al., 2014; Serluca, 2008) and detected the normal expression of these marker genes in epi-KO hearts (Figure 4F). This result suggests that epicardium-specific loss of shha does not directly affect the development of the epicardium, consistent with the observation that epicardial development is normal in mutant mouse hearts lacking the Shh receptor smoothened (Rudat et al., 2013). However, the expression of the well-characterized epicardial marker raldh2/aldh1a2, which encodes an enzyme that plays a key role in retinoic acid (RA) synthesis (Begemann et al., 2001; Cunningham and Duester, 2015; Niederreither et al., 1999; Sucov and Evans, 1995), was strongly decreased (Figure 4F). To analyze epicardial cell development and raldh2 expression in epi-KO hearts, we crossed tcf21:CreER; shhact/+ with shhact/+ fish carrying the epicardium-specific DsRed2 reporter transgene Tg(tcf21;DsRed2) (tcf21:DsRed2) (Kikuchi et al., 2011) and obtained tcf21:DsRed2; tcf21:CreER; shhact/ct embryos after PCR genotyping of individual embryos. We induced epicardium-specific inactivation of shha expression in these embryos and found that DsRed2+ epicardial cells were normally distributed throughout the epi-KO heart (Figure 4G). Consistent with the PCR results (Figure 4F), Raldh2 protein expression was barely detectable in mutant DsRed2+ cells (Figure 4G). These findings suggest that defects in RA synthesis might underlie the developmental defect in epi-KO hearts.
study
100.0
The expression of fgf2 and wnt9b, two cardiomyocyte mitogens regulated by RA signaling during mouse heart development (Merki et al., 2005), was strongly decreased in epi-KO hearts (Figure 4H), whereas that of igf2a and igf2b, factors regulated by liver-derived erythropoietin (Brade et al., 2011), was unaffected (Figure 4H). Interestingly, the expression of fgf9, a gene encoding a key FGF regulated by RA signaling in the mouse heart (Lavine et al., 2005), as well as fgf20a and fgf20b was unaffected in epi-KO hearts, whereas fgf16 expression levels was strongly decreased (Figure 4H). These findings suggest that although FGF regulation by RA signals is evolutionary conserved, different FGF proteins may regulate cardiogenesis in zebrafish and mammals. We also analyzed the expression of fgf2, fgf16, and wnt9b in purified cardiomyocytes and epicardial cells (Figure 4A), finding that they are primarily expressed in the epicardium during zebrafish heart development (Figure 4I). Thus, epicardium-specific inactivation of shha expression revealed that Shha plays a crucial role in RA synthesis and mitogen expression in the epicardium during heart development in zebrafish.
study
100.0
The zebrafish has been used for regeneration research for more than a decade, but the cKO approach has yet to be applied to these studies. To illustrate the utility of our cKO approach in zebrafish regeneration research, we inactivated epicardial shha expression in injured hearts and investigated its effects on regeneration (Figure 5A). Studies using cyclopamine, which globally inhibits Hh signaling, indicated that epicardial Shha promotes myocardial proliferation (Choi et al., 2013) and epicardial regeneration in cultured injured hearts in vitro (Wang et al., 2015a). Consistent with the observation that the shha promoter is activated in regenerating epicardium (Choi et al., 2013), we detected shha expression in epicardial cells via in situ hybridization (Figure 4—figure supplement 1D–F) and semi-qRT-PCR in purified epicardial cells from tcf21:DsRed2 hearts at 7 days postinjury (dpi) (Figure 5B). Next, we analyzed adult tcf21:DsRed2; tcf21:CreER; shhact/ct fish treated with 4-HT, which induced a nearly complete inversion of the gene trap cassette (Figure 4—figure supplement 2E). We used tcf21:DsRed2; shhact/ct fish treated with 4-HT as a negative control (CreER−) because a low level of cassette inversion was detected in tcf21:CreER; shhact/ct fish in adults, but not in embryos, in the absence of 4-HT (Figure 5—figure supplement 1A–C). Semi-qRT-PCR confirmed that shha expression was depleted in epicardial cells purified from injured hearts (Figure 5C). Interestingly, unlike the embryonic epi-KO heart (Figure 4B and F), ptch1 and raldh2 expression was unchanged in shha mutant adult hearts (Figure 5C and Figure 5—figure supplement 2A and B). Moreover, epicardial cells normally infiltrated the injury site (Figure 5—figure supplement 2A–C) and incorporated 5-ethynyl-2′-deoxyuridine (EdU), a DNA synthesis marker, in epi-KO hearts (Figure 5—figure supplement 2C and D). Consistent with the previous report (Wang et al., 2015a), we detected the expression of other Hh ligand genes, namely dhh and ihhb, in epicardial cells purified from injured hearts (Figure 5B) and epi-KO hearts (Figure 5C). These results suggest that redundant Hh proteins or non-epicardium-derived Shha plays a role in RA synthesis and epicardial regeneration in the adult zebrafish heart.10.7554/eLife.24635.015Figure 5.Epicardial shha expression promotes subepicardial cardiomyocyte proliferation during heart regeneration.(A) Schematic of the experiment. (B) Semi-qRT-PCR analysis of shha in purified tcf21:DsRed2+ epicardial cells obtained from uninjured and injured (7 dpi) tcf21:DsRed hearts. Injury was confirmed by the induction of raldh2 expression. (C) Semi-qRT-PCR analysis of Hh pathway genes using purified tcf21:DsRed2+ epicardial cells obtained from 4-HT–treated 7 dpi tcf21:DsRed; shhact/ct (control, left) and tcf21:DsRed; tcf21:CreER; shhact/ct hearts (right). (D) Immunofluorescence images of the subepicardial (top) and trabecular areas (bottom) of heart sections obtained from 4-HT–treated 7 dpi shhact/ct (control) or tcf21:CreER; shhact/ct hearts. Brackets, subepicardial areas. Dotted lines, approximate amputation plane. Arrows indicate proliferating cardiomyocytes. (E) Quantification of cardiomyocyte proliferation in the subepicardial and trabecular areas of the heart sections obtained from 4-HT–treated 7 dpi shhact/ct (control) or tcf21:CreER; shhact/ct hearts shown in D (n = 6 each). The data are presented as the mean ± SEM (**p<0.01, Mann–Whitney U test). N.S., not significant (p=0.3367). (F) Image of heart sections obtained from 7 dpi gata4:EGFP; tcf21;DsRed2 fish. Subepicardial cardiomyocytes (green, arrows) associate with epicardial cells (magenta, arrowheads). (G) Semi-qRT-PCR analysis of shha pathway genes using purified subepicardial cardiomyocytes obtained from 7 dpi gata4:EGFP ventricles. Cardiomyocytes purified from uninjured cmlc2:EGFP ventricles were used as negative controls. Scale bar, 50 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.01510.7554/eLife.24635.016Figure 5—figure supplement 1.Assessment of spontaneous cassette inversion.(A) Schematic of the inverted shhact allele and primers sites. (B) The inverted allele was not detected in untreated tcf21:CreER; shhact/ct embryos using 40 cycles of PCR. (C) The inverted allele was not detected in untreated adult tcf21:CreER; shhact/ct fish using 30 cycles of PCR, but it was detected when the cycle number was increased to 35. Each PCR sample used genomic DNA isolated from five pooled hearts as a template.DOI: http://dx.doi.org/10.7554/eLife.24635.01610.7554/eLife.24635.017Figure 5—figure supplement 2.A redundant role for epicardial shha in epicardial migration and proliferation during heart regeneration.(A) In situ hybridization analysis of raldh2 expression in sections obtained from 4-HT-treated 7 dpi shhact/ct (control) and tcf21:CreER; shhact/ct hearts. Brackets, injury site. (B) Immunofluorescence staining of Raldh2 and DsRed2 using sections obtained from 4-HT-treated 7 dpi tcf21:DsRed; shhact/ct (control) and tcf21:DsRed; tcf21:CreER; shhact/ct hearts (right). Single-channel images of the rectangle are shown at the bottom. Dotted line, approximate amputation plane. (C) Immunofluorescence staining of DsRed2 and EdU in sections obtained from 4-HT-treated 7 dpi tcf21:DsRed; shhact/ct (control) and tcf21:CreER; tcf21:DsRed; shhact/ct hearts. Inset, non-injured area. Arrows indicate proliferating tcf21+ epicardial cells, which were defined as epicardial cells colabeled DsRed2 and EdU. (D) Quantification of epicardial cell proliferation in the sections obtained from 4-HT-treated 7 dpi tcf21:DsRed; shhact/ct (control) and tcf21:CreER; tcf21:DsRed; shhact/ct hearts shown in C (n = 5 each). The data represent the mean ± SEM (wound, p=0.754; remote, p=0.602; Mann–Whitney U test). N.S., not significant. Single confocal slice images are shown in B and C. Scale bar, 50 μm.DOI: http://dx.doi.org/10.7554/eLife.24635.017
study
99.94
(A) Schematic of the experiment. (B) Semi-qRT-PCR analysis of shha in purified tcf21:DsRed2+ epicardial cells obtained from uninjured and injured (7 dpi) tcf21:DsRed hearts. Injury was confirmed by the induction of raldh2 expression. (C) Semi-qRT-PCR analysis of Hh pathway genes using purified tcf21:DsRed2+ epicardial cells obtained from 4-HT–treated 7 dpi tcf21:DsRed; shhact/ct (control, left) and tcf21:DsRed; tcf21:CreER; shhact/ct hearts (right). (D) Immunofluorescence images of the subepicardial (top) and trabecular areas (bottom) of heart sections obtained from 4-HT–treated 7 dpi shhact/ct (control) or tcf21:CreER; shhact/ct hearts. Brackets, subepicardial areas. Dotted lines, approximate amputation plane. Arrows indicate proliferating cardiomyocytes. (E) Quantification of cardiomyocyte proliferation in the subepicardial and trabecular areas of the heart sections obtained from 4-HT–treated 7 dpi shhact/ct (control) or tcf21:CreER; shhact/ct hearts shown in D (n = 6 each). The data are presented as the mean ± SEM (**p<0.01, Mann–Whitney U test). N.S., not significant (p=0.3367). (F) Image of heart sections obtained from 7 dpi gata4:EGFP; tcf21;DsRed2 fish. Subepicardial cardiomyocytes (green, arrows) associate with epicardial cells (magenta, arrowheads). (G) Semi-qRT-PCR analysis of shha pathway genes using purified subepicardial cardiomyocytes obtained from 7 dpi gata4:EGFP ventricles. Cardiomyocytes purified from uninjured cmlc2:EGFP ventricles were used as negative controls. Scale bar, 50 μm.
study
100.0
10.7554/eLife.24635.016Figure 5—figure supplement 1.Assessment of spontaneous cassette inversion.(A) Schematic of the inverted shhact allele and primers sites. (B) The inverted allele was not detected in untreated tcf21:CreER; shhact/ct embryos using 40 cycles of PCR. (C) The inverted allele was not detected in untreated adult tcf21:CreER; shhact/ct fish using 30 cycles of PCR, but it was detected when the cycle number was increased to 35. Each PCR sample used genomic DNA isolated from five pooled hearts as a template.DOI: http://dx.doi.org/10.7554/eLife.24635.016
study
100.0